throwing in the simplest fix i can think of while i ponder about what the real fix should be (#4559)

## Summary of the Pull Request
The issue seems to be how `SwapChainScaleChanged` gets fired and attempts to tell the renderer
to `UpdateDPI` when the renderer is gone. So, as a quick bandaid, we'll put a quick check to only do the thing if the renderer is alive.

## PR Checklist
* [x] Closes #4539
* [x] CLA signed. If not, go over [here](https://cla.opensource.microsoft.com/microsoft/Terminal) and sign the CLA
* [x] Tests added/passed

## Validation Steps Performed
Held my new tab button for about thirty seconds then held the close tab button until all tabs closed without a crash.

.clang-format

@@ -0,0 +1,93 @@
+
+AccessModifierOffset: -4
+AlignAfterOpenBracket: Align
+#AllowAllArgumentsOnNextLine: false
+AlignConsecutiveAssignments: false
+AlignConsecutiveDeclarations: false
+#AllowAllConstructorInitializersOnNextLine: false
+AlignEscapedNewlines: Left
+AlignOperands: true
+AlignTrailingComments: false
+AllowAllParametersOfDeclarationOnNextLine: false
+AllowShortFunctionsOnASingleLine: Inline
+AllowShortCaseLabelsOnASingleLine: false
+AllowShortIfStatementsOnASingleLine: false
+#AllowShortLambdasOnASingleLine: Inline
+AllowShortLoopsOnASingleLine: false
+AlwaysBreakAfterReturnType: None
+AlwaysBreakBeforeMultilineStrings: false
+AlwaysBreakTemplateDeclarations: Yes
+BinPackArguments: false
+BinPackParameters: false
+BraceWrapping:
+  AfterClass: true
+  AfterControlStatement: true
+  AfterEnum: true
+  AfterFunction: true
+  AfterNamespace: true
+  AfterObjCDeclaration: true
+  AfterStruct: true
+  AfterUnion: true
+  AfterExternBlock: false
+  BeforeCatch: true
+  BeforeElse: true
+  IndentBraces: false
+  SplitEmptyFunction: true
+  SplitEmptyRecord: true
+  SplitEmptyNamespace: true
+BreakBeforeBinaryOperators: None
+BreakBeforeBraces: Custom
+BreakBeforeTernaryOperators: false
+BreakConstructorInitializers: AfterColon
+BreakInheritanceList: AfterColon
+ColumnLimit: 0
+CommentPragmas: "suppress"
+CompactNamespaces: false
+ConstructorInitializerAllOnOneLineOrOnePerLine: true
+ConstructorInitializerIndentWidth: 4
+ContinuationIndentWidth: 4
+Cpp11BracedListStyle: false
+DerivePointerAlignment: false
+FixNamespaceComments: false
+IncludeBlocks: Regroup
+IncludeCategories:
+    - Regex: '^.*(precomp|pch|stdafx)'
+      Priority: -1
+    - Regex: '^".*"'
+      Priority: 1
+    - Regex: '^<.*>'
+      Priority: 2
+    - Regex: '.*'
+      Priority: 3
+IndentCaseLabels: false
+IndentPPDirectives: None
+IndentWidth: 4
+IndentWrappedFunctionNames: false
+KeepEmptyLinesAtTheStartOfBlocks: false
+MacroBlockBegin: "BEGIN_TEST_METHOD_PROPERTIES|BEGIN_MODULE|BEGIN_TEST_CLASS|BEGIN_TEST_METHOD"
+MacroBlockEnd: "END_TEST_METHOD_PROPERTIES|END_MODULE|END_TEST_CLASS|END_TEST_METHOD"
+MaxEmptyLinesToKeep: 1
+NamespaceIndentation: All
+PointerAlignment: Left
+ReflowComments: false
+SortIncludes: false
+SortUsingDeclarations: true
+SpaceAfterCStyleCast: false
+#SpaceAfterLogicalNot: false
+SpaceAfterTemplateKeyword: false
+SpaceBeforeAssignmentOperators: true
+SpaceBeforeCpp11BracedList: false
+SpaceBeforeCtorInitializerColon: true
+SpaceBeforeInheritanceColon: true
+SpaceBeforeParens: ControlStatements
+SpaceBeforeRangeBasedForLoopColon: true
+SpaceInEmptyParentheses: false
+SpacesBeforeTrailingComments: 1
+SpacesInAngles: false
+SpacesInCStyleCastParentheses: false
+SpacesInContainerLiterals: false
+SpacesInParentheses: false
+SpacesInSquareBrackets: false
+Standard: Cpp11
+TabWidth: 4
+UseTab: Never

.editorconfig

@@ -0,0 +1,13 @@
+root = true
+
+[*]
+trim_trailing_whitespace = true
+insert_final_newline = true
+
+[{*.cpp,*.c,*.hpp,*.h,*.cs}]
+indent_style = space
+indent_size = 4
+
+[*.json]
+indent_style = space
+indent_size = 2

.github/ISSUE_TEMPLATE/Bug_Report.md

@@ -0,0 +1,54 @@
+---
+name: "Bug report 🐛"
+about: Report errors or unexpected behavior
+title: ''
+labels: ''
+assignees: ''
+
+---
+
+<!-- 
+🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨
+
+I ACKNOWLEDGE THE FOLLOWING BEFORE PROCEEDING:
+1. If I delete this entire template and go my own path, the core team may close my issue without further explanation or engagement.
+2. If I list multiple bugs/concerns in this one issue, the core team may close my issue without further explanation or engagement.
+3. If I write an issue that has many duplicates, the core team may close my issue without further explanation or engagement (and without necessarily spending time to find the exact duplicate ID number).
+4. If I leave the title incomplete when filing the issue, the core team may close my issue without further explanation or engagement.
+5. If I file something completely blank in the body, the core team may close my issue without further explanation or engagement.
+
+All good? Then proceed!
+-->
+
+<!--
+This bug tracker is monitored by Windows Terminal development team and other technical folks.
+
+**Important: When reporting BSODs or security issues, DO NOT attach memory dumps, logs, or traces to Github issues**.
+Instead, send dumps/traces to secure@microsoft.com, referencing this GitHub issue.
+
+If this is an application crash, please also provide a Feedback Hub submission link so we can find your diagnostic data on the backend. Use the category "Apps > Windows Terminal (Preview)" and choose "Share My Feedback" after submission to get the link.
+
+Please use this form and describe your issue, concisely but precisely, with as much detail as possible.
+
+-->
+
+# Environment
+
+```none
+Windows build number: [run `[Environment]::OSVersion` for powershell, or `ver` for cmd]
+Windows Terminal version (if applicable):
+
+Any other software?
+```
+
+# Steps to reproduce
+
+<!-- A description of how to trigger this bug. -->
+
+# Expected behavior
+
+<!-- A description of what you're expecting, possibly containing screenshots or reference material. -->
+
+# Actual behavior
+
+<!-- What's actually happening? -->

.github/ISSUE_TEMPLATE/Documentation_Issue.md

@@ -0,0 +1,10 @@
+---
+name: "Documentation Issue 📚"
+about: Report issues in our documentation
+title: ''
+labels: Issue-Docs
+assignees: ''
+
+---
+
+<!-- Briefly describe which document needs to be corrected and why. -->

.github/ISSUE_TEMPLATE/Feature_Request.md

@@ -0,0 +1,35 @@
+---
+name: "Feature Request/Idea 🚀"
+about: Suggest a new feature or improvement (this does not mean you have to implement
+  it)
+title: ''
+labels: Issue-Feature
+assignees: ''
+
+---
+
+<!-- 
+🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨
+
+I ACKNOWLEDGE THE FOLLOWING BEFORE PROCEEDING:
+1. If I delete this entire template and go my own path, the core team may close my issue without further explanation or engagement.
+2. If I list multiple bugs/concerns in this one issue, the core team may close my issue without further explanation or engagement.
+3. If I write an issue that has many duplicates, the core team may close my issue without further explanation or engagement (and without necessarily spending time to find the exact duplicate ID number).
+4. If I leave the title incomplete when filing the issue, the core team may close my issue without further explanation or engagement.
+5. If I file something completely blank in the body, the core team may close my issue without further explanation or engagement.
+
+All good? Then proceed!
+-->
+
+# Description of the new feature/enhancement
+
+<!-- 
+A clear and concise description of what the problem is that the new feature would solve.
+Describe why and how a user would use this new functionality (if applicable).
+-->
+
+# Proposed technical implementation details (optional)
+
+<!-- 
+A clear and concise description of what you want to happen.
+-->

.github/PULL_REQUEST_TEMPLATE.md

@@ -0,0 +1,19 @@
+<!-- Enter a brief description/summary of your PR here. What does it fix/what does it change/how was it tested (even manually, if necessary)? -->
+## Summary of the Pull Request
+
+<!-- Other than the issue solved, is this relevant to any other issues/existing PRs? --> 
+## References
+
+<!-- Please review the items on the PR checklist before submitting-->
+## PR Checklist
+* [ ] Closes #xxx
+* [ ] CLA signed. If not, go over [here](https://cla.opensource.microsoft.com/microsoft/Terminal) and sign the CLA
+* [ ] Tests added/passed
+* [ ] Requires documentation to be updated
+* [ ] I've discussed this with core contributors already. If not checked, I'm ready to accept this work might be rejected in favor of a different grand plan. Issue number where discussion took place: #xxx
+
+<!-- Provide a more detailed description of the PR, other things fixed or any additional comments/features here -->
+## Detailed Description of the Pull Request / Additional comments
+
+<!-- Describe how you validated the behavior. Add automated tests wherever possible, but list manual validation steps taken as well -->
+## Validation Steps Performed

.gitignore

@@ -17,10 +17,11 @@
 [Rr]eleases/
 x64/
 x86/
-build/
+ARM64/
 bld/
 [Bb]in/
 [Oo]bj/
+[Ll]og/
 objfre/
 objchk/
 
@@ -79,14 +80,18 @@ _Chutzpah*
 ipch/
 *.aps
 *.ncb
+*.opendb
 *.opensdf
 *.sdf
 *.cachefile
+*.VC.db
+*.VC.VC.opendb
 
 # Visual Studio profiler
 *.psess
 *.vsp
 *.vspx
+*.sap
 
 # TFS 2012 Local Workspace
 $tf/
@@ -111,6 +116,7 @@ _TeamCity*
 # NCrunch
 _NCrunch_*
 .*crunch*.local.xml
+nCrunchTemp_*
 
 # MightyMoose
 *.mm.*
@@ -138,13 +144,16 @@ publish/
 # Publish Web Output
 *.[Pp]ublish.xml
 *.azurePubxml
-## TODO: Comment the next line if you want to checkin your
-## web deploy settings but do note that will include unencrypted
-## passwords
-#*.pubxml
-
+# TODO: Comment the next line if you want to checkin your web deploy settings
+# but database connection strings (with potential passwords) will be unencrypted
+*.pubxml
 *.publishproj
 
+# Microsoft Azure Web App publish settings. Comment the next line if you want to
+# checkin your Azure Web App publish settings, but sensitive information contained
+# in these scripts will be unencrypted
+PublishScripts/
+
 # NuGet Packages
 *.nupkg
 # The packages folder can be ignored because of Package Restore
@@ -153,13 +162,23 @@ publish/
 !**/packages/build/
 # Uncomment if necessary however generally it will be regenerated when needed
 #!**/packages/repositories.config
+# NuGet v3's project.json files produces more ignorable files
+*.nuget.props
+*.nuget.targets
 
-# Windows Azure Build Output
+# Microsoft Azure Build Output
 csx/
 *.build.csdef
 
-# Windows Store app package directory
+# Microsoft Azure Emulator
+ecf/
+rcf/
+
+# Windows Store app package directories and files
 AppPackages/
+BundleArtifacts/
+Package.StoreAssociation.xml
+_pkginfo.txt
 
 # Visual Studio cache files
 # files ending in .cache can be ignored
@@ -211,12 +230,25 @@ FakesAssemblies/
 # Visual Studio 6 workspace options file
 *.opt
 
-# LightSwitch generated files
-GeneratedArtifacts/
-_Pvt_Extensions/
-ModelManifest.xml
-*.opendb
-*.db
+# Visual Studio LightSwitch build output
+**/*.HTMLClient/GeneratedArtifacts
+**/*.DesktopClient/GeneratedArtifacts
+**/*.DesktopClient/ModelManifest.xml
+**/*.Server/GeneratedArtifacts
+**/*.Server/ModelManifest.xml
+_Pvt_Extensions
+
+# Paket dependency manager
+.paket/paket.exe
+paket-files/
+
+# FAKE - F# Make
+.fake/
+
+# JetBrains Rider
+.idea/
+*.sln.iml
+*.exe
 
 # Windows Build System files
 build*.dbb
@@ -229,8 +261,14 @@ build*.rec
 build*.wrn
 build*.metadata
 
+# MS Build binary logs
+*.binlog
+
 # .razzlerc.cmd file - used by dev environment
 tools/.razzlerc.*
+# .PowershellModules - if one needs a powershell module dependency, one
+# can save it here. used by tools/OpenConsole.psm1
+.PowershellModules
 # message compiler output
 MSG*.bin
 /*.exe
@@ -238,8 +276,9 @@ MSG*.bin
 # python
 *.pyc
 
-**Generated Files/
+**/Generated Files/
 **/Merged/*
 **/Unmerged/*
 profiles.json
 *.metaproj
+*.swp

.gitmodules

@@ -1,6 +1,6 @@
 [submodule "dep/gsl"]
 	path = dep/gsl
-	url = https://github.com/Microsoft/gsl
+	url = https://github.com/microsoft/gsl
 [submodule "dep/wil"]
 	path = dep/wil
-	url = https://github.com/Microsoft/wil
+	url = https://github.com/microsoft/wil

.nuget/packages.config

@@ -0,0 +1,4 @@
+<?xml version="1.0" encoding="utf-8"?>
+<packages>
+  <package id="vswhere" version="2.6.7" />
+</packages>

.vsconfig

@@ -0,0 +1,34 @@
+{
+  "version": "1.0",
+  "components": [
+    "Microsoft.VisualStudio.Component.CoreEditor",
+    "Microsoft.VisualStudio.Workload.CoreEditor",
+    "Microsoft.VisualStudio.Workload.Universal",
+    "Microsoft.VisualStudio.Workload.NativeDesktop",
+    "Microsoft.VisualStudio.Workload.ManagedDesktop",
+    "Microsoft.VisualStudio.Component.ManagedDesktop.Prerequisites",
+    "Microsoft.VisualStudio.Component.NuGet",
+    "Microsoft.VisualStudio.Component.Roslyn.Compiler",
+    "Microsoft.VisualStudio.Component.Roslyn.LanguageServices",
+    "Microsoft.Net.ComponentGroup.DevelopmentPrerequisites",
+    "Microsoft.Component.MSBuild",
+    "Microsoft.VisualStudio.Component.ManagedDesktop.Core",
+    "Microsoft.Net.Component.4.TargetingPack",
+    "Microsoft.Net.Component.4.5.TargetingPack",
+    "Microsoft.VisualStudio.Component.DiagnosticTools",
+    "Microsoft.VisualStudio.Component.Debugger.JustInTime",
+    "Microsoft.VisualStudio.Component.Windows10SDK.18362",
+    "Microsoft.VisualStudio.ComponentGroup.UWP.Support",
+    "Microsoft.VisualStudio.Component.VC.CoreIde",
+    "Microsoft.VisualStudio.ComponentGroup.NativeDesktop.Core",
+    "Microsoft.VisualStudio.Component.Graphics",
+    "Microsoft.VisualStudio.Component.VC.Redist.14.Latest",
+    "Microsoft.VisualStudio.Component.VC.Tools.x86.x64",
+    "Microsoft.VisualStudio.Component.VC.Tools.ARM64",
+    "Microsoft.VisualStudio.Component.VC.v142.x86.x64",
+    "Microsoft.VisualStudio.Component.VC.v142.ARM64",
+    "Microsoft.VisualStudio.ComponentGroup.UWP.VC",
+    "Microsoft.VisualStudio.ComponentGroup.UWP.VC.v142",
+    "Microsoft.VisualStudio.Component.UWP.VC.ARM64"
+  ]
+}

CODE_OF_CONDUCT.md

@@ -0,0 +1,8 @@
+# Code of Conduct
+
+This project has adopted the [Microsoft Open Source Code of Conduct][conduct-code].
+For more information see the [Code of Conduct FAQ][conduct-FAQ] or contact [opencode@microsoft.com][conduct-email] with any additional questions or comments.
+
+[conduct-code]: https://opensource.microsoft.com/codeofconduct/
+[conduct-FAQ]: https://opensource.microsoft.com/codeofconduct/faq/
+[conduct-email]: mailto:opencode@microsoft.com

CONTRIBUTING.md

@@ -0,0 +1,158 @@
+# Terminal Contributor's Guide
+
+Below is our guidance for how to report issues, propose new features, and submit contributions via Pull Requests (PRs).
+
+## Open Development Workflow
+
+The Windows Terminal team is VERY active in this GitHub Repo. In fact, we live in it all day long and carry out all our development in the open!
+
+When the team finds issues we file them in the repo. When we propose new ideas or think-up new features, we file new feature requests. When we work on fixes or features, we create branches and work on those improvements. And when PRs are reviewed, we review in public - including all the good, the bad, and the ugly parts.
+
+The point of doing all this work in public is to ensure that we are holding ourselves to a high degree of transparency, and so that the community sees that we apply the same processes and hold ourselves to the same quality-bar as we do to community-submitted issues and PRs. We also want to make sure that we expose our team culture and "tribal knowledge" that is inherent in any closely-knit team, which often contains considerable value to those new to the project who are trying to figure out "why the heck does this thing look/work like this???"
+
+### Repo Bot
+
+The team triages new issues several times a week. During triage, the team uses labels to categorize, manage, and drive the project workflow.
+
+We employ [a bot engine](https://github.com/microsoft/terminal/blob/master/doc/bot.md) to help us automate common processes within our workflow.
+
+We drive the bot by tagging issues with specific labels which cause the bot engine to close issues, merge branches, etc. This bot engine helps us keep the repo clean by automating the process of notifying appropriate parties if/when information/follow-up is needed, and closing stale issues/PRs after reminders have remained unanswered for several days.
+
+Therefore, if you do file issues, or create PRs, please keep an eye on your GitHub notifications. If you do not respond to requests for information, your issues/PRs may be closed automatically.
+
+---
+## Reporting Security Issues
+
+**Please do not report security vulnerabilities through public GitHub issues.** Instead, please report them to the Microsoft Security Response Center (MSRC). See [SECURITY.md](./SECURITY.md) for more information.
+
+## Before you start, file an issue
+
+Please follow this simple rule to help us eliminate any unnecessary wasted effort & frustration, and ensure an efficient and effective use of everyone's time - yours, ours, and other community members':
+
+> 👉 If you have a question, think you've discovered an issue, would like to propose a new feature, etc., then find/file an issue **BEFORE** starting work to fix/implement it.
+
+### Search existing issues first
+
+Before filing a new issue, search existing open and closed issues first: This project is moving fast! It is likely someone else has found the problem you're seeing, and someone may be working on or have already contributed a fix!
+
+If no existing item describes your issue/feature, great - please file a new issue:
+
+### File a new Issue
+
+* Don't know whether you're reporting an issue or requesting a feature? File an issue
+* Have a question that you don't see answered in docs, videos, etc.? File an issue
+* Want to know if we're planning on building a particular feature? File an issue
+* Got a great idea for a new feature? File an issue/request/idea
+* Don't understand how to do something? File an issue/Community Guidance Request
+* Found an existing issue that describes yours? Great - upvote and add additional commentary / info / repro-steps / etc.
+
+When you hit "New Issue", select the type of issue closest to what you want to report/ask/request:
+![New issue types](/doc/images/new-issue-template.png)
+
+### Complete the template
+
+**Complete the information requested in the issue template, providing as much information as possible**. The more information you provide, the more likely your issue/ask will be understood and implemented. Helpful information includes:
+
+* What device you're running (inc. CPU type, memory, disk, etc.)
+* What build of Windows your device is running
+
+  👉 Tip: Run the following in PowerShell Core
+
+  ```powershell
+  C:\> $PSVersionTable.OS
+  Microsoft Windows 10.0.18909
+  ```
+
+  ... or in Windows PowerShell
+
+  ```powershell
+  C:\> $PSVersionTable.BuildVersion
+
+  Major  Minor  Build  Revision
+  -----  -----  -----  --------
+  10     0      18912  1001
+  ```
+
+  ... or Cmd:
+
+  ```cmd
+  C:\> ver
+
+  Microsoft Windows [Version 10.0.18900.1001]
+  ```
+
+* What tools and apps you're using (e.g. VS 2019, VSCode, etc.)
+* Don't assume we're experts in setting up YOUR environment and don't assume we are experts in `<your distro/tool of choice>`. Teach us to help you!
+* **We LOVE detailed repro steps!** What steps do we need to take to reproduce the issue? Assume we love to read repro steps. As much detail as you can stand is probably _barely_ enough detail for us!
+* If you're reporting a particular character/glyph not rendering correctly, the specific Unicode codepoint would be MOST welcome (e.g. U+1F4AF, U+4382)
+* Prefer error message text where possible or screenshots of errors if text cannot be captured
+* We MUCH prefer text command-line script than screenshots of command-line script.
+* **If you intend to implement the fix/feature yourself then say so!** If you do not indicate otherwise we will assume that the issue is our to solve, or may label the issue as `Help-Wanted`.
+
+### DO NOT post "+1" comments
+
+> ⚠ DO NOT post "+1", "me too", or similar comments - they just add noise to an issue.
+
+If you don't have any additional info/context to add but would like to indicate that you're affected by the issue, upvote the original issue by clicking its [+😊] button and hitting 👍 (+1) icon. This way we can actually measure how impactful an issue is.
+
+---
+
+## Contributing fixes / features
+
+For those able & willing to help fix issues and/or implement features ...
+
+### To Spec or not to Spec
+
+Some issues/features may be quick and simple to describe and understand. For such scenarios, once a team member has agreed with your approach, skip ahead to the section headed "Fork, Branch, and Create your PR", below.
+
+Small issues that do not require a spec will be labelled Issue-Bug or Issue-Task.
+
+However, some issues/features will require careful thought & formal design before implementation. For these scenarios, we'll request that a spec is written and the associated issue will be labeled Issue-Feature.
+
+Specs help collaborators discuss different approaches to solve a problem, describe how the feature will behave, how the feature will impact the user, what happens if something goes wrong, etc. Driving towards agreement in a spec, before any code is written, often results in simpler code, and less wasted effort in the long run.
+
+Specs will be managed in a very similar manner as code contributions so please follow the "Fork, Branch and Create your PR" below.
+
+### Writing / Contributing-to a Spec
+
+To write/contribute to a spec: fork, branch and commit via PRs, as you would with any code changes.
+
+Specs are written in markdown, stored under the `\doc\spec` folder and named `[issue id] - [spec description].md`.
+
+👉 **It is important to follow the spec templates and complete the requested information**. The available spec templates will help ensure that specs contain the minimum information & decisions necessary to permit development to begin. In particular, specs require you to confirm that you've already discussed the issue/idea with the team in an issue and that you provide the issue ID for reference.
+
+Team members will be happy to help review specs and guide them to completion.
+
+### Help Wanted
+
+Once the team have approved an issue/spec, development can proceed. If no developers are immediately available, the spec can be parked ready for a developer to get started. Parked specs' issues will be labeled "Help Wanted". To find a list of development opportunities waiting for developer involvement, visit the Issues and filter on [the Help-Wanted label](https://github.com/microsoft/terminal/labels/Help-Wanted).
+
+---
+
+## Development
+
+### Fork, Clone, Branch and Create your PR
+
+Once you've discussed your proposed feature/fix/etc. with a team member, and you've agreed an approach or a spec has been written and approved, it's time to start development:
+
+1. Fork the repo if you haven't already
+1. Clone your fork locally
+1. Create & push a feature branch
+1. Create a [Draft Pull Request (PR)](https://github.blog/2019-02-14-introducing-draft-pull-requests/)
+1. Work on your changes
+
+### Code Review
+
+When you'd like the team to take a look, (even if the work is not yet fully-complete), mark the PR as 'Ready For Review' so that the team can review your work and provide comments, suggestions, and request changes. It may take several cycles, but the end result will be solid, testable, conformant code that is safe for us to merge.
+
+> ⚠ Remember: **changes you make may affect both Windows Terminal and Windows Console and may end up being re-incorporated into Windows itself!** Because of this, we will treat community PR's with the same level of scrutiny and rigor as commits submitted to the official Windows source by team members and partners.
+
+### Merge
+
+Once your code has been reviewed and approved by the requisite number of team members, it will be merged into the master branch. Once merged, your PR will be automatically closed.
+
+---
+
+## Thank you
+
+Thank you in advance for your contribution! Now, [what's next on the list](https://github.com/microsoft/terminal/labels/Help-Wanted)? 😜

NOTICE.md

@@ -0,0 +1,115 @@
+# NOTICES AND INFORMATION
+Do Not Translate or Localize
+
+This software incorporates material from third parties. Microsoft makes certain
+open source code available at http://3rdpartysource.microsoft.com, or you may
+send a check or money order for US $5.00, including the product name, the open
+source component name, and version number, to:
+
+```
+Source Code Compliance Team
+Microsoft Corporation
+One Microsoft Way
+Redmond, WA 98052
+USA
+```
+
+Notwithstanding any other terms, you may reverse engineer this software to the
+extent required to debug changes to any libraries licensed under the GNU Lesser
+General Public License.
+
+## jsoncpp
+
+**Source**: https://github.com/open-source-parsers/jsoncpp
+
+### License
+
+```
+Copyright (c) 2007-2010 Baptiste Lepilleur and The JsonCpp Authors
+
+Permission is hereby granted, free of charge, to any person
+obtaining a copy of this software and associated documentation
+files (the "Software"), to deal in the Software without
+restriction, including without limitation the rights to use, copy,
+modify, merge, publish, distribute, sublicense, and/or sell copies
+of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be
+included in all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
+```
+
+## telnetpp
+
+**Source**: https://github.com/KazDragon/telnetpp
+
+### License 
+
+```
+The MIT License (MIT)
+
+Copyright (c) 2015-2017 Matthew Chaplain a.k.a KazDragon
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
+```
+
+## chromium/base/numerics
+
+**Source**: 
+
+### License
+
+```
+Copyright 2015 The Chromium Authors. All rights reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are
+met:
+
+   * Redistributions of source code must retain the above copyright
+notice, this list of conditions and the following disclaimer.
+   * Redistributions in binary form must reproduce the above
+copyright notice, this list of conditions and the following disclaimer
+in the documentation and/or other materials provided with the
+distribution.
+   * Neither the name of Google Inc. nor the names of its
+contributors may be used to endorse or promote products derived from
+this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+```

NuGet.Config

@@ -1,14 +1,20 @@
 <?xml version="1.0" encoding="utf-8"?>
 <configuration>
-  <packageSources>
-    <add key="NuGet.org" value="https://api.nuget.org/v3/index.json" />
-    <!-- Add repositories here to the list of available repositories -->
-    
-    <!-- Dependencies that we must carry because they're not on public nuget feeds right now. -->
-    <add key="Static Package Dependencies" value="dep\packages" />
-    
-    <!-- Internal NuGet feeds that may not be accessible outside Microsoft corporate network -->
-    <!--<add key="TAEF - internal" value="https://microsoft.pkgs.visualstudio.com/DefaultCollection/_packaging/Taef/nuget/v3/index.json" />
-    <add key="OpenConsole - Internal" value="https://microsoft.pkgs.visualstudio.com/_packaging/OpenConsole/nuget/v3/index.json" />-->
-  </packageSources>
+    <packageSources>
+        <add key="NuGet.org" value="https://api.nuget.org/v3/index.json" />
+        <!-- Add repositories here to the list of available repositories -->
+        
+        <!-- Dependencies that we must carry because they're not on public nuget feeds right now. -->
+        <!--<add key="Static Package Dependencies" value="dep\packages" />-->
+        
+        <!-- Use our own NuGet Feed -->
+        <add key="Windows Terminal NuGet Feed" value="https://terminalnuget.blob.core.windows.net/feed/index.json" />
+        
+        <!-- Internal NuGet feeds that may not be accessible outside Microsoft corporate network -->
+        <!--<add key="TAEF - internal" value="https://microsoft.pkgs.visualstudio.com/DefaultCollection/_packaging/Taef/nuget/v3/index.json" />
+        <add key="OpenConsole - Internal" value="https://microsoft.pkgs.visualstudio.com/_packaging/OpenConsole/nuget/v3/index.json" />-->
+    </packageSources>
+    <config>
+        <add key="repositorypath" value=".\packages" />
+    </config>
 </configuration>

README.md

@@ -1,45 +1,221 @@
-# Welcome to the Console Project!
-
-This project is currently controlled by the Windows Developer Platform Tools & Runtimes' Open Source Software team (*WDG > DEP > DART > OSS*).
-
-Our team can be reached at `dartcon@microsoft.com`.
-
-The code is stored at <https://microsoft.visualstudio.com/Dart/_git/OpenConsole>.
-
-The area path within the Microsoft.VisualStudio.com database for our Work Items is `OS\CORE-OS Core\DEP-Developer Ecosystem Platform\DART-Developer Tools and Runtimes\Open Source Software\Console`.
-
-## Jumping In
-
-To get started, feel free to read up on some of our documentation on the way we get things done and hop in.
-
-Make a branch off of `dev/main` for yourself of the pattern `dev/myalias/foo` and feel free to push it to the server to get automatic builds and unit test runs.
-
-Choose a bit of code to clean up, try to add a new feature, or improve something that you try to use every day.
-
-When you are ready, use the [web portal](https://microsoft.visualstudio.com/Dart/_git/OpenConsole/pullrequests) to send a pull request into our `dev/main` branch and we'll be happy to help you get your code in line with the rest of the console.
-
-## Building
-
-OpenConsole uses submodules for some of its dependencies. To make sure submodules are restored or updated:
-
-```
-git submodule update --init --recursive
-```
-
-OpenConsole.sln may be built from within Visual Studio or from the command line using msbuild. To build from the command line:
-
-```
-nuget.exe restore OpenConsole.sln
-msbuild.exe OpenConsole.sln
-```
-
-We provide a set of convienence scripts in the /tools directory to help automate the process of building and running tests.
-
-## Assorted Notes
-
-Here's some assorted notes on the way we do things. If you learn something about how we do things, feel free to contribute to any of our documentation files anywhere in the repository (or make some new ones!) This is a work in progress as we try to learn what we'll need to train people on in order to be effective contributors to our project. We're pretty blind to these things after staring at this code for so long... so mind the gaps and ask us plenty of questions!
-
-* [Coding Style](./doc/STYLE.md)
-* [Code Organization](./doc/ORGANIZATION.md)
-* [Exceptions in our legacy codebase](./doc/EXCEPTIONS.md)
-* [Helpful smart pointers and macros for interfacing with Windows in WIL](./doc/WIL.md)
+# Welcome to the Windows Terminal, Console and Command-Line repo
+
+This repository contains the source code for:
+
+* [Windows Terminal](https://www.microsoft.com/en-us/p/windows-terminal-preview/9n0dx20hk701)
+* The Windows console host (`conhost.exe`)
+* Components shared between the two projects
+* [ColorTool](https://github.com/Microsoft/Terminal/tree/master/src/tools/ColorTool)
+* [Sample projects](https://github.com/Microsoft/Terminal/tree/master/samples) that show how to consume the Windows Console APIs
+
+Related repositories include:
+
+* [Console API Documentation](https://github.com/MicrosoftDocs/Console-Docs)
+* [Cascadia Code Font](https://github.com/Microsoft/Cascadia-Code)
+
+## Installing and running Windows Terminal
+
+> 👉 Note: Windows Terminal requires Windows 10 1903 (build 18362) or later
+
+### Manually installing builds from this repository
+
+For users who are unable to install Terminal from the Microsoft Store, Terminal builds can be manually downloaded from this repository's [Releases page](https://github.com/microsoft/terminal/releases).
+
+> ⚠ Note: If you install Terminal manually:
+>
+> * Be sure to install the [Desktop Bridge VC++ v14 Redistributable Package](https://www.microsoft.com/en-us/download/details.aspx?id=53175) otherwise Terminal may not install and/or run and may crash at startup
+> * Terminal will not auto-update when new builds are released so you will need to regularly install the latest Terminal release to receive all the latest fixes and improvements!
+
+### Install via Chocolatey (unofficial)
+
+[Chocolatey](https://chocolatey.org) users can download and install the latest Terminal release by installing the `microsoft-windows-terminal` package:
+
+```powershell
+choco install microsoft-windows-terminal
+```
+
+To upgrade Windows Terminal using Chocolatey, run the following:
+
+```powershell
+choco upgrade microsoft-windows-terminal
+```
+
+If you have any issues when installing/upgrading the package please go to the [Windows Terminal package page](https://chocolatey.org/packages/microsoft-windows-terminal) and follow the [Chocolatey triage process](https://chocolatey.org/docs/package-triage-process)
+
+---
+
+## Project Build Status
+
+Project|Build Status
+---|---
+Terminal|[![Build Status](https://dev.azure.com/ms/Terminal/_apis/build/status/Terminal%20CI?branchName=master)](https://dev.azure.com/ms/Terminal/_build?definitionId=136)
+ColorTool|![](https://microsoft.visualstudio.com/_apis/public/build/definitions/c93e867a-8815-43c1-92c4-e7dd5404f1e1/17023/badge)
+
+---
+
+## Windows Terminal v1.0 Roadmap
+
+The plan for delivering Windows Terminal v1.0 [is described here](/doc/terminal-v1-roadmap.md), and will be updated as the project proceeds.
+
+---
+
+## Terminal & Console Overview
+
+Please take a few minutes to review the overview below before diving into the code:
+
+### Windows Terminal
+
+Windows Terminal is a new, modern, feature-rich, productive terminal application for command-line users. It includes many of the features most frequently requested by the Windows command-line community including support for tabs, rich text, globalization, configurability, theming & styling, and more.
+
+The Terminal will also need to meet our goals and measures to ensure it remains fast and efficient, and doesn't consume vast amounts of memory or power.
+
+### The Windows Console Host
+
+The Windows Console host, `conhost.exe`, is Windows' original command-line user experience. It also hosts Windows' command-line infrastructure and the Windows Console API server, input engine, rendering engine, user preferences, etc. The console host code in this repository is the actual source from which the `conhost.exe` in Windows itself is built.
+
+Since taking ownership of the Windows command-line in 2014, the team added several new features to the Console, including background transparency, line-based selection, support for [ANSI / Virtual Terminal sequences](https://en.wikipedia.org/wiki/ANSI_escape_code), [24-bit color](https://devblogs.microsoft.com/commandline/24-bit-color-in-the-windows-console/), a [Pseudoconsole ("ConPTY")](https://devblogs.microsoft.com/commandline/windows-command-line-introducing-the-windows-pseudo-console-conpty/), and more.
+
+However, because Windows Console's primary goal is to maintain backward compatibility, we have been unable to add many of the features the community (and the team) have been wanting for the last several years including tabs, unicode text, and emoji.
+
+These limitations led us to create the new Windows Terminal.
+
+> You can read more about the evolution of the command-line in general, and the Windows command-line specifically in [this accompanying series of blog posts](https://devblogs.microsoft.com/commandline/windows-command-line-backgrounder/) on the Command-Line team's blog.
+
+### Shared Components
+
+While overhauling Windows Console, we modernized its codebase considerably, cleanly separating logical entities into modules and classes, introduced some key extensibility points, replaced several old, home-grown collections and containers with safer, more efficient [STL containers](https://docs.microsoft.com/en-us/cpp/standard-library/stl-containers?view=vs-2019), and made the code simpler and safer by using Microsoft's [Windows Implementation Libraries - WIL](https://github.com/Microsoft/wil).
+
+This overhaul resulted in several of Console's key components being available for re-use in any terminal implementation on Windows. These components include a new DirectWrite-based text layout and rendering engine, a text buffer capable of storing both UTF-16 and UTF-8, a VT parser/emitter, and more.
+
+### Creating the new Windows Terminal
+
+When we started planning the new Windows Terminal application, we explored and evaluated several approaches and technology stacks. We ultimately decided that our goals would be best met by continuing our investment in our C++ codebase, which would allow us to reuse several of the aforementioned modernized components in both the existing Console and the new Terminal. Further, we realized that this would allow us to build much of the Terminal's core itself as a reusable UI control that others can incorporate into their own applications.
+
+The result of this work is contained within this repo and delivered as the Windows Terminal application you can download from the Microsoft Store, or [directly from this repo's releases](https://github.com/microsoft/terminal/releases).
+
+---
+
+## Resources
+
+For more information about Windows Terminal, you may find some of these resources useful and interesting:
+
+* [Command-Line Blog](https://devblogs.microsoft.com/commandline)
+* [Command-Line Backgrounder Blog Series](https://devblogs.microsoft.com/commandline/windows-command-line-backgrounder/)
+* Windows Terminal Launch: [Terminal "Sizzle Video"](https://www.youtube.com/watch?v=8gw0rXPMMPE&list=PLEHMQNlPj-Jzh9DkNpqipDGCZZuOwrQwR&index=2&t=0s)
+* Windows Terminal Launch: [Build 2019 Session](https://www.youtube.com/watch?v=KMudkRcwjCw)
+* Run As Radio: [Show 645 - Windows Terminal with Richard Turner](http://www.runasradio.com/Shows/Show/645)
+* Azure Devops Podcast: [Episode 54 - Kayla Cinnamon and Rich Turner on DevOps on the Windows Terminal](http://azuredevopspodcast.clear-measure.com/kayla-cinnamon-and-rich-turner-on-devops-on-the-windows-terminal-team-episode-54)
+* Microsoft Ignite 2019 Session: [The Modern Windows Command Line: Windows Terminal - BRK3321](https://myignite.techcommunity.microsoft.com/sessions/81329?source=sessions)
+
+---
+
+## FAQ
+
+### I built and ran the new Terminal, but it looks just like the old console
+
+Cause: You're launching the incorrect solution in Visual Studio.
+
+Solution: Make sure you're building & deploying the `CascadiaPackage` project in Visual Studio.
+
+> ⚠ Note: `OpenConsole.exe` is just a locally-built `conhost.exe`, the classic Windows Console that hosts Windows' command-line infrastructure. OpenConsole is used by Windows Terminal to connect to and communicate with command-line applications (via [ConPty](https://devblogs.microsoft.com/commandline/windows-command-line-introducing-the-windows-pseudo-console-conpty/)).
+
+---
+
+## Documentation
+
+All project documentation is located in the `./doc` folder. If you would like to contribute to the documentation, please submit a pull request.
+
+---
+
+## Contributing
+
+We are excited to work alongside you, our amazing community, to build and enhance Windows Terminal\!
+
+***BEFORE you start work on a feature/fix***, please read & follow our [Contributor's Guide](https://github.com/microsoft/terminal/blob/master/CONTRIBUTING.md) to help avoid any wasted or duplicate effort.
+
+## Communicating with the Team
+
+The easiest way to communicate with the team is via GitHub issues.
+
+Please file new issues, feature requests and suggestions, but **DO search for similar open/closed pre-existing issues before creating a new issue.**
+
+If you would like to ask a question that you feel doesn't warrant an issue (yet), please reach out to us via Twitter:
+
+* Kayla Cinnamon, Program Manager: [@cinnamon\_msft](https://twitter.com/cinnamon_msft)
+* Rich Turner, Program Manager: [@richturn\_ms](https://twitter.com/richturn_ms)
+* Dustin Howett, Engineering Lead: [@dhowett](https://twitter.com/DHowett)
+* Michael Niksa, Senior Developer: [@michaelniksa](https://twitter.com/MichaelNiksa)
+* Mike Griese, Developer: [@zadjii](https://twitter.com/zadjii)
+* Carlos Zamora, Developer: [@cazamor_msft](https://twitter.com/cazamor_msft)
+* Leon Liang, Developer: [@leonmsft](https://twitter.com/leonmsft)
+
+## Developer Guidance
+
+## Prerequisites
+
+* You must be running Windows 1903 (build >= 10.0.18362.0) or later to run Windows Terminal
+* You must [enable Developer Mode in the Windows Settings app](https://docs.microsoft.com/en-us/windows/uwp/get-started/enable-your-device-for-development) to locally install and run Windows Terminal
+* You must have the [Windows 10 1903 SDK](https://developer.microsoft.com/en-us/windows/downloads/windows-10-sdk) installed
+* You must have at least [VS 2019](https://visualstudio.microsoft.com/downloads/) installed
+* You must install the following Workloads via the VS Installer. Note: Opening the solution in VS 2019 will [prompt you to install missing components automatically](https://devblogs.microsoft.com/setup/configure-visual-studio-across-your-organization-with-vsconfig/):
+  * Desktop Development with C++
+  * Universal Windows Platform Development
+  * **The following Individual Components**
+    * C++ (v142) Universal Windows Platform Tools
+
+## Building the Code
+
+This repository uses [git submodules](https://git-scm.com/book/en/v2/Git-Tools-Submodules) for some of its dependencies. To make sure submodules are restored or updated, be sure to run the following prior to building:
+
+```shell
+git submodule update --init --recursive
+```
+
+OpenConsole.sln may be built from within Visual Studio or from the command-line using a set of convenience scripts & tools in the **/tools** directory:
+
+### Building in PowerShell
+
+```powershell
+Import-Module .\tools\OpenConsole.psm1
+Set-MsBuildDevEnvironment
+Invoke-OpenConsoleBuild
+```
+
+### Building in Cmd
+
+```shell
+.\tools\razzle.cmd
+bcz
+```
+
+## Running & Debugging
+
+To debug the Windows Terminal in VS, right click on `CascadiaPackage` (in the Solution Explorer) and go to properties. In the Debug menu, change "Application process" and "Background task process" to "Native Only".
+
+You should then be able to build & debug the Terminal project by hitting <kbd>F5</kbd>.
+
+> 👉 You will _not_ be able to launch the Terminal directly by running the WindowsTerminal.exe. For more details on why, see [#926](https://github.com/microsoft/terminal/issues/926), [#4043](https://github.com/microsoft/terminal/issues/4043)
+
+### Coding Guidance
+
+Please review these brief docs below about our coding practices.
+
+> 👉 If you find something missing from these docs, feel free to contribute to any of our documentation files anywhere in the repository (or write some new ones!)
+
+This is a work in progress as we learn what we'll need to provide people in order to be effective contributors to our project.
+
+* [Coding Style](https://github.com/Microsoft/Terminal/blob/master/doc/STYLE.md)
+* [Code Organization](https://github.com/Microsoft/Terminal/blob/master/doc/ORGANIZATION.md)
+* [Exceptions in our legacy codebase](https://github.com/Microsoft/Terminal/blob/master/doc/EXCEPTIONS.md)
+* [Helpful smart pointers and macros for interfacing with Windows in WIL](https://github.com/Microsoft/Terminal/blob/master/doc/WIL.md)
+
+---
+
+# Code of Conduct
+
+This project has adopted the [Microsoft Open Source Code of Conduct][conduct-code].
+For more information see the [Code of Conduct FAQ][conduct-FAQ] or contact [opencode@microsoft.com][conduct-email] with any additional questions or comments.
+
+[conduct-code]: https://opensource.microsoft.com/codeofconduct/
+[conduct-FAQ]: https://opensource.microsoft.com/codeofconduct/faq/
+[conduct-email]: mailto:opencode@microsoft.com

SECURITY.md

@@ -0,0 +1,41 @@
+<!-- BEGIN MICROSOFT SECURITY.MD V0.0.2 BLOCK -->
+
+## Security
+
+Microsoft takes the security of our software products and services seriously, which includes all source code repositories managed through our GitHub organizations, which include [Microsoft](https://github.com/Microsoft), [Azure](https://github.com/Azure), [DotNet](https://github.com/dotnet), [AspNet](https://github.com/aspnet), [Xamarin](https://github.com/xamarin), and [many more](https://opensource.microsoft.com/).
+
+If you believe you have found a security vulnerability in any Microsoft-owned repository that meets Microsoft's [definition](https://docs.microsoft.com/en-us/previous-versions/tn-archive/cc751383(v=technet.10)) of a security vulnerability, please report it to us as described below.
+
+## Reporting Security Issues
+
+**Please do not report security vulnerabilities through public GitHub issues.**
+
+Instead, please report them to the Microsoft Security Response Center (MSRC) at [https://msrc.microsoft.com/create-report](https://msrc.microsoft.com/create-report).
+
+If you prefer to submit without logging in, send email to [secure@microsoft.com](mailto:secure@microsoft.com).  If possible, encrypt your message with our PGP key; please download it from the [Microsoft Security Response Center PGP Key page](https://www.microsoft.com/en-us/msrc/pgp-key-msrc).
+
+You should receive a response within 24 hours. If for some reason you do not, please follow up via email to ensure we received your original message. Additional information can be found at [microsoft.com/msrc](https://www.microsoft.com/msrc). 
+
+Please include the requested information listed below (as much as you can provide) to help us better understand the nature and scope of the possible issue:
+
+  * Type of issue (e.g. buffer overflow, SQL injection, cross-site scripting, etc.)
+  * Full paths of source file(s) related to the manifestation of the issue
+  * The location of the affected source code (tag/branch/commit or direct URL)
+  * Any special configuration required to reproduce the issue
+  * Step-by-step instructions to reproduce the issue
+  * Proof-of-concept or exploit code (if possible)
+  * Impact of the issue, including how an attacker might exploit the issue
+
+This information will help us triage your report more quickly.
+
+If you are reporting for a bug bounty, more complete reports can contribute to a higher bounty award. Please visit our [Microsoft Bug Bounty Program](https://microsoft.com/msrc/bounty) page for more details about our active programs.
+
+## Preferred Languages
+
+We prefer all communications to be in English.
+
+## Policy
+
+Microsoft follows the principle of [Coordinated Vulnerability Disclosure](https://www.microsoft.com/en-us/msrc/cvd).
+
+<!-- END MICROSOFT SECURITY.MD BLOCK -->

build/config/SignConfig.NuGet.xml

@@ -0,0 +1,5 @@
+<SignConfigXML>
+  <job platform="" configuration="" dest="__INPATHROOT__" jobname="EngFunSimpleSign" approvers="">
+   <file src="__INPATHROOT__\Microsoft.Terminal*.nupkg" signType="NuGet" /> 
+  </job>
+</SignConfigXML>

build/config/SignConfig.WindowsTerminal.xml

@@ -1,5 +1,6 @@
 <SignConfigXML>
   <job platform="" configuration="" certSubject="CN=Microsoft Corporation, O=Microsoft Corporation, L=Redmond, S=Washington, C=US" jobname="EngFunSimpleSign" approvers="">
    <file src="__INPATHROOT__\Microsoft.WindowsTerminal_8wekyb3d8bbwe.msixbundle" signType="136020001" dest="__OUTPATHROOT__\Microsoft.WindowsTerminal_8wekyb3d8bbwe.msixbundle" /> 
+   <file src="__INPATHROOT__\Microsoft.WindowsTerminalUniversal_8wekyb3d8bbwe.msixbundle" signType="136020001" dest="__OUTPATHROOT__\Microsoft.WindowsTerminalUniversal_8wekyb3d8bbwe.msixbundle" /> 
   </job>
 </SignConfigXML>

build/pipelines/ci.yml

@@ -0,0 +1,47 @@
+trigger:
+  batch: true
+  branches:
+    include:
+      - master
+  paths:
+    exclude:
+      - doc/*
+      - samples/*
+      - tools/*
+
+pr:
+  branches:
+    include:
+      - master
+  paths:
+    exclude:
+      - doc/*
+      - samples/*
+      - tools/*
+
+variables:
+  - name: runCodesignValidationInjectionBG
+    value: false
+
+#     0.0.yyMM.dd##
+#     0.0.1904.0900
+name: 0.0.$(Date:yyMM).$(Date:dd)$(Rev:rr)
+
+jobs:
+  - template: ./templates/build-console-audit-job.yml
+    parameters:
+      platform: x64
+
+  - template: ./templates/build-console-ci.yml
+    parameters:
+      platform: x64
+
+  - template: ./templates/build-console-ci.yml
+    parameters:
+      platform: x86
+
+  - template: ./templates/build-console-ci.yml
+    parameters:
+      platform: ARM64
+
+  - template: ./templates/check-formatting.yml

build/pipelines/release.yml

@@ -0,0 +1,40 @@
+trigger: none
+pr: none
+
+variables:
+  baseYearForVersioning: 2019 # Used by build-console-int
+  versionMajor: 0
+  versionMinor: 1
+
+# When we move off PackageES for Versioning, we'll need to switch
+# name to this format. For now, though, we need to use DayOfYear.Rev
+# to unique our builds, as mandated by PackageES's Setup task.
+# name: '$(versionMajor).$(versionMinor).$(DayOfYear)$(Rev:r).0'
+#
+# Build name/version number above must end with .0 to make the
+# store publication machinery happy.
+name: 'Terminal_$(date:yyMM).$(date:dd)$(rev:rrr)'
+
+jobs:
+  - template: ./templates/build-console-audit-job.yml
+    parameters:
+      platform: x64
+
+  - template: ./templates/build-console-int.yml
+    parameters:
+      platform: x64
+      additionalBuildArguments: /p:WindowsTerminalReleaseBuild=true
+
+  - template: ./templates/build-console-int.yml
+    parameters:
+      platform: x86
+      additionalBuildArguments: /p:WindowsTerminalReleaseBuild=true
+
+  - template: ./templates/build-console-int.yml
+    parameters:
+      platform: arm64
+      additionalBuildArguments: /p:WindowsTerminalReleaseBuild=true
+
+  - template: ./templates/check-formatting.yml
+
+  - template: ./templates/release-sign-and-bundle.yml

build/pipelines/templates/build-console-audit-job.yml

@@ -0,0 +1,43 @@
+parameters:
+  platform: ''
+  additionalBuildArguments: ''
+
+jobs:
+- job: Build${{ parameters.platform }}AuditMode
+  displayName: Static Analysis Build ${{ parameters.platform }}
+  variables:
+    BuildConfiguration: AuditMode
+    BuildPlatform: ${{ parameters.platform }}
+  pool: { vmImage: windows-2019 }
+
+  steps:
+  - checkout: self
+    submodules: true
+    clean: true
+
+  - task: NuGetToolInstaller@0
+    displayName: Ensure NuGet 4.8.1
+    inputs:
+      versionSpec: 4.8.1
+
+  # In the Microsoft Azure DevOps tenant, NuGetCommand is ambiguous.
+  # This should be `task: NuGetCommand@2`
+  - task: 333b11bd-d341-40d9-afcf-b32d5ce6f23b@2
+    displayName: Restore NuGet packages
+    inputs:
+      command: restore
+      feedsToUse: config
+      configPath: NuGet.config
+      restoreSolution: OpenConsole.sln
+      restoreDirectory: '$(Build.SourcesDirectory)\packages'
+
+  - task: VSBuild@1
+    displayName: 'Build solution **\OpenConsole.sln'
+    inputs:
+      solution: '**\OpenConsole.sln'
+      vsVersion: 16.0
+      platform: '$(BuildPlatform)'
+      configuration: '$(BuildConfiguration)'
+      msbuildArgs: ${{ parameters.additionalBuildArguments }}
+      clean: true
+      maximumCpuCount: true

build/pipelines/templates/build-console-ci.yml

@@ -0,0 +1,17 @@
+parameters:
+  configuration: 'Release'
+  platform: ''
+  additionalBuildArguments: ''
+
+jobs:
+- job: Build${{ parameters.platform }}${{ parameters.configuration }}
+  displayName: Build ${{ parameters.platform }} ${{ parameters.configuration }}
+  variables:
+    BuildConfiguration: ${{ parameters.configuration }}
+    BuildPlatform: ${{ parameters.platform }}
+  pool: { vmImage: windows-2019 }
+
+  steps:
+  - template: build-console-steps.yml
+    parameters:
+      additionalBuildArguments: ${{ parameters.additionalBuildArguments }}

build/pipelines/templates/build-console-int.yml

@@ -0,0 +1,30 @@
+parameters:
+  configuration: 'Release'
+  platform: ''
+  additionalBuildArguments: ''
+
+jobs:
+- job: Build${{ parameters.platform }}${{ parameters.configuration }}
+  displayName: Build ${{ parameters.platform }} ${{ parameters.configuration }}
+  variables:
+    BuildConfiguration: ${{ parameters.configuration }}
+    BuildPlatform: ${{ parameters.platform }}
+
+  pool:
+    name: Package ES Lab E
+    demands:
+    - msbuild
+    - visualstudio
+    - vstest
+
+  steps:
+  - task: PkgESSetupBuild@10
+    displayName: 'Package ES - Setup Build'
+    inputs:
+      useDfs: false
+      productName: WindowsTerminal
+      disableOutputRedirect: true
+
+  - template: build-console-steps.yml
+    parameters:
+      additionalBuildArguments: "/p:XesUseOneStoreVersioning=true;XesBaseYearForStoreVersion=$(baseYearForVersioning) ${{ parameters.additionalBuildArguments }}"

build/pipelines/templates/build-console-steps.yml

@@ -0,0 +1,98 @@
+parameters:
+  additionalBuildArguments: ''
+
+steps:
+- checkout: self
+  submodules: true
+  clean: true
+
+- task: NuGetToolInstaller@0
+  displayName: Ensure NuGet 4.8.1
+  inputs:
+    versionSpec: 4.8.1
+
+- task: VisualStudioTestPlatformInstaller@1
+  displayName: Ensure VSTest Platform
+
+# In the Microsoft Azure DevOps tenant, NuGetCommand is ambiguous.
+# This should be `task: NuGetCommand@2`
+- task: 333b11bd-d341-40d9-afcf-b32d5ce6f23b@2
+  displayName: Restore NuGet packages
+  inputs:
+    command: restore
+    feedsToUse: config
+    configPath: NuGet.config
+    restoreSolution: OpenConsole.sln
+    restoreDirectory: '$(Build.SourcesDirectory)\packages'
+
+- task: VSBuild@1
+  displayName: 'Build solution **\OpenConsole.sln'
+  inputs:
+    solution: '**\OpenConsole.sln'
+    vsVersion: 16.0
+    platform: '$(BuildPlatform)'
+    configuration: '$(BuildConfiguration)'
+    msbuildArgs: "${{ parameters.additionalBuildArguments }}"
+    clean: true
+    maximumCpuCount: true
+
+- task: PowerShell@2
+  displayName: 'Check MSIX for common regressions'
+  inputs:
+    targetType: inline
+    script: |
+      $Package = Get-ChildItem -Recurse -Filter "CascadiaPackage_*.msix"
+      .\build\scripts\Test-WindowsTerminalPackage.ps1 -Verbose -Path $Package.FullName
+
+- task: powershell@2
+  displayName: 'Source Index PDBs'
+  inputs:
+    targetType: filePath
+    filePath: build\scripts\Index-Pdbs.ps1
+    arguments: -SearchDir '$(Build.SourcesDirectory)' -SourceRoot '$(Build.SourcesDirectory)' -recursive -Verbose -CommitId $(Build.SourceVersion)
+    errorActionPreference: silentlyContinue
+
+- task: PowerShell@2
+  displayName: 'Rationalize build platform'
+  inputs:
+    targetType: inline
+    script: |
+      $Arch = "$(BuildPlatform)"
+      If ($Arch -Eq "x86") { $Arch = "Win32" }
+      Write-Host "##vso[task.setvariable variable=RationalizedBuildPlatform]${Arch}"
+
+- task: PowerShell@2
+  displayName: 'Run Unit Tests'
+  inputs:
+    targetType: filePath
+    filePath: build\scripts\Run-Tests.ps1
+    arguments: -MatchPattern '*unit.test*.dll' -Platform '$(RationalizedBuildPlatform)' -Configuration '$(BuildConfiguration)'
+  condition: and(succeeded(), or(eq(variables['BuildPlatform'], 'x64'), eq(variables['BuildPlatform'], 'x86')))
+
+- task: PowerShell@2
+  displayName: 'Run Feature Tests (x64 only)'
+  inputs:
+    targetType: filePath
+    filePath: build\scripts\Run-Tests.ps1
+    arguments: -MatchPattern '*feature.test*.dll' -Platform '$(RationalizedBuildPlatform)' -Configuration '$(BuildConfiguration)'
+  condition: and(succeeded(), eq(variables['BuildPlatform'], 'x64'))
+
+- task: CopyFiles@2
+  displayName: 'Copy *.appx/*.msix to Artifacts (Non-PR builds only)'
+  inputs:
+    Contents: |
+     **/*.appx
+     **/*.msix
+     **/*.appxsym
+     !**/Microsoft.VCLibs*.appx
+    TargetFolder: '$(Build.ArtifactStagingDirectory)/appx'
+    OverWrite: true
+    flattenFolders: true
+  condition: and(succeeded(), ne(variables['Build.Reason'], 'PullRequest'))
+
+- task: PublishBuildArtifacts@1
+  displayName: 'Publish Artifact (appx) (Non-PR builds only)'
+  inputs:
+    PathtoPublish: '$(Build.ArtifactStagingDirectory)/appx'
+    ArtifactName: 'appx-$(BuildConfiguration)'
+  condition: and(succeeded(), ne(variables['Build.Reason'], 'PullRequest'))

build/pipelines/templates/check-formatting.yml

@@ -0,0 +1,16 @@
+
+jobs:
+- job: CodeFormatCheck
+  displayName: Proper Code Formatting Check
+  pool: { vmImage: windows-2019 }
+
+  steps:
+  - checkout: self
+    submodules: false
+    clean: true
+
+  - task: PowerShell@2
+    displayName: 'Code Formattting Check'
+    inputs:
+      targetType: filePath
+      filePath: '.\build\scripts\Invoke-FormattingCheck.ps1'

build/pipelines/templates/release-sign-and-bundle.yml

@@ -0,0 +1,74 @@
+parameters:
+  configuration: 'Release'
+
+jobs:
+- job: SignDeploy${{ parameters.configuration }}
+  displayName: Sign and Deploy for ${{ parameters.configuration }}
+
+  dependsOn:
+    - Buildx64AuditMode
+    - Buildx64Release
+    - Buildx86Release
+    - Buildarm64Release
+    - CodeFormatCheck
+  condition: |
+    and
+    (
+      in(dependencies.Buildx64AuditMode.result, 'Succeeded', 'SucceededWithIssues', 'Skipped'),
+      in(dependencies.Buildx64Release.result,   'Succeeded', 'SucceededWithIssues', 'Skipped'),
+      in(dependencies.Buildx86Release.result,   'Succeeded', 'SucceededWithIssues', 'Skipped'),
+      in(dependencies.Buildarm64Release.result, 'Succeeded', 'SucceededWithIssues', 'Skipped'),
+      in(dependencies.CodeFormatCheck.result,   'Succeeded', 'SucceededWithIssues', 'Skipped')
+    )
+
+  variables:
+    BuildConfiguration: ${{ parameters.configuration }}
+    AppxProjectName: CascadiaPackage
+    AppxBundleName: Microsoft.WindowsTerminal_8wekyb3d8bbwe.msixbundle
+
+  pool:
+    name: Package ES Lab E
+
+  steps:
+  - checkout: self
+    clean: true
+
+  - task: PkgESSetupBuild@10
+    displayName: 'Package ES - Setup Build'
+    inputs:
+      useDfs: false
+      productName: WindowsTerminal
+      disableOutputRedirect: true
+
+  - task: ms.vss-governance-buildtask.governance-build-task-component-detection.ComponentGovernanceComponentDetection@0
+    displayName: 'Component Detection'
+
+  - task: DownloadBuildArtifacts@0
+    displayName: Download AppX artifacts
+    inputs:
+      artifactName: 'appx-$(BuildConfiguration)'
+      itemPattern: |
+        **/*.appx
+        **/*.msix
+      downloadPath: '$(Build.ArtifactStagingDirectory)\appx'
+
+  - task: PowerShell@2
+    displayName: 'Create $(AppxBundleName)'
+    inputs:
+      targetType: filePath
+      filePath: '.\build\scripts\Create-AppxBundle.ps1'
+      arguments: |
+        -InputPath "$(Build.ArtifactStagingDirectory)\appx" -ProjectName $(AppxProjectName) -BundleVersion 0.0.0.0 -OutputPath "$(Build.ArtifactStagingDirectory)\$(AppxBundleName)"
+
+  - task: PkgESCodeSign@10
+    displayName: 'Package ES - SignConfig.WindowsTerminal.xml'
+    inputs:
+      signConfigXml: 'build\config\SignConfig.WindowsTerminal.xml'
+      inPathRoot: '$(Build.ArtifactStagingDirectory)'
+      outPathRoot: '$(Build.ArtifactStagingDirectory)\signed'
+
+  - task: PublishBuildArtifacts@1
+    displayName: 'Publish Signed AppX'
+    inputs:
+      PathtoPublish: '$(Build.ArtifactStagingDirectory)\signed'
+      ArtifactName: 'appxbundle-signed-$(BuildConfiguration)'

build/rules/GenerateSxsManifestsFromWinmds.targets

@@ -0,0 +1,41 @@
+<?xml version="1.0" encoding="utf-8"?>
+<Project DefaultTargets="Build" ToolsVersion="16.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003">
+  <PropertyGroup>
+    <BeforeLinkTargets Condition="'$(WindowsTargetPlatformVersion)' &gt;= '10.0.18362.0'">
+      $(BeforeLinkTargets);
+      _ConsoleGenerateAdditionalWinmdManifests;
+    </BeforeLinkTargets>
+  </PropertyGroup>
+
+  <Target Name="_ConsoleMapWinmdsToManifestFiles" DependsOnTargets="ResolveAssemblyReferences">
+    <ItemGroup>
+      <!-- For each non-system .winmd file in References, generate a .manifest in IntDir for it. -->
+      <_ConsoleWinmdManifest Include="@(ReferencePath->'$(IntDir)\%(FileName).manifest')" Condition="'%(ReferencePath.IsSystemReference)' != 'true' and '%(ReferencePath.WinMDFile)' == 'true' and '%(ReferencePath.ReferenceSourceTarget)' == 'ResolveAssemblyReference' and '%(ReferencePath.Implementation)' != ''">
+        <WinMDPath>%(ReferencePath.FullPath)</WinMDPath>
+        <Implementation>%(ReferencePath.Implementation)</Implementation>
+      </_ConsoleWinmdManifest>
+      <!-- For each referenced project that _produces_ a winmd, generate a temporary item that maps to
+           the winmd, and use that temporary item to generate a .manifest in IntDir for it.
+           We don't set Implementation here because it's inherited from the _ResolvedNativeProjectReferencePaths. -->
+      <_ConsoleWinmdProjectReference Condition="'%(_ResolvedNativeProjectReferencePaths.ProjectType)' != 'StaticLibrary'" Include="@(_ResolvedNativeProjectReferencePaths-&gt;WithMetadataValue('FileType','winmd')-&gt;'%(RootDir)%(Directory)%(TargetPath)')" />
+      <_ConsoleWinmdManifest Include="@(_ConsoleWinmdProjectReference->'$(IntDir)\%(FileName).manifest')">
+        <WinMDPath>%(Identity)</WinMDPath>
+      </_ConsoleWinmdManifest>
+    </ItemGroup>
+  </Target>
+
+  <Target Name="_ConsoleGenerateAdditionalWinmdManifests"
+          Inputs="@(_ConsoleWinmdManifest.WinMDPath)"
+          Outputs="@(_ConsoleWinmdManifest)"
+          DependsOnTargets="_ConsoleMapWinmdsToManifestFiles">
+
+    <!-- This target is batched and a new Exec is spawned for each entry in _ConsoleWinmdManifest. -->
+    <Exec Command="mt.exe -winmd:%(_ConsoleWinmdManifest.WinMDPath) -dll:%(_ConsoleWinmdManifest.Implementation) -out:%(_ConsoleWinmdManifest.Identity)" />
+
+    <ItemGroup>
+      <!-- Emit the generated manifest into the Link inputs. -->
+      <Manifest Include="@(_ConsoleWinmdManifest)" />
+    </ItemGroup>
+
+  </Target>
+</Project>

build/scripts/Index-Pdbs.ps1

@@ -0,0 +1,85 @@
+[CmdLetBinding()]
+Param(
+    [Parameter(Mandatory=$true, Position=0)][string]$SearchDir,
+    [Parameter(Mandatory=$true, Position=1)][string]$SourceRoot,
+    [Parameter(Mandatory=$true, Position=2)][string]$CommitId,
+    [string]$Organization = "microsoft",
+    [string]$Repo = "terminal",
+    [switch]$recursive
+)
+
+$debuggerPath = (Get-ItemProperty -path "HKLM:\SOFTWARE\WOW6432Node\Microsoft\Windows Kits\Installed Roots" -name WindowsDebuggersRoot10).WindowsDebuggersRoot10
+$srcsrvPath = Join-Path $debuggerPath "x64\srcsrv"
+$srctoolExe = Join-Path $srcsrvPath "srctool.exe"
+$pdbstrExe = Join-Path $srcsrvPath "pdbstr.exe"
+
+$fileTable = @{}
+foreach ($gitFile in & git ls-files)
+{
+    $fileTable[$gitFile] = $gitFile
+}
+
+$mappedFiles = New-Object System.Collections.ArrayList
+
+foreach ($file in (Get-ChildItem -r:$recursive "$SearchDir\*.pdb"))
+{
+    Write-Verbose "Found $file"
+
+    $ErrorActionPreference = "Continue" # Azure Pipelines defaults to "Stop", continue past errors in this script.
+    
+    $allFiles = & $srctoolExe -r "$file"
+
+    # If the pdb didn't have enough files then skip it (the srctool output has a blank line even when there's no info
+    # so check for less than 2 lines)
+    if ($allFiles.Length -lt 2)
+    {
+        continue
+    }
+
+    for ($i = 0; $i -lt $allFiles.Length; $i++)
+    {
+        if ($allFiles[$i].StartsWith($SourceRoot, [StringComparison]::OrdinalIgnoreCase))
+        {
+            $relative = $allFiles[$i].Substring($SourceRoot.Length).TrimStart("\")
+            $relative = $relative.Replace("\", "/")
+
+            # Git urls are case-sensitive but the PDB might contain a lowercased version of the file path.
+            # Look up the relative url in the output of "ls-files". If it's not there then it's not something
+            # in git, so don't index it.
+            $relative = $fileTable[$relative]
+            if ($relative)
+            {
+                $mapping = $allFiles[$i] + "*$relative"
+                $mappedFiles.Add($mapping)
+
+                Write-Verbose "Mapped path $($i): $mapping"
+            }
+        }
+    }
+
+    $pdbstrFile = Join-Path "$env:TEMP" "pdbstr.txt"
+
+    Write-Verbose "pdbstr.txt = $pdbstrFile"
+
+    @"
+SRCSRV: ini ------------------------------------------------
+VERSION=2
+VERCTRL=http
+SRCSRV: variables ------------------------------------------
+ORGANIZATION=$Organization
+REPO=$Repo
+COMMITID=$CommitId
+HTTP_ALIAS=https://raw.githubusercontent.com/%ORGANIZATION%/%REPO%/%COMMITID%/
+HTTP_EXTRACT_TARGET=%HTTP_ALIAS%%var2%
+SRCSRVTRG=%HTTP_EXTRACT_TARGET%
+SRC_INDEX=public
+SRCSRV: source files ---------------------------------------
+$($mappedFiles -join "`r`n")
+SRCSRV: end ------------------------------------------------
+"@ | Set-Content $pdbstrFile
+
+    & $pdbstrExe -p:"$file" -w -s:srcsrv -i:$pdbstrFile
+}
+
+# Return with exit 0 to override any weird error code from other tools
+Exit 0

build/scripts/Invoke-FormattingCheck.ps1

@@ -0,0 +1,14 @@
+
+#.SYNOPSIS
+# Checks for code formatting errors. Will throw exception if any are found.
+function Invoke-CheckBadCodeFormatting() {
+    Import-Module ./tools/OpenConsole.psm1
+    Invoke-CodeFormat
+    # returns a non-zero exit code if there are any diffs in the tracked files in the repo
+    git diff-index --quiet HEAD --
+    if ($lastExitCode -eq 1) {
+        throw "code formatting bad, run Invoke-CodeFormat on branch"
+    }
+}
+
+Invoke-CheckBadCodeFormatting

build/scripts/Run-Tests.ps1

@@ -0,0 +1,14 @@
+[CmdLetBinding()]
+Param(
+    [Parameter(Mandatory=$true, Position=0)][string]$MatchPattern,
+    [Parameter(Mandatory=$true, Position=1)][string]$Platform,
+    [Parameter(Mandatory=$true, Position=2)][string]$Configuration
+)
+
+$testdlls = Get-ChildItem -Path ".\bin\$Platform\$Configuration" -Recurse -Filter $MatchPattern
+
+&".\bin\$Platform\$Configuration\te.exe" $testdlls.FullName
+
+if ($lastexitcode -Ne 0) { Exit $lastexitcode }
+
+Exit 0

build/scripts/Test-WindowsTerminalPackage.ps1

@@ -0,0 +1,85 @@
+[CmdletBinding()]
+Param(
+    [Parameter(Mandatory=$true, ValueFromPipeline=$true,
+      HelpMessage="Path to the .appx/.msix to validate")]
+    [string]
+    $Path,
+
+    [Parameter(HelpMessage="Path to Windows Kit")]
+    [ValidateScript({Test-Path $_ -Type Leaf})]
+    [string]
+    $WindowsKitPath = "C:\Program Files (x86)\Windows Kits\10\bin\10.0.18362.0"
+)
+
+$ErrorActionPreference = "Stop"
+
+If ($null -Eq (Get-Item $WindowsKitPath -EA:SilentlyContinue)) {
+    Write-Error "Could not find a windows SDK at at `"$WindowsKitPath`".`nMake sure that WindowsKitPath points to a valid SDK."
+    Exit 1
+}
+
+$makeAppx = "$WindowsKitPath\x86\MakeAppx.exe"
+$makePri = "$WindowsKitPath\x86\MakePri.exe"
+
+Function Expand-ApplicationPackage {
+    Param(
+        [Parameter(Mandatory, ValueFromPipeline)]
+        [string]
+        $Path
+    )
+
+    $sentinelFile = New-TemporaryFile
+    $directory = New-Item -Type Directory "$($sentinelFile.FullName)_Package"
+    Remove-Item $sentinelFile -Force -EA:Ignore
+
+    & $makeAppx unpack /p $Path /d $directory /nv /o
+
+    If ($LastExitCode -Ne 0) {
+        Throw "Failed to expand AppX"
+    }
+
+    $directory
+}
+
+Write-Verbose "Expanding $Path"
+$AppxPackageRoot = Expand-ApplicationPackage $Path
+$AppxPackageRootPath = $AppxPackageRoot.FullName
+
+Write-Verbose "Expanded to $AppxPackageRootPath"
+
+Try {
+    & $makePri dump /if "$AppxPackageRootPath\resources.pri" /of "$AppxPackageRootPath\resources.pri.xml" /o
+    If ($LastExitCode -Ne 0) {
+        Throw "Failed to dump PRI"
+    }
+
+    $Manifest = [xml](Get-Content "$AppxPackageRootPath\AppxManifest.xml")
+    $PRIFile = [xml](Get-Content "$AppxPackageRootPath\resources.pri.xml")
+
+    ### Check the activatable class entries for a few DLLs we need.
+    $inProcServers = $Manifest.Package.Extensions.Extension.InProcessServer.Path
+    $RequiredInProcServers = ("TerminalApp.dll", "TerminalControl.dll", "TerminalConnection.dll")
+
+    Write-Verbose "InProc Servers: $inProcServers"
+
+    ForEach ($req in $RequiredInProcServers) {
+        If ($req -NotIn $inProcServers) {
+            Throw "Failed to find $req in InProcServer list $inProcServers"
+        }
+    }
+
+    ### Check that we have an App.xbf (which is a proxy for our resources having been merged)
+    $resourceXpath = '/PriInfo/ResourceMap/ResourceMapSubtree[@name="Files"]/NamedResource[@name="App.xbf"]'
+    $AppXbf = $PRIFile.SelectSingleNode($resourceXpath)
+    If ($null -eq $AppXbf) {
+        Throw "Failed to find App.xbf (TerminalApp project) in resources.pri"
+    }
+
+    If (($null -eq (Get-Item "$AppxPackageRootPath\cpprest142_2_10.dll" -EA:Ignore)) -And
+        ($null -eq (Get-Item "$AppxPackageRootPath\cpprest142_2_10d.dll" -EA:Ignore))) {
+        Throw "Failed to find cpprest142_2_10.dll -- check the WAP packaging project"
+    }
+
+} Finally {
+    Remove-Item -Recurse -Force $AppxPackageRootPath
+}

consolegit2gitfilters.json

@@ -11,7 +11,13 @@
     "/packages/",
     "/ipch/",
     "/dep/",
-    "/.vs/"
+    "/.vs/",
+    "/build/",
+    "/src/cascadia/",
+    "/src/winconpty/",
+    "/.nuget/",
+    "/.github/",
+    "/samples/"
   ],
   "SuffixFilters": [
     ".dbb",

custom.props

@@ -0,0 +1,11 @@
+<?xml version="1.0" encoding="utf-8"?>
+<Project ToolsVersion="4.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003">
+  <!-- This file is read by XES, which we use in our Release builds. -->
+  <PropertyGroup Label="Version">
+    <XesUseOneStoreVersioning>true</XesUseOneStoreVersioning>
+    <XesBaseYearForStoreVersion>2020</XesBaseYearForStoreVersion>
+    <VersionMajor>0</VersionMajor>
+    <VersionMinor>9</VersionMinor>
+    <VersionInfoProductName>Windows Terminal</VersionInfoProductName>
+  </PropertyGroup>
+</Project>

dep/CLI11/README.md

@@ -0,0 +1,5 @@
+# CLI11
+
+Taken from [release v1.8.0](https://github.com/CLIUtils/CLI11/releases/tag/v1.8.0), source commit
+[13becad](https://github.com/CLIUtils/CLI11/commit/13becaddb657eacd090537719a669d66d393b8b2)
+

dep/chromium/LICENSE

@@ -0,0 +1,27 @@
+// Copyright 2015 The Chromium Authors. All rights reserved.
+//
+// Redistribution and use in source and binary forms, with or without
+// modification, are permitted provided that the following conditions are
+// met:
+//
+//    * Redistributions of source code must retain the above copyright
+// notice, this list of conditions and the following disclaimer.
+//    * Redistributions in binary form must reproduce the above
+// copyright notice, this list of conditions and the following disclaimer
+// in the documentation and/or other materials provided with the
+// distribution.
+//    * Neither the name of Google Inc. nor the names of its
+// contributors may be used to endorse or promote products derived from
+// this software without specific prior written permission.
+//
+// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

dep/chromium/MAINTAINER_README.md

@@ -0,0 +1,17 @@
+### Notes for Future Maintainers
+
+This was originally imported by @miniksa in January 2020.
+
+The provenance information (where it came from and which commit) is stored in the file `cgmanifest.json` in the same directory as this readme.
+Please update the provenance information in that file when ingesting an updated version of the dependent library.
+That provenance file is automatically read and inventoried by Microsoft systems to ensure compliance with appropiate governance standards.
+
+## What should be done to update this in the future?
+
+1. Go to chromium/chromium repository on GitHub.
+2. Take the entire contents of the base/numerics directory wholesale and drop it in the base/numerics directory here.
+3. Don't change anything about it.
+4. Validate that the license in the root of the repository didn't change and update it if so. It is sitting in the same directory as this readme. 
+   If it changed dramatically, ensure that it is still compatible with our license scheme. Also update the NOTICE file in the root of our repository to declare the third-party usage.
+5. Submit the pull.
+

dep/chromium/base/numerics/BUILD.gn

@@ -0,0 +1,28 @@
+# Copyright (c) 2017 The Chromium Authors. All rights reserved.
+# Use of this source code is governed by a BSD-style license that can be
+# found in the LICENSE file.
+
+# This is a dependency-free, header-only, library, and it needs to stay that
+# way to facilitate pulling it into various third-party projects. So, this
+# file is here to protect against accidentally introducing external
+# dependencies or depending on internal implementation details.
+source_set("base_numerics") {
+  visibility = [ "//base/*" ]
+  sources = [
+    "checked_math_impl.h",
+    "clamped_math_impl.h",
+    "safe_conversions_arm_impl.h",
+    "safe_conversions_impl.h",
+    "safe_math_arm_impl.h",
+    "safe_math_clang_gcc_impl.h",
+    "safe_math_shared_impl.h",
+  ]
+  public = [
+    "checked_math.h",
+    "clamped_math.h",
+    "math_constants.h",
+    "ranges.h",
+    "safe_conversions.h",
+    "safe_math.h",
+  ]
+}

dep/chromium/base/numerics/DEPS

@@ -0,0 +1,7 @@
+# This is a dependency-free, header-only, library, and it needs to stay that
+# way to facilitate pulling it into various third-party projects. So, this
+# file is here to protect against accidentally introducing dependencies.
+include_rules = [
+  "-base",
+  "+base/numerics",
+]

dep/chromium/base/numerics/OWNERS

@@ -0,0 +1,5 @@
+jschuh@chromium.org
+tsepez@chromium.org
+
+
+# COMPONENT: Internals

dep/chromium/base/numerics/README.md

@@ -0,0 +1,409 @@
+# `base/numerics`
+
+This directory contains a dependency-free, header-only library of templates
+providing well-defined semantics for safely and performantly handling a variety
+of numeric operations, including most common arithmetic operations and
+conversions.
+
+The public API is broken out into the following header files:
+
+*   `checked_math.h` contains the `CheckedNumeric` template class and helper
+    functions for performing arithmetic and conversion operations that detect
+    errors and boundary conditions (e.g. overflow, truncation, etc.).
+*   `clamped_math.h` contains the `ClampedNumeric` template class and
+    helper functions for performing fast, clamped (i.e. [non-sticky](#notsticky)
+    saturating) arithmetic operations and conversions.
+*   `safe_conversions.h` contains the `StrictNumeric` template class and
+    a collection of custom casting templates and helper functions for safely
+    converting between a range of numeric types.
+*   `safe_math.h` includes all of the previously mentioned headers.
+
+*** aside
+**Note:** The `Numeric` template types implicitly convert from C numeric types
+and `Numeric` templates that are convertable to an underlying C numeric type.
+The conversion priority for `Numeric` type coercions is:
+
+*   `StrictNumeric` coerces to `ClampedNumeric` and `CheckedNumeric`
+*   `ClampedNumeric` coerces to `CheckedNumeric`
+***
+
+[TOC]
+
+## Common patterns and use-cases
+
+The following covers the preferred style for the most common uses of this
+library. Please don't cargo-cult from anywhere else. 😉
+
+### Performing checked arithmetic type conversions
+
+The `checked_cast` template converts between arbitrary arithmetic types, and is
+used for cases where a conversion failure should result in program termination:
+
+```cpp
+// Crash if signed_value is out of range for buff_size.
+size_t buff_size = checked_cast<size_t>(signed_value);
+```
+
+### Performing saturated (clamped) arithmetic type conversions
+
+The `saturated_cast` template converts between arbitrary arithmetic types, and
+is used in cases where an out-of-bounds source value should be saturated to the
+corresponding maximum or minimum of the destination type:
+
+```cpp
+// Convert from float with saturation to INT_MAX, INT_MIN, or 0 for NaN.
+int int_value = saturated_cast<int>(floating_point_value);
+```
+
+### Enforcing arithmetic type conversions at compile-time
+
+The `strict_cast` emits code that is identical to `static_cast`. However,
+provides static checks that will cause a compilation failure if the
+destination type cannot represent the full range of the source type:
+
+```cpp
+// Throw a compiler error if byte_value is changed to an out-of-range-type.
+int int_value = strict_cast<int>(byte_value);
+```
+
+You can also enforce these compile-time restrictions on function parameters by
+using the `StrictNumeric` template:
+
+```cpp
+// Throw a compiler error if the size argument cannot be represented by a
+// size_t (e.g. passing an int will fail to compile).
+bool AllocateBuffer(void** buffer, StrictCast<size_t> size);
+```
+
+### Comparing values between arbitrary arithmetic types
+
+Both the `StrictNumeric` and `ClampedNumeric` types provide well defined
+comparisons between arbitrary arithmetic types. This allows you to perform
+comparisons that are not legal or would trigger compiler warnings or errors
+under the normal arithmetic promotion rules:
+
+```cpp
+bool foo(unsigned value, int upper_bound) {
+  // Converting to StrictNumeric allows this comparison to work correctly.
+  if (MakeStrictNum(value) >= upper_bound)
+    return false;
+```
+
+*** note
+**Warning:** Do not perform manual conversions using the comparison operators.
+Instead, use the cast templates described in the previous sections, or the
+constexpr template functions `IsValueInRangeForNumericType` and
+`IsTypeInRangeForNumericType`, as these templates properly handle the full range
+of corner cases and employ various optimizations.
+***
+
+### Calculating a buffer size (checked arithmetic)
+
+When making exact calculations—such as for buffer lengths—it's often necessary
+to know when those calculations trigger an overflow, undefined behavior, or
+other boundary conditions. The `CheckedNumeric` template does this by storing
+a bit determining whether or not some arithmetic operation has occured that
+would put the variable in an "invalid" state. Attempting to extract the value
+from a variable in an invalid state will trigger a check/trap condition, that
+by default will result in process termination.
+
+Here's an example of a buffer calculation using a `CheckedNumeric` type (note:
+the AssignIfValid method will trigger a compile error if the result is ignored).
+
+```cpp
+// Calculate the buffer size and detect if an overflow occurs.
+size_t size;
+if (!CheckAdd(kHeaderSize, CheckMul(count, kItemSize)).AssignIfValid(&size)) {
+  // Handle an overflow error...
+}
+```
+
+### Calculating clamped coordinates (non-sticky saturating arithmetic)
+
+Certain classes of calculations—such as coordinate calculations—require
+well-defined semantics that always produce a valid result on boundary
+conditions. The `ClampedNumeric` template addresses this by providing
+performant, non-sticky saturating arithmetic operations.
+
+Here's an example of using a `ClampedNumeric` to calculate an operation
+insetting a rectangle.
+
+```cpp
+// Use clamped arithmetic since inset calculations might overflow.
+void Rect::Inset(int left, int top, int right, int bottom) {
+  origin_ += Vector2d(left, top);
+  set_width(ClampSub(width(), ClampAdd(left, right)));
+  set_height(ClampSub(height(), ClampAdd(top, bottom)));
+}
+```
+
+*** note
+<a name="notsticky"></a>
+The `ClampedNumeric` type is not "sticky", which means the saturation is not
+retained across individual operations. As such, one arithmetic operation may
+result in a saturated value, while the next operation may then "desaturate"
+the value. Here's an example:
+
+```cpp
+ClampedNumeric<int> value = INT_MAX;
+++value;  // value is still INT_MAX, due to saturation.
+--value;  // value is now (INT_MAX - 1), because saturation is not sticky.
+```
+
+***
+
+## Conversion functions and StrictNumeric<> in safe_conversions.h
+
+This header includes a collection of helper `constexpr` templates for safely
+performing a range of conversions, assignments, and tests.
+
+### Safe casting templates
+
+*   `as_signed()` - Returns the supplied integral value as a signed type of
+    the same width.
+*   `as_unsigned()` - Returns the supplied integral value as an unsigned type
+    of the same width.
+*   `checked_cast<>()` - Analogous to `static_cast<>` for numeric types, except
+    that by default it will trigger a crash on an out-of-bounds conversion (e.g.
+    overflow, underflow, NaN to integral) or a compile error if the conversion
+    error can be detected at compile time. The crash handler can be overridden
+    to perform a behavior other than crashing.
+*   `saturated_cast<>()` - Analogous to `static_cast` for numeric types, except
+    that it returns a saturated result when the specified numeric conversion
+    would otherwise overflow or underflow. An NaN source returns 0 by
+    default, but can be overridden to return a different result.
+*   `strict_cast<>()` - Analogous to `static_cast` for numeric types, except
+    this causes a compile failure if the destination type is not large
+    enough to contain any value in the source type. It performs no runtime
+    checking and thus introduces no runtime overhead.
+
+### Other helper and conversion functions
+
+*   `IsValueInRangeForNumericType<>()` - A convenience function that returns
+    true if the type supplied as the template parameter can represent the value
+    passed as an argument to the function.
+*   `IsTypeInRangeForNumericType<>()` - A convenience function that evaluates
+    entirely at compile-time and returns true if the destination type (first
+    template parameter) can represent the full range of the source type
+    (second template parameter).
+*   `IsValueNegative()` - A convenience function that will accept any
+    arithmetic type as an argument and will return whether the value is less
+    than zero. Unsigned types always return false.
+*   `SafeUnsignedAbs()` - Returns the absolute value of the supplied integer
+    parameter as an unsigned result (thus avoiding an overflow if the value
+    is the signed, two's complement minimum).
+
+### StrictNumeric<>
+
+`StrictNumeric<>` is a wrapper type that performs assignments and copies via
+the `strict_cast` template, and can perform valid arithmetic comparisons
+across any range of arithmetic types. `StrictNumeric` is the return type for
+values extracted from a `CheckedNumeric` class instance. The raw numeric value
+is extracted via `static_cast` to the underlying type or any type with
+sufficient range to represent the underlying type.
+
+*   `MakeStrictNum()` - Creates a new `StrictNumeric` from the underlying type
+    of the supplied arithmetic or StrictNumeric type.
+*   `SizeT` - Alias for `StrictNumeric<size_t>`.
+
+## CheckedNumeric<> in checked_math.h
+
+`CheckedNumeric<>` implements all the logic and operators for detecting integer
+boundary conditions such as overflow, underflow, and invalid conversions.
+The `CheckedNumeric` type implicitly converts from floating point and integer
+data types, and contains overloads for basic arithmetic operations (i.e.: `+`,
+`-`, `*`, `/` for all types and `%`, `<<`, `>>`, `&`, `|`, `^` for integers).
+However, *the [variadic template functions
+](#CheckedNumeric_in-checked_math_h-Non_member-helper-functions)
+are the prefered API,* as they remove type ambiguities and help prevent a number
+of common errors. The variadic functions can also be more performant, as they
+eliminate redundant expressions that are unavoidable with the with the operator
+overloads. (Ideally the compiler should optimize those away, but better to avoid
+them in the first place.)
+
+Type promotions are a slightly modified version of the [standard C/C++ numeric
+promotions
+](http://en.cppreference.com/w/cpp/language/implicit_conversion#Numeric_promotions)
+with the two differences being that *there is no default promotion to int*
+and *bitwise logical operations always return an unsigned of the wider type.*
+
+### Members
+
+The unary negation, increment, and decrement operators are supported, along
+with the following unary arithmetic methods, which return a new
+`CheckedNumeric` as a result of the operation:
+
+*   `Abs()` - Absolute value.
+*   `UnsignedAbs()` - Absolute value as an equal-width unsigned underlying type
+    (valid for only integral types).
+*   `Max()` - Returns whichever is greater of the current instance or argument.
+    The underlying return type is whichever has the greatest magnitude.
+*   `Min()` - Returns whichever is lowest of the current instance or argument.
+    The underlying return type is whichever has can represent the lowest
+    number in the smallest width (e.g. int8_t over unsigned, int over
+    int8_t, and float over int).
+
+The following are for converting `CheckedNumeric` instances:
+
+*   `type` - The underlying numeric type.
+*   `AssignIfValid()` - Assigns the underlying value to the supplied
+    destination pointer if the value is currently valid and within the
+    range supported by the destination type. Returns true on success.
+*   `Cast<>()` - Instance method returning a `CheckedNumeric` derived from
+    casting the current instance to a `CheckedNumeric` of the supplied
+    destination type.
+
+*** aside
+The following member functions return a `StrictNumeric`, which is valid for
+comparison and assignment operations, but will trigger a compile failure on
+attempts to assign to a type of insufficient range. The underlying value can
+be extracted by an explicit `static_cast` to the underlying type or any type
+with sufficient range to represent the underlying type.
+***
+
+*   `IsValid()` - Returns true if the underlying numeric value is valid (i.e.
+    has not wrapped or saturated and is not the result of an invalid
+    conversion).
+*   `ValueOrDie()` - Returns the underlying value. If the state is not valid
+    this call will trigger a crash by default (but may be overridden by
+    supplying an alternate handler to the template).
+*   `ValueOrDefault()` - Returns the current value, or the supplied default if
+    the state is not valid (but will not crash).
+
+**Comparison operators are explicitly not provided** for `CheckedNumeric`
+types because they could result in a crash if the type is not in a valid state.
+Patterns like the following should be used instead:
+
+```cpp
+// Either input or padding (or both) may be arbitrary sizes.
+size_t buff_size;
+if (!CheckAdd(input, padding, kHeaderLength).AssignIfValid(&buff_size) ||
+     buff_size >= kMaxBuffer) {
+  // Handle an error...
+} else {
+  // Do stuff on success...
+}
+```
+
+### Non-member helper functions
+
+The following variadic convenience functions, which accept standard arithmetic
+or `CheckedNumeric` types, perform arithmetic operations, and return a
+`CheckedNumeric` result. The supported functions are:
+
+*   `CheckAdd()` - Addition.
+*   `CheckSub()` - Subtraction.
+*   `CheckMul()` - Multiplication.
+*   `CheckDiv()` - Division.
+*   `CheckMod()` - Modulus (integer only).
+*   `CheckLsh()` - Left integer shift (integer only).
+*   `CheckRsh()` - Right integer shift (integer only).
+*   `CheckAnd()` - Bitwise AND (integer only with unsigned result).
+*   `CheckOr()`  - Bitwise OR (integer only with unsigned result).
+*   `CheckXor()` - Bitwise XOR (integer only with unsigned result).
+*   `CheckMax()` - Maximum of supplied arguments.
+*   `CheckMin()` - Minimum of supplied arguments.
+
+The following wrapper functions can be used to avoid the template
+disambiguator syntax when converting a destination type.
+
+*   `IsValidForType<>()` in place of: `a.template IsValid<>()`
+*   `ValueOrDieForType<>()` in place of: `a.template ValueOrDie<>()`
+*   `ValueOrDefaultForType<>()` in place of: `a.template ValueOrDefault<>()`
+
+The following general utility methods is are useful for converting from
+arithmetic types to `CheckedNumeric` types:
+
+*   `MakeCheckedNum()` - Creates a new `CheckedNumeric` from the underlying type
+    of the supplied arithmetic or directly convertible type.
+
+## ClampedNumeric<> in clamped_math.h
+
+`ClampedNumeric<>` implements all the logic and operators for clamped
+(non-sticky saturating) arithmetic operations and conversions. The
+`ClampedNumeric` type implicitly converts back and forth between floating point
+and integer data types, saturating on assignment as appropriate. It contains
+overloads for basic arithmetic operations (i.e.: `+`, `-`, `*`, `/` for
+all types and `%`, `<<`, `>>`, `&`, `|`, `^` for integers) along with comparison
+operators for arithmetic types of any size. However, *the [variadic template
+functions
+](#ClampedNumeric_in-clamped_math_h-Non_member-helper-functions)
+are the prefered API,* as they remove type ambiguities and help prevent
+a number of common errors. The variadic functions can also be more performant,
+as they eliminate redundant expressions that are unavoidable with the operator
+overloads. (Ideally the compiler should optimize those away, but better to avoid
+them in the first place.)
+
+Type promotions are a slightly modified version of the [standard C/C++ numeric
+promotions
+](http://en.cppreference.com/w/cpp/language/implicit_conversion#Numeric_promotions)
+with the two differences being that *there is no default promotion to int*
+and *bitwise logical operations always return an unsigned of the wider type.*
+
+*** aside
+Most arithmetic operations saturate normally, to the numeric limit in the
+direction of the sign. The potentially unusual cases are:
+
+*   **Division:** Division by zero returns the saturated limit in the direction
+    of sign of the dividend (first argument). The one exception is 0/0, which
+	returns zero (although logically is NaN).
+*   **Modulus:** Division by zero returns the dividend (first argument).
+*   **Left shift:** Non-zero values saturate in the direction of the signed
+    limit (max/min), even for shifts larger than the bit width. 0 shifted any
+    amount results in 0.
+*   **Right shift:** Negative values saturate to -1. Positive or 0 saturates
+    to 0. (Effectively just an unbounded arithmetic-right-shift.)
+*   **Bitwise operations:** No saturation; bit pattern is identical to
+    non-saturated bitwise operations.
+***
+
+### Members
+
+The unary negation, increment, and decrement operators are supported, along
+with the following unary arithmetic methods, which return a new
+`ClampedNumeric` as a result of the operation:
+
+*   `Abs()` - Absolute value.
+*   `UnsignedAbs()` - Absolute value as an equal-width unsigned underlying type
+    (valid for only integral types).
+*   `Max()` - Returns whichever is greater of the current instance or argument.
+    The underlying return type is whichever has the greatest magnitude.
+*   `Min()` - Returns whichever is lowest of the current instance or argument.
+    The underlying return type is whichever has can represent the lowest
+    number in the smallest width (e.g. int8_t over unsigned, int over
+    int8_t, and float over int).
+
+The following are for converting `ClampedNumeric` instances:
+
+*   `type` - The underlying numeric type.
+*   `RawValue()` - Returns the raw value as the underlying arithmetic type. This
+    is useful when e.g. assigning to an auto type or passing as a deduced
+    template parameter.
+*   `Cast<>()` - Instance method returning a `ClampedNumeric` derived from
+    casting the current instance to a `ClampedNumeric` of the supplied
+    destination type.
+
+### Non-member helper functions
+
+The following variadic convenience functions, which accept standard arithmetic
+or `ClampedNumeric` types, perform arithmetic operations, and return a
+`ClampedNumeric` result. The supported functions are:
+
+*   `ClampAdd()` - Addition.
+*   `ClampSub()` - Subtraction.
+*   `ClampMul()` - Multiplication.
+*   `ClampDiv()` - Division.
+*   `ClampMod()` - Modulus (integer only).
+*   `ClampLsh()` - Left integer shift (integer only).
+*   `ClampRsh()` - Right integer shift (integer only).
+*   `ClampAnd()` - Bitwise AND (integer only with unsigned result).
+*   `ClampOr()`  - Bitwise OR (integer only with unsigned result).
+*   `ClampXor()` - Bitwise XOR (integer only with unsigned result).
+*   `ClampMax()` - Maximum of supplied arguments.
+*   `ClampMin()` - Minimum of supplied arguments.
+
+The following is a general utility method that is useful for converting
+to a `ClampedNumeric` type:
+
+*   `MakeClampedNum()` - Creates a new `ClampedNumeric` from the underlying type
+    of the supplied arithmetic or directly convertible type.

dep/chromium/base/numerics/checked_math.h

@@ -0,0 +1,393 @@
+// Copyright 2017 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#ifndef BASE_NUMERICS_CHECKED_MATH_H_
+#define BASE_NUMERICS_CHECKED_MATH_H_
+
+#include <stddef.h>
+
+#include <limits>
+#include <type_traits>
+
+#include "base/numerics/checked_math_impl.h"
+
+namespace base {
+namespace internal {
+
+template <typename T>
+class CheckedNumeric {
+  static_assert(std::is_arithmetic<T>::value,
+                "CheckedNumeric<T>: T must be a numeric type.");
+
+ public:
+  using type = T;
+
+  constexpr CheckedNumeric() = default;
+
+  // Copy constructor.
+  template <typename Src>
+  constexpr CheckedNumeric(const CheckedNumeric<Src>& rhs)
+      : state_(rhs.state_.value(), rhs.IsValid()) {}
+
+  template <typename Src>
+  friend class CheckedNumeric;
+
+  // This is not an explicit constructor because we implicitly upgrade regular
+  // numerics to CheckedNumerics to make them easier to use.
+  template <typename Src>
+  constexpr CheckedNumeric(Src value)  // NOLINT(runtime/explicit)
+      : state_(value) {
+    static_assert(std::is_arithmetic<Src>::value, "Argument must be numeric.");
+  }
+
+  // This is not an explicit constructor because we want a seamless conversion
+  // from StrictNumeric types.
+  template <typename Src>
+  constexpr CheckedNumeric(
+      StrictNumeric<Src> value)  // NOLINT(runtime/explicit)
+      : state_(static_cast<Src>(value)) {}
+
+  // IsValid() - The public API to test if a CheckedNumeric is currently valid.
+  // A range checked destination type can be supplied using the Dst template
+  // parameter.
+  template <typename Dst = T>
+  constexpr bool IsValid() const {
+    return state_.is_valid() &&
+           IsValueInRangeForNumericType<Dst>(state_.value());
+  }
+
+  // AssignIfValid(Dst) - Assigns the underlying value if it is currently valid
+  // and is within the range supported by the destination type. Returns true if
+  // successful and false otherwise.
+  template <typename Dst>
+#if defined(__clang__) || defined(__GNUC__)
+  __attribute__((warn_unused_result))
+#elif defined(_MSC_VER)
+  _Check_return_
+#endif
+  constexpr bool
+  AssignIfValid(Dst* result) const {
+    return BASE_NUMERICS_LIKELY(IsValid<Dst>())
+               ? ((*result = static_cast<Dst>(state_.value())), true)
+               : false;
+  }
+
+  // ValueOrDie() - The primary accessor for the underlying value. If the
+  // current state is not valid it will CHECK and crash.
+  // A range checked destination type can be supplied using the Dst template
+  // parameter, which will trigger a CHECK if the value is not in bounds for
+  // the destination.
+  // The CHECK behavior can be overridden by supplying a handler as a
+  // template parameter, for test code, etc. However, the handler cannot access
+  // the underlying value, and it is not available through other means.
+  template <typename Dst = T, class CheckHandler = CheckOnFailure>
+  constexpr StrictNumeric<Dst> ValueOrDie() const {
+    return BASE_NUMERICS_LIKELY(IsValid<Dst>())
+               ? static_cast<Dst>(state_.value())
+               : CheckHandler::template HandleFailure<Dst>();
+  }
+
+  // ValueOrDefault(T default_value) - A convenience method that returns the
+  // current value if the state is valid, and the supplied default_value for
+  // any other state.
+  // A range checked destination type can be supplied using the Dst template
+  // parameter. WARNING: This function may fail to compile or CHECK at runtime
+  // if the supplied default_value is not within range of the destination type.
+  template <typename Dst = T, typename Src>
+  constexpr StrictNumeric<Dst> ValueOrDefault(const Src default_value) const {
+    return BASE_NUMERICS_LIKELY(IsValid<Dst>())
+               ? static_cast<Dst>(state_.value())
+               : checked_cast<Dst>(default_value);
+  }
+
+  // Returns a checked numeric of the specified type, cast from the current
+  // CheckedNumeric. If the current state is invalid or the destination cannot
+  // represent the result then the returned CheckedNumeric will be invalid.
+  template <typename Dst>
+  constexpr CheckedNumeric<typename UnderlyingType<Dst>::type> Cast() const {
+    return *this;
+  }
+
+  // This friend method is available solely for providing more detailed logging
+  // in the the tests. Do not implement it in production code, because the
+  // underlying values may change at any time.
+  template <typename U>
+  friend U GetNumericValueForTest(const CheckedNumeric<U>& src);
+
+  // Prototypes for the supported arithmetic operator overloads.
+  template <typename Src>
+  constexpr CheckedNumeric& operator+=(const Src rhs);
+  template <typename Src>
+  constexpr CheckedNumeric& operator-=(const Src rhs);
+  template <typename Src>
+  constexpr CheckedNumeric& operator*=(const Src rhs);
+  template <typename Src>
+  constexpr CheckedNumeric& operator/=(const Src rhs);
+  template <typename Src>
+  constexpr CheckedNumeric& operator%=(const Src rhs);
+  template <typename Src>
+  constexpr CheckedNumeric& operator<<=(const Src rhs);
+  template <typename Src>
+  constexpr CheckedNumeric& operator>>=(const Src rhs);
+  template <typename Src>
+  constexpr CheckedNumeric& operator&=(const Src rhs);
+  template <typename Src>
+  constexpr CheckedNumeric& operator|=(const Src rhs);
+  template <typename Src>
+  constexpr CheckedNumeric& operator^=(const Src rhs);
+
+  constexpr CheckedNumeric operator-() const {
+    // The negation of two's complement int min is int min, so we simply
+    // check for that in the constexpr case.
+    // We use an optimized code path for a known run-time variable.
+    return MustTreatAsConstexpr(state_.value()) || !std::is_signed<T>::value ||
+                   std::is_floating_point<T>::value
+               ? CheckedNumeric<T>(
+                     NegateWrapper(state_.value()),
+                     IsValid() && (!std::is_signed<T>::value ||
+                                   std::is_floating_point<T>::value ||
+                                   NegateWrapper(state_.value()) !=
+                                       std::numeric_limits<T>::lowest()))
+               : FastRuntimeNegate();
+  }
+
+  constexpr CheckedNumeric operator~() const {
+    return CheckedNumeric<decltype(InvertWrapper(T()))>(
+        InvertWrapper(state_.value()), IsValid());
+  }
+
+  constexpr CheckedNumeric Abs() const {
+    return !IsValueNegative(state_.value()) ? *this : -*this;
+  }
+
+  template <typename U>
+  constexpr CheckedNumeric<typename MathWrapper<CheckedMaxOp, T, U>::type> Max(
+      const U rhs) const {
+    using R = typename UnderlyingType<U>::type;
+    using result_type = typename MathWrapper<CheckedMaxOp, T, U>::type;
+    // TODO(jschuh): This can be converted to the MathOp version and remain
+    // constexpr once we have C++14 support.
+    return CheckedNumeric<result_type>(
+        static_cast<result_type>(
+            IsGreater<T, R>::Test(state_.value(), Wrapper<U>::value(rhs))
+                ? state_.value()
+                : Wrapper<U>::value(rhs)),
+        state_.is_valid() && Wrapper<U>::is_valid(rhs));
+  }
+
+  template <typename U>
+  constexpr CheckedNumeric<typename MathWrapper<CheckedMinOp, T, U>::type> Min(
+      const U rhs) const {
+    using R = typename UnderlyingType<U>::type;
+    using result_type = typename MathWrapper<CheckedMinOp, T, U>::type;
+    // TODO(jschuh): This can be converted to the MathOp version and remain
+    // constexpr once we have C++14 support.
+    return CheckedNumeric<result_type>(
+        static_cast<result_type>(
+            IsLess<T, R>::Test(state_.value(), Wrapper<U>::value(rhs))
+                ? state_.value()
+                : Wrapper<U>::value(rhs)),
+        state_.is_valid() && Wrapper<U>::is_valid(rhs));
+  }
+
+  // This function is available only for integral types. It returns an unsigned
+  // integer of the same width as the source type, containing the absolute value
+  // of the source, and properly handling signed min.
+  constexpr CheckedNumeric<typename UnsignedOrFloatForSize<T>::type>
+  UnsignedAbs() const {
+    return CheckedNumeric<typename UnsignedOrFloatForSize<T>::type>(
+        SafeUnsignedAbs(state_.value()), state_.is_valid());
+  }
+
+  constexpr CheckedNumeric& operator++() {
+    *this += 1;
+    return *this;
+  }
+
+  constexpr CheckedNumeric operator++(int) {
+    CheckedNumeric value = *this;
+    *this += 1;
+    return value;
+  }
+
+  constexpr CheckedNumeric& operator--() {
+    *this -= 1;
+    return *this;
+  }
+
+  constexpr CheckedNumeric operator--(int) {
+    CheckedNumeric value = *this;
+    *this -= 1;
+    return value;
+  }
+
+  // These perform the actual math operations on the CheckedNumerics.
+  // Binary arithmetic operations.
+  template <template <typename, typename, typename> class M,
+            typename L,
+            typename R>
+  static constexpr CheckedNumeric MathOp(const L lhs, const R rhs) {
+    using Math = typename MathWrapper<M, L, R>::math;
+    T result = 0;
+    bool is_valid =
+        Wrapper<L>::is_valid(lhs) && Wrapper<R>::is_valid(rhs) &&
+        Math::Do(Wrapper<L>::value(lhs), Wrapper<R>::value(rhs), &result);
+    return CheckedNumeric<T>(result, is_valid);
+  }
+
+  // Assignment arithmetic operations.
+  template <template <typename, typename, typename> class M, typename R>
+  constexpr CheckedNumeric& MathOp(const R rhs) {
+    using Math = typename MathWrapper<M, T, R>::math;
+    T result = 0;  // Using T as the destination saves a range check.
+    bool is_valid = state_.is_valid() && Wrapper<R>::is_valid(rhs) &&
+                    Math::Do(state_.value(), Wrapper<R>::value(rhs), &result);
+    *this = CheckedNumeric<T>(result, is_valid);
+    return *this;
+  }
+
+ private:
+  CheckedNumericState<T> state_;
+
+  CheckedNumeric FastRuntimeNegate() const {
+    T result;
+    bool success = CheckedSubOp<T, T>::Do(T(0), state_.value(), &result);
+    return CheckedNumeric<T>(result, IsValid() && success);
+  }
+
+  template <typename Src>
+  constexpr CheckedNumeric(Src value, bool is_valid)
+      : state_(value, is_valid) {}
+
+  // These wrappers allow us to handle state the same way for both
+  // CheckedNumeric and POD arithmetic types.
+  template <typename Src>
+  struct Wrapper {
+    static constexpr bool is_valid(Src) { return true; }
+    static constexpr Src value(Src value) { return value; }
+  };
+
+  template <typename Src>
+  struct Wrapper<CheckedNumeric<Src>> {
+    static constexpr bool is_valid(const CheckedNumeric<Src> v) {
+      return v.IsValid();
+    }
+    static constexpr Src value(const CheckedNumeric<Src> v) {
+      return v.state_.value();
+    }
+  };
+
+  template <typename Src>
+  struct Wrapper<StrictNumeric<Src>> {
+    static constexpr bool is_valid(const StrictNumeric<Src>) { return true; }
+    static constexpr Src value(const StrictNumeric<Src> v) {
+      return static_cast<Src>(v);
+    }
+  };
+};
+
+// Convenience functions to avoid the ugly template disambiguator syntax.
+template <typename Dst, typename Src>
+constexpr bool IsValidForType(const CheckedNumeric<Src> value) {
+  return value.template IsValid<Dst>();
+}
+
+template <typename Dst, typename Src>
+constexpr StrictNumeric<Dst> ValueOrDieForType(
+    const CheckedNumeric<Src> value) {
+  return value.template ValueOrDie<Dst>();
+}
+
+template <typename Dst, typename Src, typename Default>
+constexpr StrictNumeric<Dst> ValueOrDefaultForType(
+    const CheckedNumeric<Src> value,
+    const Default default_value) {
+  return value.template ValueOrDefault<Dst>(default_value);
+}
+
+// Convience wrapper to return a new CheckedNumeric from the provided arithmetic
+// or CheckedNumericType.
+template <typename T>
+constexpr CheckedNumeric<typename UnderlyingType<T>::type> MakeCheckedNum(
+    const T value) {
+  return value;
+}
+
+// These implement the variadic wrapper for the math operations.
+template <template <typename, typename, typename> class M,
+          typename L,
+          typename R>
+constexpr CheckedNumeric<typename MathWrapper<M, L, R>::type> CheckMathOp(
+    const L lhs,
+    const R rhs) {
+  using Math = typename MathWrapper<M, L, R>::math;
+  return CheckedNumeric<typename Math::result_type>::template MathOp<M>(lhs,
+                                                                        rhs);
+}
+
+// General purpose wrapper template for arithmetic operations.
+template <template <typename, typename, typename> class M,
+          typename L,
+          typename R,
+          typename... Args>
+constexpr CheckedNumeric<typename ResultType<M, L, R, Args...>::type>
+CheckMathOp(const L lhs, const R rhs, const Args... args) {
+  return CheckMathOp<M>(CheckMathOp<M>(lhs, rhs), args...);
+}
+
+BASE_NUMERIC_ARITHMETIC_OPERATORS(Checked, Check, Add, +, +=)
+BASE_NUMERIC_ARITHMETIC_OPERATORS(Checked, Check, Sub, -, -=)
+BASE_NUMERIC_ARITHMETIC_OPERATORS(Checked, Check, Mul, *, *=)
+BASE_NUMERIC_ARITHMETIC_OPERATORS(Checked, Check, Div, /, /=)
+BASE_NUMERIC_ARITHMETIC_OPERATORS(Checked, Check, Mod, %, %=)
+BASE_NUMERIC_ARITHMETIC_OPERATORS(Checked, Check, Lsh, <<, <<=)
+BASE_NUMERIC_ARITHMETIC_OPERATORS(Checked, Check, Rsh, >>, >>=)
+BASE_NUMERIC_ARITHMETIC_OPERATORS(Checked, Check, And, &, &=)
+BASE_NUMERIC_ARITHMETIC_OPERATORS(Checked, Check, Or, |, |=)
+BASE_NUMERIC_ARITHMETIC_OPERATORS(Checked, Check, Xor, ^, ^=)
+BASE_NUMERIC_ARITHMETIC_VARIADIC(Checked, Check, Max)
+BASE_NUMERIC_ARITHMETIC_VARIADIC(Checked, Check, Min)
+
+// These are some extra StrictNumeric operators to support simple pointer
+// arithmetic with our result types. Since wrapping on a pointer is always
+// bad, we trigger the CHECK condition here.
+template <typename L, typename R>
+L* operator+(L* lhs, const StrictNumeric<R> rhs) {
+  uintptr_t result = CheckAdd(reinterpret_cast<uintptr_t>(lhs),
+                              CheckMul(sizeof(L), static_cast<R>(rhs)))
+                         .template ValueOrDie<uintptr_t>();
+  return reinterpret_cast<L*>(result);
+}
+
+template <typename L, typename R>
+L* operator-(L* lhs, const StrictNumeric<R> rhs) {
+  uintptr_t result = CheckSub(reinterpret_cast<uintptr_t>(lhs),
+                              CheckMul(sizeof(L), static_cast<R>(rhs)))
+                         .template ValueOrDie<uintptr_t>();
+  return reinterpret_cast<L*>(result);
+}
+
+}  // namespace internal
+
+using internal::CheckedNumeric;
+using internal::IsValidForType;
+using internal::ValueOrDieForType;
+using internal::ValueOrDefaultForType;
+using internal::MakeCheckedNum;
+using internal::CheckMax;
+using internal::CheckMin;
+using internal::CheckAdd;
+using internal::CheckSub;
+using internal::CheckMul;
+using internal::CheckDiv;
+using internal::CheckMod;
+using internal::CheckLsh;
+using internal::CheckRsh;
+using internal::CheckAnd;
+using internal::CheckOr;
+using internal::CheckXor;
+
+}  // namespace base
+
+#endif  // BASE_NUMERICS_CHECKED_MATH_H_

dep/chromium/base/numerics/checked_math_impl.h

@@ -0,0 +1,567 @@
+// Copyright 2017 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#ifndef BASE_NUMERICS_CHECKED_MATH_IMPL_H_
+#define BASE_NUMERICS_CHECKED_MATH_IMPL_H_
+
+#include <stddef.h>
+#include <stdint.h>
+
+#include <climits>
+#include <cmath>
+#include <cstdlib>
+#include <limits>
+#include <type_traits>
+
+#include "base/numerics/safe_conversions.h"
+#include "base/numerics/safe_math_shared_impl.h"
+
+namespace base {
+namespace internal {
+
+template <typename T>
+constexpr bool CheckedAddImpl(T x, T y, T* result) {
+  static_assert(std::is_integral<T>::value, "Type must be integral");
+  // Since the value of x+y is undefined if we have a signed type, we compute
+  // it using the unsigned type of the same size.
+  using UnsignedDst = typename std::make_unsigned<T>::type;
+  using SignedDst = typename std::make_signed<T>::type;
+  UnsignedDst ux = static_cast<UnsignedDst>(x);
+  UnsignedDst uy = static_cast<UnsignedDst>(y);
+  UnsignedDst uresult = static_cast<UnsignedDst>(ux + uy);
+  *result = static_cast<T>(uresult);
+  // Addition is valid if the sign of (x + y) is equal to either that of x or
+  // that of y.
+  return (std::is_signed<T>::value)
+             ? static_cast<SignedDst>((uresult ^ ux) & (uresult ^ uy)) >= 0
+             : uresult >= uy;  // Unsigned is either valid or underflow.
+}
+
+template <typename T, typename U, class Enable = void>
+struct CheckedAddOp {};
+
+template <typename T, typename U>
+struct CheckedAddOp<T,
+                    U,
+                    typename std::enable_if<std::is_integral<T>::value &&
+                                            std::is_integral<U>::value>::type> {
+  using result_type = typename MaxExponentPromotion<T, U>::type;
+  template <typename V>
+  static constexpr bool Do(T x, U y, V* result) {
+    // TODO(jschuh) Make this "constexpr if" once we're C++17.
+    if (CheckedAddFastOp<T, U>::is_supported)
+      return CheckedAddFastOp<T, U>::Do(x, y, result);
+
+    // Double the underlying type up to a full machine word.
+    using FastPromotion = typename FastIntegerArithmeticPromotion<T, U>::type;
+    using Promotion =
+        typename std::conditional<(IntegerBitsPlusSign<FastPromotion>::value >
+                                   IntegerBitsPlusSign<intptr_t>::value),
+                                  typename BigEnoughPromotion<T, U>::type,
+                                  FastPromotion>::type;
+    // Fail if either operand is out of range for the promoted type.
+    // TODO(jschuh): This could be made to work for a broader range of values.
+    if (BASE_NUMERICS_UNLIKELY(!IsValueInRangeForNumericType<Promotion>(x) ||
+                               !IsValueInRangeForNumericType<Promotion>(y))) {
+      return false;
+    }
+
+    Promotion presult = {};
+    bool is_valid = true;
+    if (IsIntegerArithmeticSafe<Promotion, T, U>::value) {
+      presult = static_cast<Promotion>(x) + static_cast<Promotion>(y);
+    } else {
+      is_valid = CheckedAddImpl(static_cast<Promotion>(x),
+                                static_cast<Promotion>(y), &presult);
+    }
+    *result = static_cast<V>(presult);
+    return is_valid && IsValueInRangeForNumericType<V>(presult);
+  }
+};
+
+template <typename T>
+constexpr bool CheckedSubImpl(T x, T y, T* result) {
+  static_assert(std::is_integral<T>::value, "Type must be integral");
+  // Since the value of x+y is undefined if we have a signed type, we compute
+  // it using the unsigned type of the same size.
+  using UnsignedDst = typename std::make_unsigned<T>::type;
+  using SignedDst = typename std::make_signed<T>::type;
+  UnsignedDst ux = static_cast<UnsignedDst>(x);
+  UnsignedDst uy = static_cast<UnsignedDst>(y);
+  UnsignedDst uresult = static_cast<UnsignedDst>(ux - uy);
+  *result = static_cast<T>(uresult);
+  // Subtraction is valid if either x and y have same sign, or (x-y) and x have
+  // the same sign.
+  return (std::is_signed<T>::value)
+             ? static_cast<SignedDst>((uresult ^ ux) & (ux ^ uy)) >= 0
+             : x >= y;
+}
+
+template <typename T, typename U, class Enable = void>
+struct CheckedSubOp {};
+
+template <typename T, typename U>
+struct CheckedSubOp<T,
+                    U,
+                    typename std::enable_if<std::is_integral<T>::value &&
+                                            std::is_integral<U>::value>::type> {
+  using result_type = typename MaxExponentPromotion<T, U>::type;
+  template <typename V>
+  static constexpr bool Do(T x, U y, V* result) {
+    // TODO(jschuh) Make this "constexpr if" once we're C++17.
+    if (CheckedSubFastOp<T, U>::is_supported)
+      return CheckedSubFastOp<T, U>::Do(x, y, result);
+
+    // Double the underlying type up to a full machine word.
+    using FastPromotion = typename FastIntegerArithmeticPromotion<T, U>::type;
+    using Promotion =
+        typename std::conditional<(IntegerBitsPlusSign<FastPromotion>::value >
+                                   IntegerBitsPlusSign<intptr_t>::value),
+                                  typename BigEnoughPromotion<T, U>::type,
+                                  FastPromotion>::type;
+    // Fail if either operand is out of range for the promoted type.
+    // TODO(jschuh): This could be made to work for a broader range of values.
+    if (BASE_NUMERICS_UNLIKELY(!IsValueInRangeForNumericType<Promotion>(x) ||
+                               !IsValueInRangeForNumericType<Promotion>(y))) {
+      return false;
+    }
+
+    Promotion presult = {};
+    bool is_valid = true;
+    if (IsIntegerArithmeticSafe<Promotion, T, U>::value) {
+      presult = static_cast<Promotion>(x) - static_cast<Promotion>(y);
+    } else {
+      is_valid = CheckedSubImpl(static_cast<Promotion>(x),
+                                static_cast<Promotion>(y), &presult);
+    }
+    *result = static_cast<V>(presult);
+    return is_valid && IsValueInRangeForNumericType<V>(presult);
+  }
+};
+
+template <typename T>
+constexpr bool CheckedMulImpl(T x, T y, T* result) {
+  static_assert(std::is_integral<T>::value, "Type must be integral");
+  // Since the value of x*y is potentially undefined if we have a signed type,
+  // we compute it using the unsigned type of the same size.
+  using UnsignedDst = typename std::make_unsigned<T>::type;
+  using SignedDst = typename std::make_signed<T>::type;
+  const UnsignedDst ux = SafeUnsignedAbs(x);
+  const UnsignedDst uy = SafeUnsignedAbs(y);
+  UnsignedDst uresult = static_cast<UnsignedDst>(ux * uy);
+  const bool is_negative =
+      std::is_signed<T>::value && static_cast<SignedDst>(x ^ y) < 0;
+  *result = is_negative ? 0 - uresult : uresult;
+  // We have a fast out for unsigned identity or zero on the second operand.
+  // After that it's an unsigned overflow check on the absolute value, with
+  // a +1 bound for a negative result.
+  return uy <= UnsignedDst(!std::is_signed<T>::value || is_negative) ||
+         ux <= (std::numeric_limits<T>::max() + UnsignedDst(is_negative)) / uy;
+}
+
+template <typename T, typename U, class Enable = void>
+struct CheckedMulOp {};
+
+template <typename T, typename U>
+struct CheckedMulOp<T,
+                    U,
+                    typename std::enable_if<std::is_integral<T>::value &&
+                                            std::is_integral<U>::value>::type> {
+  using result_type = typename MaxExponentPromotion<T, U>::type;
+  template <typename V>
+  static constexpr bool Do(T x, U y, V* result) {
+    // TODO(jschuh) Make this "constexpr if" once we're C++17.
+    if (CheckedMulFastOp<T, U>::is_supported)
+      return CheckedMulFastOp<T, U>::Do(x, y, result);
+
+    using Promotion = typename FastIntegerArithmeticPromotion<T, U>::type;
+    // Verify the destination type can hold the result (always true for 0).
+    if (BASE_NUMERICS_UNLIKELY((!IsValueInRangeForNumericType<Promotion>(x) ||
+                                !IsValueInRangeForNumericType<Promotion>(y)) &&
+                               x && y)) {
+      return false;
+    }
+
+    Promotion presult = {};
+    bool is_valid = true;
+    if (CheckedMulFastOp<Promotion, Promotion>::is_supported) {
+      // The fast op may be available with the promoted type.
+      is_valid = CheckedMulFastOp<Promotion, Promotion>::Do(x, y, &presult);
+    } else if (IsIntegerArithmeticSafe<Promotion, T, U>::value) {
+      presult = static_cast<Promotion>(x) * static_cast<Promotion>(y);
+    } else {
+      is_valid = CheckedMulImpl(static_cast<Promotion>(x),
+                                static_cast<Promotion>(y), &presult);
+    }
+    *result = static_cast<V>(presult);
+    return is_valid && IsValueInRangeForNumericType<V>(presult);
+  }
+};
+
+// Division just requires a check for a zero denominator or an invalid negation
+// on signed min/-1.
+template <typename T, typename U, class Enable = void>
+struct CheckedDivOp {};
+
+template <typename T, typename U>
+struct CheckedDivOp<T,
+                    U,
+                    typename std::enable_if<std::is_integral<T>::value &&
+                                            std::is_integral<U>::value>::type> {
+  using result_type = typename MaxExponentPromotion<T, U>::type;
+  template <typename V>
+  static constexpr bool Do(T x, U y, V* result) {
+    if (BASE_NUMERICS_UNLIKELY(!y))
+      return false;
+
+    // The overflow check can be compiled away if we don't have the exact
+    // combination of types needed to trigger this case.
+    using Promotion = typename BigEnoughPromotion<T, U>::type;
+    if (BASE_NUMERICS_UNLIKELY(
+            (std::is_signed<T>::value && std::is_signed<U>::value &&
+             IsTypeInRangeForNumericType<T, Promotion>::value &&
+             static_cast<Promotion>(x) ==
+                 std::numeric_limits<Promotion>::lowest() &&
+             y == static_cast<U>(-1)))) {
+      return false;
+    }
+
+    // This branch always compiles away if the above branch wasn't removed.
+    if (BASE_NUMERICS_UNLIKELY((!IsValueInRangeForNumericType<Promotion>(x) ||
+                                !IsValueInRangeForNumericType<Promotion>(y)) &&
+                               x)) {
+      return false;
+    }
+
+    Promotion presult = Promotion(x) / Promotion(y);
+    *result = static_cast<V>(presult);
+    return IsValueInRangeForNumericType<V>(presult);
+  }
+};
+
+template <typename T, typename U, class Enable = void>
+struct CheckedModOp {};
+
+template <typename T, typename U>
+struct CheckedModOp<T,
+                    U,
+                    typename std::enable_if<std::is_integral<T>::value &&
+                                            std::is_integral<U>::value>::type> {
+  using result_type = typename MaxExponentPromotion<T, U>::type;
+  template <typename V>
+  static constexpr bool Do(T x, U y, V* result) {
+    using Promotion = typename BigEnoughPromotion<T, U>::type;
+    if (BASE_NUMERICS_LIKELY(y)) {
+      Promotion presult = static_cast<Promotion>(x) % static_cast<Promotion>(y);
+      *result = static_cast<Promotion>(presult);
+      return IsValueInRangeForNumericType<V>(presult);
+    }
+    return false;
+  }
+};
+
+template <typename T, typename U, class Enable = void>
+struct CheckedLshOp {};
+
+// Left shift. Shifts less than 0 or greater than or equal to the number
+// of bits in the promoted type are undefined. Shifts of negative values
+// are undefined. Otherwise it is defined when the result fits.
+template <typename T, typename U>
+struct CheckedLshOp<T,
+                    U,
+                    typename std::enable_if<std::is_integral<T>::value &&
+                                            std::is_integral<U>::value>::type> {
+  using result_type = T;
+  template <typename V>
+  static constexpr bool Do(T x, U shift, V* result) {
+    // Disallow negative numbers and verify the shift is in bounds.
+    if (BASE_NUMERICS_LIKELY(!IsValueNegative(x) &&
+                             as_unsigned(shift) <
+                                 as_unsigned(std::numeric_limits<T>::digits))) {
+      // Shift as unsigned to avoid undefined behavior.
+      *result = static_cast<V>(as_unsigned(x) << shift);
+      // If the shift can be reversed, we know it was valid.
+      return *result >> shift == x;
+    }
+
+    // Handle the legal corner-case of a full-width signed shift of zero.
+    return std::is_signed<T>::value && !x &&
+           as_unsigned(shift) == as_unsigned(std::numeric_limits<T>::digits);
+  }
+};
+
+template <typename T, typename U, class Enable = void>
+struct CheckedRshOp {};
+
+// Right shift. Shifts less than 0 or greater than or equal to the number
+// of bits in the promoted type are undefined. Otherwise, it is always defined,
+// but a right shift of a negative value is implementation-dependent.
+template <typename T, typename U>
+struct CheckedRshOp<T,
+                    U,
+                    typename std::enable_if<std::is_integral<T>::value &&
+                                            std::is_integral<U>::value>::type> {
+  using result_type = T;
+  template <typename V>
+  static bool Do(T x, U shift, V* result) {
+    // Use the type conversion push negative values out of range.
+    if (BASE_NUMERICS_LIKELY(as_unsigned(shift) <
+                             IntegerBitsPlusSign<T>::value)) {
+      T tmp = x >> shift;
+      *result = static_cast<V>(tmp);
+      return IsValueInRangeForNumericType<V>(tmp);
+    }
+    return false;
+  }
+};
+
+template <typename T, typename U, class Enable = void>
+struct CheckedAndOp {};
+
+// For simplicity we support only unsigned integer results.
+template <typename T, typename U>
+struct CheckedAndOp<T,
+                    U,
+                    typename std::enable_if<std::is_integral<T>::value &&
+                                            std::is_integral<U>::value>::type> {
+  using result_type = typename std::make_unsigned<
+      typename MaxExponentPromotion<T, U>::type>::type;
+  template <typename V>
+  static constexpr bool Do(T x, U y, V* result) {
+    result_type tmp = static_cast<result_type>(x) & static_cast<result_type>(y);
+    *result = static_cast<V>(tmp);
+    return IsValueInRangeForNumericType<V>(tmp);
+  }
+};
+
+template <typename T, typename U, class Enable = void>
+struct CheckedOrOp {};
+
+// For simplicity we support only unsigned integers.
+template <typename T, typename U>
+struct CheckedOrOp<T,
+                   U,
+                   typename std::enable_if<std::is_integral<T>::value &&
+                                           std::is_integral<U>::value>::type> {
+  using result_type = typename std::make_unsigned<
+      typename MaxExponentPromotion<T, U>::type>::type;
+  template <typename V>
+  static constexpr bool Do(T x, U y, V* result) {
+    result_type tmp = static_cast<result_type>(x) | static_cast<result_type>(y);
+    *result = static_cast<V>(tmp);
+    return IsValueInRangeForNumericType<V>(tmp);
+  }
+};
+
+template <typename T, typename U, class Enable = void>
+struct CheckedXorOp {};
+
+// For simplicity we support only unsigned integers.
+template <typename T, typename U>
+struct CheckedXorOp<T,
+                    U,
+                    typename std::enable_if<std::is_integral<T>::value &&
+                                            std::is_integral<U>::value>::type> {
+  using result_type = typename std::make_unsigned<
+      typename MaxExponentPromotion<T, U>::type>::type;
+  template <typename V>
+  static constexpr bool Do(T x, U y, V* result) {
+    result_type tmp = static_cast<result_type>(x) ^ static_cast<result_type>(y);
+    *result = static_cast<V>(tmp);
+    return IsValueInRangeForNumericType<V>(tmp);
+  }
+};
+
+// Max doesn't really need to be implemented this way because it can't fail,
+// but it makes the code much cleaner to use the MathOp wrappers.
+template <typename T, typename U, class Enable = void>
+struct CheckedMaxOp {};
+
+template <typename T, typename U>
+struct CheckedMaxOp<
+    T,
+    U,
+    typename std::enable_if<std::is_arithmetic<T>::value &&
+                            std::is_arithmetic<U>::value>::type> {
+  using result_type = typename MaxExponentPromotion<T, U>::type;
+  template <typename V>
+  static constexpr bool Do(T x, U y, V* result) {
+    result_type tmp = IsGreater<T, U>::Test(x, y) ? static_cast<result_type>(x)
+                                                  : static_cast<result_type>(y);
+    *result = static_cast<V>(tmp);
+    return IsValueInRangeForNumericType<V>(tmp);
+  }
+};
+
+// Min doesn't really need to be implemented this way because it can't fail,
+// but it makes the code much cleaner to use the MathOp wrappers.
+template <typename T, typename U, class Enable = void>
+struct CheckedMinOp {};
+
+template <typename T, typename U>
+struct CheckedMinOp<
+    T,
+    U,
+    typename std::enable_if<std::is_arithmetic<T>::value &&
+                            std::is_arithmetic<U>::value>::type> {
+  using result_type = typename LowestValuePromotion<T, U>::type;
+  template <typename V>
+  static constexpr bool Do(T x, U y, V* result) {
+    result_type tmp = IsLess<T, U>::Test(x, y) ? static_cast<result_type>(x)
+                                               : static_cast<result_type>(y);
+    *result = static_cast<V>(tmp);
+    return IsValueInRangeForNumericType<V>(tmp);
+  }
+};
+
+// This is just boilerplate that wraps the standard floating point arithmetic.
+// A macro isn't the nicest solution, but it beats rewriting these repeatedly.
+#define BASE_FLOAT_ARITHMETIC_OPS(NAME, OP)                              \
+  template <typename T, typename U>                                      \
+  struct Checked##NAME##Op<                                              \
+      T, U,                                                              \
+      typename std::enable_if<std::is_floating_point<T>::value ||        \
+                              std::is_floating_point<U>::value>::type> { \
+    using result_type = typename MaxExponentPromotion<T, U>::type;       \
+    template <typename V>                                                \
+    static constexpr bool Do(T x, U y, V* result) {                      \
+      using Promotion = typename MaxExponentPromotion<T, U>::type;       \
+      Promotion presult = x OP y;                                        \
+      *result = static_cast<V>(presult);                                 \
+      return IsValueInRangeForNumericType<V>(presult);                   \
+    }                                                                    \
+  };
+
+BASE_FLOAT_ARITHMETIC_OPS(Add, +)
+BASE_FLOAT_ARITHMETIC_OPS(Sub, -)
+BASE_FLOAT_ARITHMETIC_OPS(Mul, *)
+BASE_FLOAT_ARITHMETIC_OPS(Div, /)
+
+#undef BASE_FLOAT_ARITHMETIC_OPS
+
+// Floats carry around their validity state with them, but integers do not. So,
+// we wrap the underlying value in a specialization in order to hide that detail
+// and expose an interface via accessors.
+enum NumericRepresentation {
+  NUMERIC_INTEGER,
+  NUMERIC_FLOATING,
+  NUMERIC_UNKNOWN
+};
+
+template <typename NumericType>
+struct GetNumericRepresentation {
+  static const NumericRepresentation value =
+      std::is_integral<NumericType>::value
+          ? NUMERIC_INTEGER
+          : (std::is_floating_point<NumericType>::value ? NUMERIC_FLOATING
+                                                        : NUMERIC_UNKNOWN);
+};
+
+template <typename T,
+          NumericRepresentation type = GetNumericRepresentation<T>::value>
+class CheckedNumericState {};
+
+// Integrals require quite a bit of additional housekeeping to manage state.
+template <typename T>
+class CheckedNumericState<T, NUMERIC_INTEGER> {
+ private:
+  // is_valid_ precedes value_ because member intializers in the constructors
+  // are evaluated in field order, and is_valid_ must be read when initializing
+  // value_.
+  bool is_valid_;
+  T value_;
+
+  // Ensures that a type conversion does not trigger undefined behavior.
+  template <typename Src>
+  static constexpr T WellDefinedConversionOrZero(const Src value,
+                                                 const bool is_valid) {
+    using SrcType = typename internal::UnderlyingType<Src>::type;
+    return (std::is_integral<SrcType>::value || is_valid)
+               ? static_cast<T>(value)
+               : static_cast<T>(0);
+  }
+
+ public:
+  template <typename Src, NumericRepresentation type>
+  friend class CheckedNumericState;
+
+  constexpr CheckedNumericState() : is_valid_(true), value_(0) {}
+
+  template <typename Src>
+  constexpr CheckedNumericState(Src value, bool is_valid)
+      : is_valid_(is_valid && IsValueInRangeForNumericType<T>(value)),
+        value_(WellDefinedConversionOrZero(value, is_valid_)) {
+    static_assert(std::is_arithmetic<Src>::value, "Argument must be numeric.");
+  }
+
+  // Copy constructor.
+  template <typename Src>
+  constexpr CheckedNumericState(const CheckedNumericState<Src>& rhs)
+      : is_valid_(rhs.IsValid()),
+        value_(WellDefinedConversionOrZero(rhs.value(), is_valid_)) {}
+
+  template <typename Src>
+  constexpr explicit CheckedNumericState(Src value)
+      : is_valid_(IsValueInRangeForNumericType<T>(value)),
+        value_(WellDefinedConversionOrZero(value, is_valid_)) {}
+
+  constexpr bool is_valid() const { return is_valid_; }
+  constexpr T value() const { return value_; }
+};
+
+// Floating points maintain their own validity, but need translation wrappers.
+template <typename T>
+class CheckedNumericState<T, NUMERIC_FLOATING> {
+ private:
+  T value_;
+
+  // Ensures that a type conversion does not trigger undefined behavior.
+  template <typename Src>
+  static constexpr T WellDefinedConversionOrNaN(const Src value,
+                                                const bool is_valid) {
+    using SrcType = typename internal::UnderlyingType<Src>::type;
+    return (StaticDstRangeRelationToSrcRange<T, SrcType>::value ==
+                NUMERIC_RANGE_CONTAINED ||
+            is_valid)
+               ? static_cast<T>(value)
+               : std::numeric_limits<T>::quiet_NaN();
+  }
+
+ public:
+  template <typename Src, NumericRepresentation type>
+  friend class CheckedNumericState;
+
+  constexpr CheckedNumericState() : value_(0.0) {}
+
+  template <typename Src>
+  constexpr CheckedNumericState(Src value, bool is_valid)
+      : value_(WellDefinedConversionOrNaN(value, is_valid)) {}
+
+  template <typename Src>
+  constexpr explicit CheckedNumericState(Src value)
+      : value_(WellDefinedConversionOrNaN(
+            value,
+            IsValueInRangeForNumericType<T>(value))) {}
+
+  // Copy constructor.
+  template <typename Src>
+  constexpr CheckedNumericState(const CheckedNumericState<Src>& rhs)
+      : value_(WellDefinedConversionOrNaN(
+            rhs.value(),
+            rhs.is_valid() && IsValueInRangeForNumericType<T>(rhs.value()))) {}
+
+  constexpr bool is_valid() const {
+    // Written this way because std::isfinite is not reliably constexpr.
+    return MustTreatAsConstexpr(value_)
+               ? value_ <= std::numeric_limits<T>::max() &&
+                     value_ >= std::numeric_limits<T>::lowest()
+               : std::isfinite(value_);
+  }
+  constexpr T value() const { return value_; }
+};
+
+}  // namespace internal
+}  // namespace base
+
+#endif  // BASE_NUMERICS_CHECKED_MATH_IMPL_H_

dep/chromium/base/numerics/clamped_math.h

@@ -0,0 +1,264 @@
+// Copyright 2017 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#ifndef BASE_NUMERICS_CLAMPED_MATH_H_
+#define BASE_NUMERICS_CLAMPED_MATH_H_
+
+#include <stddef.h>
+
+#include <limits>
+#include <type_traits>
+
+#include "base/numerics/clamped_math_impl.h"
+
+namespace base {
+namespace internal {
+
+template <typename T>
+class ClampedNumeric {
+  static_assert(std::is_arithmetic<T>::value,
+                "ClampedNumeric<T>: T must be a numeric type.");
+
+ public:
+  using type = T;
+
+  constexpr ClampedNumeric() : value_(0) {}
+
+  // Copy constructor.
+  template <typename Src>
+  constexpr ClampedNumeric(const ClampedNumeric<Src>& rhs)
+      : value_(saturated_cast<T>(rhs.value_)) {}
+
+  template <typename Src>
+  friend class ClampedNumeric;
+
+  // This is not an explicit constructor because we implicitly upgrade regular
+  // numerics to ClampedNumerics to make them easier to use.
+  template <typename Src>
+  constexpr ClampedNumeric(Src value)  // NOLINT(runtime/explicit)
+      : value_(saturated_cast<T>(value)) {
+    static_assert(std::is_arithmetic<Src>::value, "Argument must be numeric.");
+  }
+
+  // This is not an explicit constructor because we want a seamless conversion
+  // from StrictNumeric types.
+  template <typename Src>
+  constexpr ClampedNumeric(
+      StrictNumeric<Src> value)  // NOLINT(runtime/explicit)
+      : value_(saturated_cast<T>(static_cast<Src>(value))) {}
+
+  // Returns a ClampedNumeric of the specified type, cast from the current
+  // ClampedNumeric, and saturated to the destination type.
+  template <typename Dst>
+  constexpr ClampedNumeric<typename UnderlyingType<Dst>::type> Cast() const {
+    return *this;
+  }
+
+  // Prototypes for the supported arithmetic operator overloads.
+  template <typename Src>
+  constexpr ClampedNumeric& operator+=(const Src rhs);
+  template <typename Src>
+  constexpr ClampedNumeric& operator-=(const Src rhs);
+  template <typename Src>
+  constexpr ClampedNumeric& operator*=(const Src rhs);
+  template <typename Src>
+  constexpr ClampedNumeric& operator/=(const Src rhs);
+  template <typename Src>
+  constexpr ClampedNumeric& operator%=(const Src rhs);
+  template <typename Src>
+  constexpr ClampedNumeric& operator<<=(const Src rhs);
+  template <typename Src>
+  constexpr ClampedNumeric& operator>>=(const Src rhs);
+  template <typename Src>
+  constexpr ClampedNumeric& operator&=(const Src rhs);
+  template <typename Src>
+  constexpr ClampedNumeric& operator|=(const Src rhs);
+  template <typename Src>
+  constexpr ClampedNumeric& operator^=(const Src rhs);
+
+  constexpr ClampedNumeric operator-() const {
+    // The negation of two's complement int min is int min, so that's the
+    // only overflow case where we will saturate.
+    return ClampedNumeric<T>(SaturatedNegWrapper(value_));
+  }
+
+  constexpr ClampedNumeric operator~() const {
+    return ClampedNumeric<decltype(InvertWrapper(T()))>(InvertWrapper(value_));
+  }
+
+  constexpr ClampedNumeric Abs() const {
+    // The negation of two's complement int min is int min, so that's the
+    // only overflow case where we will saturate.
+    return ClampedNumeric<T>(SaturatedAbsWrapper(value_));
+  }
+
+  template <typename U>
+  constexpr ClampedNumeric<typename MathWrapper<ClampedMaxOp, T, U>::type> Max(
+      const U rhs) const {
+    using result_type = typename MathWrapper<ClampedMaxOp, T, U>::type;
+    return ClampedNumeric<result_type>(
+        ClampedMaxOp<T, U>::Do(value_, Wrapper<U>::value(rhs)));
+  }
+
+  template <typename U>
+  constexpr ClampedNumeric<typename MathWrapper<ClampedMinOp, T, U>::type> Min(
+      const U rhs) const {
+    using result_type = typename MathWrapper<ClampedMinOp, T, U>::type;
+    return ClampedNumeric<result_type>(
+        ClampedMinOp<T, U>::Do(value_, Wrapper<U>::value(rhs)));
+  }
+
+  // This function is available only for integral types. It returns an unsigned
+  // integer of the same width as the source type, containing the absolute value
+  // of the source, and properly handling signed min.
+  constexpr ClampedNumeric<typename UnsignedOrFloatForSize<T>::type>
+  UnsignedAbs() const {
+    return ClampedNumeric<typename UnsignedOrFloatForSize<T>::type>(
+        SafeUnsignedAbs(value_));
+  }
+
+  constexpr ClampedNumeric& operator++() {
+    *this += 1;
+    return *this;
+  }
+
+  constexpr ClampedNumeric operator++(int) {
+    ClampedNumeric value = *this;
+    *this += 1;
+    return value;
+  }
+
+  constexpr ClampedNumeric& operator--() {
+    *this -= 1;
+    return *this;
+  }
+
+  constexpr ClampedNumeric operator--(int) {
+    ClampedNumeric value = *this;
+    *this -= 1;
+    return value;
+  }
+
+  // These perform the actual math operations on the ClampedNumerics.
+  // Binary arithmetic operations.
+  template <template <typename, typename, typename> class M,
+            typename L,
+            typename R>
+  static constexpr ClampedNumeric MathOp(const L lhs, const R rhs) {
+    using Math = typename MathWrapper<M, L, R>::math;
+    return ClampedNumeric<T>(
+        Math::template Do<T>(Wrapper<L>::value(lhs), Wrapper<R>::value(rhs)));
+  }
+
+  // Assignment arithmetic operations.
+  template <template <typename, typename, typename> class M, typename R>
+  constexpr ClampedNumeric& MathOp(const R rhs) {
+    using Math = typename MathWrapper<M, T, R>::math;
+    *this =
+        ClampedNumeric<T>(Math::template Do<T>(value_, Wrapper<R>::value(rhs)));
+    return *this;
+  }
+
+  template <typename Dst>
+  constexpr operator Dst() const {
+    return saturated_cast<typename ArithmeticOrUnderlyingEnum<Dst>::type>(
+        value_);
+  }
+
+  // This method extracts the raw integer value without saturating it to the
+  // destination type as the conversion operator does. This is useful when
+  // e.g. assigning to an auto type or passing as a deduced template parameter.
+  constexpr T RawValue() const { return value_; }
+
+ private:
+  T value_;
+
+  // These wrappers allow us to handle state the same way for both
+  // ClampedNumeric and POD arithmetic types.
+  template <typename Src>
+  struct Wrapper {
+    static constexpr Src value(Src value) {
+      return static_cast<typename UnderlyingType<Src>::type>(value);
+    }
+  };
+};
+
+// Convience wrapper to return a new ClampedNumeric from the provided arithmetic
+// or ClampedNumericType.
+template <typename T>
+constexpr ClampedNumeric<typename UnderlyingType<T>::type> MakeClampedNum(
+    const T value) {
+  return value;
+}
+
+#if !BASE_NUMERICS_DISABLE_OSTREAM_OPERATORS
+// Overload the ostream output operator to make logging work nicely.
+template <typename T>
+std::ostream& operator<<(std::ostream& os, const ClampedNumeric<T>& value) {
+  os << static_cast<T>(value);
+  return os;
+}
+#endif
+
+// These implement the variadic wrapper for the math operations.
+template <template <typename, typename, typename> class M,
+          typename L,
+          typename R>
+constexpr ClampedNumeric<typename MathWrapper<M, L, R>::type> ClampMathOp(
+    const L lhs,
+    const R rhs) {
+  using Math = typename MathWrapper<M, L, R>::math;
+  return ClampedNumeric<typename Math::result_type>::template MathOp<M>(lhs,
+                                                                        rhs);
+}
+
+// General purpose wrapper template for arithmetic operations.
+template <template <typename, typename, typename> class M,
+          typename L,
+          typename R,
+          typename... Args>
+constexpr ClampedNumeric<typename ResultType<M, L, R, Args...>::type>
+ClampMathOp(const L lhs, const R rhs, const Args... args) {
+  return ClampMathOp<M>(ClampMathOp<M>(lhs, rhs), args...);
+}
+
+BASE_NUMERIC_ARITHMETIC_OPERATORS(Clamped, Clamp, Add, +, +=)
+BASE_NUMERIC_ARITHMETIC_OPERATORS(Clamped, Clamp, Sub, -, -=)
+BASE_NUMERIC_ARITHMETIC_OPERATORS(Clamped, Clamp, Mul, *, *=)
+BASE_NUMERIC_ARITHMETIC_OPERATORS(Clamped, Clamp, Div, /, /=)
+BASE_NUMERIC_ARITHMETIC_OPERATORS(Clamped, Clamp, Mod, %, %=)
+BASE_NUMERIC_ARITHMETIC_OPERATORS(Clamped, Clamp, Lsh, <<, <<=)
+BASE_NUMERIC_ARITHMETIC_OPERATORS(Clamped, Clamp, Rsh, >>, >>=)
+BASE_NUMERIC_ARITHMETIC_OPERATORS(Clamped, Clamp, And, &, &=)
+BASE_NUMERIC_ARITHMETIC_OPERATORS(Clamped, Clamp, Or, |, |=)
+BASE_NUMERIC_ARITHMETIC_OPERATORS(Clamped, Clamp, Xor, ^, ^=)
+BASE_NUMERIC_ARITHMETIC_VARIADIC(Clamped, Clamp, Max)
+BASE_NUMERIC_ARITHMETIC_VARIADIC(Clamped, Clamp, Min)
+BASE_NUMERIC_COMPARISON_OPERATORS(Clamped, IsLess, <)
+BASE_NUMERIC_COMPARISON_OPERATORS(Clamped, IsLessOrEqual, <=)
+BASE_NUMERIC_COMPARISON_OPERATORS(Clamped, IsGreater, >)
+BASE_NUMERIC_COMPARISON_OPERATORS(Clamped, IsGreaterOrEqual, >=)
+BASE_NUMERIC_COMPARISON_OPERATORS(Clamped, IsEqual, ==)
+BASE_NUMERIC_COMPARISON_OPERATORS(Clamped, IsNotEqual, !=)
+
+}  // namespace internal
+
+using internal::ClampedNumeric;
+using internal::MakeClampedNum;
+using internal::ClampMax;
+using internal::ClampMin;
+using internal::ClampAdd;
+using internal::ClampSub;
+using internal::ClampMul;
+using internal::ClampDiv;
+using internal::ClampMod;
+using internal::ClampLsh;
+using internal::ClampRsh;
+using internal::ClampAnd;
+using internal::ClampOr;
+using internal::ClampXor;
+
+}  // namespace base
+
+#endif  // BASE_NUMERICS_CLAMPED_MATH_H_

dep/chromium/base/numerics/clamped_math_impl.h

@@ -0,0 +1,341 @@
+// Copyright 2017 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#ifndef BASE_NUMERICS_CLAMPED_MATH_IMPL_H_
+#define BASE_NUMERICS_CLAMPED_MATH_IMPL_H_
+
+#include <stddef.h>
+#include <stdint.h>
+
+#include <climits>
+#include <cmath>
+#include <cstdlib>
+#include <limits>
+#include <type_traits>
+
+#include "base/numerics/checked_math.h"
+#include "base/numerics/safe_conversions.h"
+#include "base/numerics/safe_math_shared_impl.h"
+
+namespace base {
+namespace internal {
+
+template <typename T,
+          typename std::enable_if<std::is_integral<T>::value &&
+                                  std::is_signed<T>::value>::type* = nullptr>
+constexpr T SaturatedNegWrapper(T value) {
+  return MustTreatAsConstexpr(value) || !ClampedNegFastOp<T>::is_supported
+             ? (NegateWrapper(value) != std::numeric_limits<T>::lowest()
+                    ? NegateWrapper(value)
+                    : std::numeric_limits<T>::max())
+             : ClampedNegFastOp<T>::Do(value);
+}
+
+template <typename T,
+          typename std::enable_if<std::is_integral<T>::value &&
+                                  !std::is_signed<T>::value>::type* = nullptr>
+constexpr T SaturatedNegWrapper(T value) {
+  return T(0);
+}
+
+template <
+    typename T,
+    typename std::enable_if<std::is_floating_point<T>::value>::type* = nullptr>
+constexpr T SaturatedNegWrapper(T value) {
+  return -value;
+}
+
+template <typename T,
+          typename std::enable_if<std::is_integral<T>::value>::type* = nullptr>
+constexpr T SaturatedAbsWrapper(T value) {
+  // The calculation below is a static identity for unsigned types, but for
+  // signed integer types it provides a non-branching, saturated absolute value.
+  // This works because SafeUnsignedAbs() returns an unsigned type, which can
+  // represent the absolute value of all negative numbers of an equal-width
+  // integer type. The call to IsValueNegative() then detects overflow in the
+  // special case of numeric_limits<T>::min(), by evaluating the bit pattern as
+  // a signed integer value. If it is the overflow case, we end up subtracting
+  // one from the unsigned result, thus saturating to numeric_limits<T>::max().
+  return static_cast<T>(SafeUnsignedAbs(value) -
+                        IsValueNegative<T>(SafeUnsignedAbs(value)));
+}
+
+template <
+    typename T,
+    typename std::enable_if<std::is_floating_point<T>::value>::type* = nullptr>
+constexpr T SaturatedAbsWrapper(T value) {
+  return value < 0 ? -value : value;
+}
+
+template <typename T, typename U, class Enable = void>
+struct ClampedAddOp {};
+
+template <typename T, typename U>
+struct ClampedAddOp<T,
+                    U,
+                    typename std::enable_if<std::is_integral<T>::value &&
+                                            std::is_integral<U>::value>::type> {
+  using result_type = typename MaxExponentPromotion<T, U>::type;
+  template <typename V = result_type>
+  static constexpr V Do(T x, U y) {
+    if (ClampedAddFastOp<T, U>::is_supported)
+      return ClampedAddFastOp<T, U>::template Do<V>(x, y);
+
+    static_assert(std::is_same<V, result_type>::value ||
+                      IsTypeInRangeForNumericType<U, V>::value,
+                  "The saturation result cannot be determined from the "
+                  "provided types.");
+    const V saturated = CommonMaxOrMin<V>(IsValueNegative(y));
+    V result = {};
+    return BASE_NUMERICS_LIKELY((CheckedAddOp<T, U>::Do(x, y, &result)))
+               ? result
+               : saturated;
+  }
+};
+
+template <typename T, typename U, class Enable = void>
+struct ClampedSubOp {};
+
+template <typename T, typename U>
+struct ClampedSubOp<T,
+                    U,
+                    typename std::enable_if<std::is_integral<T>::value &&
+                                            std::is_integral<U>::value>::type> {
+  using result_type = typename MaxExponentPromotion<T, U>::type;
+  template <typename V = result_type>
+  static constexpr V Do(T x, U y) {
+    // TODO(jschuh) Make this "constexpr if" once we're C++17.
+    if (ClampedSubFastOp<T, U>::is_supported)
+      return ClampedSubFastOp<T, U>::template Do<V>(x, y);
+
+    static_assert(std::is_same<V, result_type>::value ||
+                      IsTypeInRangeForNumericType<U, V>::value,
+                  "The saturation result cannot be determined from the "
+                  "provided types.");
+    const V saturated = CommonMaxOrMin<V>(!IsValueNegative(y));
+    V result = {};
+    return BASE_NUMERICS_LIKELY((CheckedSubOp<T, U>::Do(x, y, &result)))
+               ? result
+               : saturated;
+  }
+};
+
+template <typename T, typename U, class Enable = void>
+struct ClampedMulOp {};
+
+template <typename T, typename U>
+struct ClampedMulOp<T,
+                    U,
+                    typename std::enable_if<std::is_integral<T>::value &&
+                                            std::is_integral<U>::value>::type> {
+  using result_type = typename MaxExponentPromotion<T, U>::type;
+  template <typename V = result_type>
+  static constexpr V Do(T x, U y) {
+    // TODO(jschuh) Make this "constexpr if" once we're C++17.
+    if (ClampedMulFastOp<T, U>::is_supported)
+      return ClampedMulFastOp<T, U>::template Do<V>(x, y);
+
+    V result = {};
+    const V saturated =
+        CommonMaxOrMin<V>(IsValueNegative(x) ^ IsValueNegative(y));
+    return BASE_NUMERICS_LIKELY((CheckedMulOp<T, U>::Do(x, y, &result)))
+               ? result
+               : saturated;
+  }
+};
+
+template <typename T, typename U, class Enable = void>
+struct ClampedDivOp {};
+
+template <typename T, typename U>
+struct ClampedDivOp<T,
+                    U,
+                    typename std::enable_if<std::is_integral<T>::value &&
+                                            std::is_integral<U>::value>::type> {
+  using result_type = typename MaxExponentPromotion<T, U>::type;
+  template <typename V = result_type>
+  static constexpr V Do(T x, U y) {
+    V result = {};
+    if (BASE_NUMERICS_LIKELY((CheckedDivOp<T, U>::Do(x, y, &result))))
+      return result;
+    // Saturation goes to max, min, or NaN (if x is zero).
+    return x ? CommonMaxOrMin<V>(IsValueNegative(x) ^ IsValueNegative(y))
+             : SaturationDefaultLimits<V>::NaN();
+  }
+};
+
+template <typename T, typename U, class Enable = void>
+struct ClampedModOp {};
+
+template <typename T, typename U>
+struct ClampedModOp<T,
+                    U,
+                    typename std::enable_if<std::is_integral<T>::value &&
+                                            std::is_integral<U>::value>::type> {
+  using result_type = typename MaxExponentPromotion<T, U>::type;
+  template <typename V = result_type>
+  static constexpr V Do(T x, U y) {
+    V result = {};
+    return BASE_NUMERICS_LIKELY((CheckedModOp<T, U>::Do(x, y, &result)))
+               ? result
+               : x;
+  }
+};
+
+template <typename T, typename U, class Enable = void>
+struct ClampedLshOp {};
+
+// Left shift. Non-zero values saturate in the direction of the sign. A zero
+// shifted by any value always results in zero.
+template <typename T, typename U>
+struct ClampedLshOp<T,
+                    U,
+                    typename std::enable_if<std::is_integral<T>::value &&
+                                            std::is_integral<U>::value>::type> {
+  using result_type = T;
+  template <typename V = result_type>
+  static constexpr V Do(T x, U shift) {
+    static_assert(!std::is_signed<U>::value, "Shift value must be unsigned.");
+    if (BASE_NUMERICS_LIKELY(shift < std::numeric_limits<T>::digits)) {
+      // Shift as unsigned to avoid undefined behavior.
+      V result = static_cast<V>(as_unsigned(x) << shift);
+      // If the shift can be reversed, we know it was valid.
+      if (BASE_NUMERICS_LIKELY(result >> shift == x))
+        return result;
+    }
+    return x ? CommonMaxOrMin<V>(IsValueNegative(x)) : 0;
+  }
+};
+
+template <typename T, typename U, class Enable = void>
+struct ClampedRshOp {};
+
+// Right shift. Negative values saturate to -1. Positive or 0 saturates to 0.
+template <typename T, typename U>
+struct ClampedRshOp<T,
+                    U,
+                    typename std::enable_if<std::is_integral<T>::value &&
+                                            std::is_integral<U>::value>::type> {
+  using result_type = T;
+  template <typename V = result_type>
+  static constexpr V Do(T x, U shift) {
+    static_assert(!std::is_signed<U>::value, "Shift value must be unsigned.");
+    // Signed right shift is odd, because it saturates to -1 or 0.
+    const V saturated = as_unsigned(V(0)) - IsValueNegative(x);
+    return BASE_NUMERICS_LIKELY(shift < IntegerBitsPlusSign<T>::value)
+               ? saturated_cast<V>(x >> shift)
+               : saturated;
+  }
+};
+
+template <typename T, typename U, class Enable = void>
+struct ClampedAndOp {};
+
+template <typename T, typename U>
+struct ClampedAndOp<T,
+                    U,
+                    typename std::enable_if<std::is_integral<T>::value &&
+                                            std::is_integral<U>::value>::type> {
+  using result_type = typename std::make_unsigned<
+      typename MaxExponentPromotion<T, U>::type>::type;
+  template <typename V>
+  static constexpr V Do(T x, U y) {
+    return static_cast<result_type>(x) & static_cast<result_type>(y);
+  }
+};
+
+template <typename T, typename U, class Enable = void>
+struct ClampedOrOp {};
+
+// For simplicity we promote to unsigned integers.
+template <typename T, typename U>
+struct ClampedOrOp<T,
+                   U,
+                   typename std::enable_if<std::is_integral<T>::value &&
+                                           std::is_integral<U>::value>::type> {
+  using result_type = typename std::make_unsigned<
+      typename MaxExponentPromotion<T, U>::type>::type;
+  template <typename V>
+  static constexpr V Do(T x, U y) {
+    return static_cast<result_type>(x) | static_cast<result_type>(y);
+  }
+};
+
+template <typename T, typename U, class Enable = void>
+struct ClampedXorOp {};
+
+// For simplicity we support only unsigned integers.
+template <typename T, typename U>
+struct ClampedXorOp<T,
+                    U,
+                    typename std::enable_if<std::is_integral<T>::value &&
+                                            std::is_integral<U>::value>::type> {
+  using result_type = typename std::make_unsigned<
+      typename MaxExponentPromotion<T, U>::type>::type;
+  template <typename V>
+  static constexpr V Do(T x, U y) {
+    return static_cast<result_type>(x) ^ static_cast<result_type>(y);
+  }
+};
+
+template <typename T, typename U, class Enable = void>
+struct ClampedMaxOp {};
+
+template <typename T, typename U>
+struct ClampedMaxOp<
+    T,
+    U,
+    typename std::enable_if<std::is_arithmetic<T>::value &&
+                            std::is_arithmetic<U>::value>::type> {
+  using result_type = typename MaxExponentPromotion<T, U>::type;
+  template <typename V = result_type>
+  static constexpr V Do(T x, U y) {
+    return IsGreater<T, U>::Test(x, y) ? saturated_cast<V>(x)
+                                       : saturated_cast<V>(y);
+  }
+};
+
+template <typename T, typename U, class Enable = void>
+struct ClampedMinOp {};
+
+template <typename T, typename U>
+struct ClampedMinOp<
+    T,
+    U,
+    typename std::enable_if<std::is_arithmetic<T>::value &&
+                            std::is_arithmetic<U>::value>::type> {
+  using result_type = typename LowestValuePromotion<T, U>::type;
+  template <typename V = result_type>
+  static constexpr V Do(T x, U y) {
+    return IsLess<T, U>::Test(x, y) ? saturated_cast<V>(x)
+                                    : saturated_cast<V>(y);
+  }
+};
+
+// This is just boilerplate that wraps the standard floating point arithmetic.
+// A macro isn't the nicest solution, but it beats rewriting these repeatedly.
+#define BASE_FLOAT_ARITHMETIC_OPS(NAME, OP)                              \
+  template <typename T, typename U>                                      \
+  struct Clamped##NAME##Op<                                              \
+      T, U,                                                              \
+      typename std::enable_if<std::is_floating_point<T>::value ||        \
+                              std::is_floating_point<U>::value>::type> { \
+    using result_type = typename MaxExponentPromotion<T, U>::type;       \
+    template <typename V = result_type>                                  \
+    static constexpr V Do(T x, U y) {                                    \
+      return saturated_cast<V>(x OP y);                                  \
+    }                                                                    \
+  };
+
+BASE_FLOAT_ARITHMETIC_OPS(Add, +)
+BASE_FLOAT_ARITHMETIC_OPS(Sub, -)
+BASE_FLOAT_ARITHMETIC_OPS(Mul, *)
+BASE_FLOAT_ARITHMETIC_OPS(Div, /)
+
+#undef BASE_FLOAT_ARITHMETIC_OPS
+
+}  // namespace internal
+}  // namespace base
+
+#endif  // BASE_NUMERICS_CLAMPED_MATH_IMPL_H_

dep/chromium/base/numerics/math_constants.h

@@ -0,0 +1,19 @@
+// Copyright 2017 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#ifndef BASE_NUMERICS_MATH_CONSTANTS_H_
+#define BASE_NUMERICS_MATH_CONSTANTS_H_
+
+namespace base {
+
+constexpr double kPiDouble = 3.14159265358979323846;
+constexpr float kPiFloat = 3.14159265358979323846f;
+
+// The mean acceleration due to gravity on Earth in m/s^2.
+constexpr double kMeanGravityDouble = 9.80665;
+constexpr float kMeanGravityFloat = 9.80665f;
+
+}  // namespace base
+
+#endif  // BASE_NUMERICS_MATH_CONSTANTS_H_

dep/chromium/base/numerics/ranges.h

@@ -0,0 +1,27 @@
+// Copyright 2017 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#ifndef BASE_NUMERICS_RANGES_H_
+#define BASE_NUMERICS_RANGES_H_
+
+#include <algorithm>
+#include <cmath>
+
+namespace base {
+
+// To be replaced with std::clamp() from C++17, someday.
+template <class T>
+constexpr const T& ClampToRange(const T& value, const T& min, const T& max) {
+  return std::min(std::max(value, min), max);
+}
+
+template <typename T>
+constexpr bool IsApproximatelyEqual(T lhs, T rhs, T tolerance) {
+  static_assert(std::is_arithmetic<T>::value, "Argument must be arithmetic");
+  return std::abs(rhs - lhs) <= tolerance;
+}
+
+}  // namespace base
+
+#endif  // BASE_NUMERICS_RANGES_H_

dep/chromium/base/numerics/safe_conversions.h

@@ -0,0 +1,358 @@
+// Copyright 2014 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#ifndef BASE_NUMERICS_SAFE_CONVERSIONS_H_
+#define BASE_NUMERICS_SAFE_CONVERSIONS_H_
+
+#include <stddef.h>
+
+#include <limits>
+#include <type_traits>
+
+#include "base/numerics/safe_conversions_impl.h"
+
+#if !defined(__native_client__) && (defined(__ARMEL__) || defined(__arch64__))
+#include "base/numerics/safe_conversions_arm_impl.h"
+#define BASE_HAS_OPTIMIZED_SAFE_CONVERSIONS (1)
+#else
+#define BASE_HAS_OPTIMIZED_SAFE_CONVERSIONS (0)
+#endif
+
+#if !BASE_NUMERICS_DISABLE_OSTREAM_OPERATORS
+#include <ostream>
+#endif
+
+namespace base {
+namespace internal {
+
+#if !BASE_HAS_OPTIMIZED_SAFE_CONVERSIONS
+template <typename Dst, typename Src>
+struct SaturateFastAsmOp {
+  static const bool is_supported = false;
+  static constexpr Dst Do(Src) {
+    // Force a compile failure if instantiated.
+    return CheckOnFailure::template HandleFailure<Dst>();
+  }
+};
+#endif  // BASE_HAS_OPTIMIZED_SAFE_CONVERSIONS
+#undef BASE_HAS_OPTIMIZED_SAFE_CONVERSIONS
+
+// The following special case a few specific integer conversions where we can
+// eke out better performance than range checking.
+template <typename Dst, typename Src, typename Enable = void>
+struct IsValueInRangeFastOp {
+  static const bool is_supported = false;
+  static constexpr bool Do(Src value) {
+    // Force a compile failure if instantiated.
+    return CheckOnFailure::template HandleFailure<bool>();
+  }
+};
+
+// Signed to signed range comparison.
+template <typename Dst, typename Src>
+struct IsValueInRangeFastOp<
+    Dst,
+    Src,
+    typename std::enable_if<
+        std::is_integral<Dst>::value && std::is_integral<Src>::value &&
+        std::is_signed<Dst>::value && std::is_signed<Src>::value &&
+        !IsTypeInRangeForNumericType<Dst, Src>::value>::type> {
+  static const bool is_supported = true;
+
+  static constexpr bool Do(Src value) {
+    // Just downcast to the smaller type, sign extend it back to the original
+    // type, and then see if it matches the original value.
+    return value == static_cast<Dst>(value);
+  }
+};
+
+// Signed to unsigned range comparison.
+template <typename Dst, typename Src>
+struct IsValueInRangeFastOp<
+    Dst,
+    Src,
+    typename std::enable_if<
+        std::is_integral<Dst>::value && std::is_integral<Src>::value &&
+        !std::is_signed<Dst>::value && std::is_signed<Src>::value &&
+        !IsTypeInRangeForNumericType<Dst, Src>::value>::type> {
+  static const bool is_supported = true;
+
+  static constexpr bool Do(Src value) {
+    // We cast a signed as unsigned to overflow negative values to the top,
+    // then compare against whichever maximum is smaller, as our upper bound.
+    return as_unsigned(value) <= as_unsigned(CommonMax<Src, Dst>());
+  }
+};
+
+// Convenience function that returns true if the supplied value is in range
+// for the destination type.
+template <typename Dst, typename Src>
+constexpr bool IsValueInRangeForNumericType(Src value) {
+  using SrcType = typename internal::UnderlyingType<Src>::type;
+  return internal::IsValueInRangeFastOp<Dst, SrcType>::is_supported
+             ? internal::IsValueInRangeFastOp<Dst, SrcType>::Do(
+                   static_cast<SrcType>(value))
+             : internal::DstRangeRelationToSrcRange<Dst>(
+                   static_cast<SrcType>(value))
+                   .IsValid();
+}
+
+// checked_cast<> is analogous to static_cast<> for numeric types,
+// except that it CHECKs that the specified numeric conversion will not
+// overflow or underflow. NaN source will always trigger a CHECK.
+template <typename Dst,
+          class CheckHandler = internal::CheckOnFailure,
+          typename Src>
+constexpr Dst checked_cast(Src value) {
+  // This throws a compile-time error on evaluating the constexpr if it can be
+  // determined at compile-time as failing, otherwise it will CHECK at runtime.
+  using SrcType = typename internal::UnderlyingType<Src>::type;
+  return BASE_NUMERICS_LIKELY((IsValueInRangeForNumericType<Dst>(value)))
+             ? static_cast<Dst>(static_cast<SrcType>(value))
+             : CheckHandler::template HandleFailure<Dst>();
+}
+
+// Default boundaries for integral/float: max/infinity, lowest/-infinity, 0/NaN.
+// You may provide your own limits (e.g. to saturated_cast) so long as you
+// implement all of the static constexpr member functions in the class below.
+template <typename T>
+struct SaturationDefaultLimits : public std::numeric_limits<T> {
+  static constexpr T NaN() {
+    return std::numeric_limits<T>::has_quiet_NaN
+               ? std::numeric_limits<T>::quiet_NaN()
+               : T();
+  }
+  using std::numeric_limits<T>::max;
+  static constexpr T Overflow() {
+    return std::numeric_limits<T>::has_infinity
+               ? std::numeric_limits<T>::infinity()
+               : std::numeric_limits<T>::max();
+  }
+  using std::numeric_limits<T>::lowest;
+  static constexpr T Underflow() {
+    return std::numeric_limits<T>::has_infinity
+               ? std::numeric_limits<T>::infinity() * -1
+               : std::numeric_limits<T>::lowest();
+  }
+};
+
+template <typename Dst, template <typename> class S, typename Src>
+constexpr Dst saturated_cast_impl(Src value, RangeCheck constraint) {
+  // For some reason clang generates much better code when the branch is
+  // structured exactly this way, rather than a sequence of checks.
+  return !constraint.IsOverflowFlagSet()
+             ? (!constraint.IsUnderflowFlagSet() ? static_cast<Dst>(value)
+                                                 : S<Dst>::Underflow())
+             // Skip this check for integral Src, which cannot be NaN.
+             : (std::is_integral<Src>::value || !constraint.IsUnderflowFlagSet()
+                    ? S<Dst>::Overflow()
+                    : S<Dst>::NaN());
+}
+
+// We can reduce the number of conditions and get slightly better performance
+// for normal signed and unsigned integer ranges. And in the specific case of
+// Arm, we can use the optimized saturation instructions.
+template <typename Dst, typename Src, typename Enable = void>
+struct SaturateFastOp {
+  static const bool is_supported = false;
+  static constexpr Dst Do(Src value) {
+    // Force a compile failure if instantiated.
+    return CheckOnFailure::template HandleFailure<Dst>();
+  }
+};
+
+template <typename Dst, typename Src>
+struct SaturateFastOp<
+    Dst,
+    Src,
+    typename std::enable_if<std::is_integral<Src>::value &&
+                            std::is_integral<Dst>::value &&
+                            SaturateFastAsmOp<Dst, Src>::is_supported>::type> {
+  static const bool is_supported = true;
+  static Dst Do(Src value) { return SaturateFastAsmOp<Dst, Src>::Do(value); }
+};
+
+template <typename Dst, typename Src>
+struct SaturateFastOp<
+    Dst,
+    Src,
+    typename std::enable_if<std::is_integral<Src>::value &&
+                            std::is_integral<Dst>::value &&
+                            !SaturateFastAsmOp<Dst, Src>::is_supported>::type> {
+  static const bool is_supported = true;
+  static Dst Do(Src value) {
+    // The exact order of the following is structured to hit the correct
+    // optimization heuristics across compilers. Do not change without
+    // checking the emitted code.
+    Dst saturated = CommonMaxOrMin<Dst, Src>(
+        IsMaxInRangeForNumericType<Dst, Src>() ||
+        (!IsMinInRangeForNumericType<Dst, Src>() && IsValueNegative(value)));
+    return BASE_NUMERICS_LIKELY(IsValueInRangeForNumericType<Dst>(value))
+               ? static_cast<Dst>(value)
+               : saturated;
+  }
+};
+
+// saturated_cast<> is analogous to static_cast<> for numeric types, except
+// that the specified numeric conversion will saturate by default rather than
+// overflow or underflow, and NaN assignment to an integral will return 0.
+// All boundary condition behaviors can be overriden with a custom handler.
+template <typename Dst,
+          template <typename> class SaturationHandler = SaturationDefaultLimits,
+          typename Src>
+constexpr Dst saturated_cast(Src value) {
+  using SrcType = typename UnderlyingType<Src>::type;
+  return !IsCompileTimeConstant(value) &&
+                 SaturateFastOp<Dst, SrcType>::is_supported &&
+                 std::is_same<SaturationHandler<Dst>,
+                              SaturationDefaultLimits<Dst>>::value
+             ? SaturateFastOp<Dst, SrcType>::Do(static_cast<SrcType>(value))
+             : saturated_cast_impl<Dst, SaturationHandler, SrcType>(
+                   static_cast<SrcType>(value),
+                   DstRangeRelationToSrcRange<Dst, SaturationHandler, SrcType>(
+                       static_cast<SrcType>(value)));
+}
+
+// strict_cast<> is analogous to static_cast<> for numeric types, except that
+// it will cause a compile failure if the destination type is not large enough
+// to contain any value in the source type. It performs no runtime checking.
+template <typename Dst, typename Src>
+constexpr Dst strict_cast(Src value) {
+  using SrcType = typename UnderlyingType<Src>::type;
+  static_assert(UnderlyingType<Src>::is_numeric, "Argument must be numeric.");
+  static_assert(std::is_arithmetic<Dst>::value, "Result must be numeric.");
+
+  // If you got here from a compiler error, it's because you tried to assign
+  // from a source type to a destination type that has insufficient range.
+  // The solution may be to change the destination type you're assigning to,
+  // and use one large enough to represent the source.
+  // Alternatively, you may be better served with the checked_cast<> or
+  // saturated_cast<> template functions for your particular use case.
+  static_assert(StaticDstRangeRelationToSrcRange<Dst, SrcType>::value ==
+                    NUMERIC_RANGE_CONTAINED,
+                "The source type is out of range for the destination type. "
+                "Please see strict_cast<> comments for more information.");
+
+  return static_cast<Dst>(static_cast<SrcType>(value));
+}
+
+// Some wrappers to statically check that a type is in range.
+template <typename Dst, typename Src, class Enable = void>
+struct IsNumericRangeContained {
+  static const bool value = false;
+};
+
+template <typename Dst, typename Src>
+struct IsNumericRangeContained<
+    Dst,
+    Src,
+    typename std::enable_if<ArithmeticOrUnderlyingEnum<Dst>::value &&
+                            ArithmeticOrUnderlyingEnum<Src>::value>::type> {
+  static const bool value = StaticDstRangeRelationToSrcRange<Dst, Src>::value ==
+                            NUMERIC_RANGE_CONTAINED;
+};
+
+// StrictNumeric implements compile time range checking between numeric types by
+// wrapping assignment operations in a strict_cast. This class is intended to be
+// used for function arguments and return types, to ensure the destination type
+// can always contain the source type. This is essentially the same as enforcing
+// -Wconversion in gcc and C4302 warnings on MSVC, but it can be applied
+// incrementally at API boundaries, making it easier to convert code so that it
+// compiles cleanly with truncation warnings enabled.
+// This template should introduce no runtime overhead, but it also provides no
+// runtime checking of any of the associated mathematical operations. Use
+// CheckedNumeric for runtime range checks of the actual value being assigned.
+template <typename T>
+class StrictNumeric {
+ public:
+  using type = T;
+
+  constexpr StrictNumeric() : value_(0) {}
+
+  // Copy constructor.
+  template <typename Src>
+  constexpr StrictNumeric(const StrictNumeric<Src>& rhs)
+      : value_(strict_cast<T>(rhs.value_)) {}
+
+  // This is not an explicit constructor because we implicitly upgrade regular
+  // numerics to StrictNumerics to make them easier to use.
+  template <typename Src>
+  constexpr StrictNumeric(Src value)  // NOLINT(runtime/explicit)
+      : value_(strict_cast<T>(value)) {}
+
+  // If you got here from a compiler error, it's because you tried to assign
+  // from a source type to a destination type that has insufficient range.
+  // The solution may be to change the destination type you're assigning to,
+  // and use one large enough to represent the source.
+  // If you're assigning from a CheckedNumeric<> class, you may be able to use
+  // the AssignIfValid() member function, specify a narrower destination type to
+  // the member value functions (e.g. val.template ValueOrDie<Dst>()), use one
+  // of the value helper functions (e.g. ValueOrDieForType<Dst>(val)).
+  // If you've encountered an _ambiguous overload_ you can use a static_cast<>
+  // to explicitly cast the result to the destination type.
+  // If none of that works, you may be better served with the checked_cast<> or
+  // saturated_cast<> template functions for your particular use case.
+  template <typename Dst,
+            typename std::enable_if<
+                IsNumericRangeContained<Dst, T>::value>::type* = nullptr>
+  constexpr operator Dst() const {
+    return static_cast<typename ArithmeticOrUnderlyingEnum<Dst>::type>(value_);
+  }
+
+ private:
+  const T value_;
+};
+
+// Convience wrapper returns a StrictNumeric from the provided arithmetic type.
+template <typename T>
+constexpr StrictNumeric<typename UnderlyingType<T>::type> MakeStrictNum(
+    const T value) {
+  return value;
+}
+
+#if !BASE_NUMERICS_DISABLE_OSTREAM_OPERATORS
+// Overload the ostream output operator to make logging work nicely.
+template <typename T>
+std::ostream& operator<<(std::ostream& os, const StrictNumeric<T>& value) {
+  os << static_cast<T>(value);
+  return os;
+}
+#endif
+
+#define BASE_NUMERIC_COMPARISON_OPERATORS(CLASS, NAME, OP)              \
+  template <typename L, typename R,                                     \
+            typename std::enable_if<                                    \
+                internal::Is##CLASS##Op<L, R>::value>::type* = nullptr> \
+  constexpr bool operator OP(const L lhs, const R rhs) {                \
+    return SafeCompare<NAME, typename UnderlyingType<L>::type,          \
+                       typename UnderlyingType<R>::type>(lhs, rhs);     \
+  }
+
+BASE_NUMERIC_COMPARISON_OPERATORS(Strict, IsLess, <)
+BASE_NUMERIC_COMPARISON_OPERATORS(Strict, IsLessOrEqual, <=)
+BASE_NUMERIC_COMPARISON_OPERATORS(Strict, IsGreater, >)
+BASE_NUMERIC_COMPARISON_OPERATORS(Strict, IsGreaterOrEqual, >=)
+BASE_NUMERIC_COMPARISON_OPERATORS(Strict, IsEqual, ==)
+BASE_NUMERIC_COMPARISON_OPERATORS(Strict, IsNotEqual, !=)
+
+}  // namespace internal
+
+using internal::as_signed;
+using internal::as_unsigned;
+using internal::checked_cast;
+using internal::strict_cast;
+using internal::saturated_cast;
+using internal::SafeUnsignedAbs;
+using internal::StrictNumeric;
+using internal::MakeStrictNum;
+using internal::IsValueInRangeForNumericType;
+using internal::IsTypeInRangeForNumericType;
+using internal::IsValueNegative;
+
+// Explicitly make a shorter size_t alias for convenience.
+using SizeT = StrictNumeric<size_t>;
+
+}  // namespace base
+
+#endif  // BASE_NUMERICS_SAFE_CONVERSIONS_H_

dep/chromium/base/numerics/safe_conversions_arm_impl.h

@@ -0,0 +1,51 @@
+// Copyright 2017 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#ifndef BASE_NUMERICS_SAFE_CONVERSIONS_ARM_IMPL_H_
+#define BASE_NUMERICS_SAFE_CONVERSIONS_ARM_IMPL_H_
+
+#include <cassert>
+#include <limits>
+#include <type_traits>
+
+#include "base/numerics/safe_conversions_impl.h"
+
+namespace base {
+namespace internal {
+
+// Fast saturation to a destination type.
+template <typename Dst, typename Src>
+struct SaturateFastAsmOp {
+  static constexpr bool is_supported =
+      std::is_signed<Src>::value && std::is_integral<Dst>::value &&
+      std::is_integral<Src>::value &&
+      IntegerBitsPlusSign<Src>::value <= IntegerBitsPlusSign<int32_t>::value &&
+      IntegerBitsPlusSign<Dst>::value <= IntegerBitsPlusSign<int32_t>::value &&
+      !IsTypeInRangeForNumericType<Dst, Src>::value;
+
+  __attribute__((always_inline)) static Dst Do(Src value) {
+    int32_t src = value;
+    typename std::conditional<std::is_signed<Dst>::value, int32_t,
+                              uint32_t>::type result;
+    if (std::is_signed<Dst>::value) {
+      asm("ssat %[dst], %[shift], %[src]"
+          : [dst] "=r"(result)
+          : [src] "r"(src), [shift] "n"(IntegerBitsPlusSign<Dst>::value <= 32
+                                            ? IntegerBitsPlusSign<Dst>::value
+                                            : 32));
+    } else {
+      asm("usat %[dst], %[shift], %[src]"
+          : [dst] "=r"(result)
+          : [src] "r"(src), [shift] "n"(IntegerBitsPlusSign<Dst>::value < 32
+                                            ? IntegerBitsPlusSign<Dst>::value
+                                            : 31));
+    }
+    return static_cast<Dst>(result);
+  }
+};
+
+}  // namespace internal
+}  // namespace base
+
+#endif  // BASE_NUMERICS_SAFE_CONVERSIONS_ARM_IMPL_H_

dep/chromium/base/numerics/safe_conversions_impl.h

@@ -0,0 +1,850 @@
+// Copyright 2014 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#ifndef BASE_NUMERICS_SAFE_CONVERSIONS_IMPL_H_
+#define BASE_NUMERICS_SAFE_CONVERSIONS_IMPL_H_
+
+#include <stdint.h>
+
+#include <limits>
+#include <type_traits>
+
+#if defined(__GNUC__) || defined(__clang__)
+#define BASE_NUMERICS_LIKELY(x) __builtin_expect(!!(x), 1)
+#define BASE_NUMERICS_UNLIKELY(x) __builtin_expect(!!(x), 0)
+#else
+#define BASE_NUMERICS_LIKELY(x) (x)
+#define BASE_NUMERICS_UNLIKELY(x) (x)
+#endif
+
+namespace base {
+namespace internal {
+
+// The std library doesn't provide a binary max_exponent for integers, however
+// we can compute an analog using std::numeric_limits<>::digits.
+template <typename NumericType>
+struct MaxExponent {
+  static const int value = std::is_floating_point<NumericType>::value
+                               ? std::numeric_limits<NumericType>::max_exponent
+                               : std::numeric_limits<NumericType>::digits + 1;
+};
+
+// The number of bits (including the sign) in an integer. Eliminates sizeof
+// hacks.
+template <typename NumericType>
+struct IntegerBitsPlusSign {
+  static const int value = std::numeric_limits<NumericType>::digits +
+                           std::is_signed<NumericType>::value;
+};
+
+// Helper templates for integer manipulations.
+
+template <typename Integer>
+struct PositionOfSignBit {
+  static const size_t value = IntegerBitsPlusSign<Integer>::value - 1;
+};
+
+// Determines if a numeric value is negative without throwing compiler
+// warnings on: unsigned(value) < 0.
+template <typename T,
+          typename std::enable_if<std::is_signed<T>::value>::type* = nullptr>
+constexpr bool IsValueNegative(T value) {
+  static_assert(std::is_arithmetic<T>::value, "Argument must be numeric.");
+  return value < 0;
+}
+
+template <typename T,
+          typename std::enable_if<!std::is_signed<T>::value>::type* = nullptr>
+constexpr bool IsValueNegative(T) {
+  static_assert(std::is_arithmetic<T>::value, "Argument must be numeric.");
+  return false;
+}
+
+// This performs a fast negation, returning a signed value. It works on unsigned
+// arguments, but probably doesn't do what you want for any unsigned value
+// larger than max / 2 + 1 (i.e. signed min cast to unsigned).
+template <typename T>
+constexpr typename std::make_signed<T>::type ConditionalNegate(
+    T x,
+    bool is_negative) {
+  static_assert(std::is_integral<T>::value, "Type must be integral");
+  using SignedT = typename std::make_signed<T>::type;
+  using UnsignedT = typename std::make_unsigned<T>::type;
+  return static_cast<SignedT>(
+      (static_cast<UnsignedT>(x) ^ -SignedT(is_negative)) + is_negative);
+}
+
+// This performs a safe, absolute value via unsigned overflow.
+template <typename T>
+constexpr typename std::make_unsigned<T>::type SafeUnsignedAbs(T value) {
+  static_assert(std::is_integral<T>::value, "Type must be integral");
+  using UnsignedT = typename std::make_unsigned<T>::type;
+  return IsValueNegative(value) ? 0 - static_cast<UnsignedT>(value)
+                                : static_cast<UnsignedT>(value);
+}
+
+// This allows us to switch paths on known compile-time constants.
+#if defined(__clang__) || defined(__GNUC__)
+constexpr bool CanDetectCompileTimeConstant() {
+  return true;
+}
+template <typename T>
+constexpr bool IsCompileTimeConstant(const T v) {
+  return __builtin_constant_p(v);
+}
+#else
+constexpr bool CanDetectCompileTimeConstant() {
+  return false;
+}
+template <typename T>
+constexpr bool IsCompileTimeConstant(const T) {
+  return false;
+}
+#endif
+template <typename T>
+constexpr bool MustTreatAsConstexpr(const T v) {
+  // Either we can't detect a compile-time constant, and must always use the
+  // constexpr path, or we know we have a compile-time constant.
+  return !CanDetectCompileTimeConstant() || IsCompileTimeConstant(v);
+}
+
+// Forces a crash, like a CHECK(false). Used for numeric boundary errors.
+// Also used in a constexpr template to trigger a compilation failure on
+// an error condition.
+struct CheckOnFailure {
+  template <typename T>
+  static T HandleFailure() {
+#if defined(_MSC_VER)
+    __debugbreak();
+#elif defined(__GNUC__) || defined(__clang__)
+    __builtin_trap();
+#else
+    ((void)(*(volatile char*)0 = 0));
+#endif
+    return T();
+  }
+};
+
+enum IntegerRepresentation {
+  INTEGER_REPRESENTATION_UNSIGNED,
+  INTEGER_REPRESENTATION_SIGNED
+};
+
+// A range for a given nunmeric Src type is contained for a given numeric Dst
+// type if both numeric_limits<Src>::max() <= numeric_limits<Dst>::max() and
+// numeric_limits<Src>::lowest() >= numeric_limits<Dst>::lowest() are true.
+// We implement this as template specializations rather than simple static
+// comparisons to ensure type correctness in our comparisons.
+enum NumericRangeRepresentation {
+  NUMERIC_RANGE_NOT_CONTAINED,
+  NUMERIC_RANGE_CONTAINED
+};
+
+// Helper templates to statically determine if our destination type can contain
+// maximum and minimum values represented by the source type.
+
+template <typename Dst,
+          typename Src,
+          IntegerRepresentation DstSign = std::is_signed<Dst>::value
+                                              ? INTEGER_REPRESENTATION_SIGNED
+                                              : INTEGER_REPRESENTATION_UNSIGNED,
+          IntegerRepresentation SrcSign = std::is_signed<Src>::value
+                                              ? INTEGER_REPRESENTATION_SIGNED
+                                              : INTEGER_REPRESENTATION_UNSIGNED>
+struct StaticDstRangeRelationToSrcRange;
+
+// Same sign: Dst is guaranteed to contain Src only if its range is equal or
+// larger.
+template <typename Dst, typename Src, IntegerRepresentation Sign>
+struct StaticDstRangeRelationToSrcRange<Dst, Src, Sign, Sign> {
+  static const NumericRangeRepresentation value =
+      MaxExponent<Dst>::value >= MaxExponent<Src>::value
+          ? NUMERIC_RANGE_CONTAINED
+          : NUMERIC_RANGE_NOT_CONTAINED;
+};
+
+// Unsigned to signed: Dst is guaranteed to contain source only if its range is
+// larger.
+template <typename Dst, typename Src>
+struct StaticDstRangeRelationToSrcRange<Dst,
+                                        Src,
+                                        INTEGER_REPRESENTATION_SIGNED,
+                                        INTEGER_REPRESENTATION_UNSIGNED> {
+  static const NumericRangeRepresentation value =
+      MaxExponent<Dst>::value > MaxExponent<Src>::value
+          ? NUMERIC_RANGE_CONTAINED
+          : NUMERIC_RANGE_NOT_CONTAINED;
+};
+
+// Signed to unsigned: Dst cannot be statically determined to contain Src.
+template <typename Dst, typename Src>
+struct StaticDstRangeRelationToSrcRange<Dst,
+                                        Src,
+                                        INTEGER_REPRESENTATION_UNSIGNED,
+                                        INTEGER_REPRESENTATION_SIGNED> {
+  static const NumericRangeRepresentation value = NUMERIC_RANGE_NOT_CONTAINED;
+};
+
+// This class wraps the range constraints as separate booleans so the compiler
+// can identify constants and eliminate unused code paths.
+class RangeCheck {
+ public:
+  constexpr RangeCheck(bool is_in_lower_bound, bool is_in_upper_bound)
+      : is_underflow_(!is_in_lower_bound), is_overflow_(!is_in_upper_bound) {}
+  constexpr RangeCheck() : is_underflow_(0), is_overflow_(0) {}
+  constexpr bool IsValid() const { return !is_overflow_ && !is_underflow_; }
+  constexpr bool IsInvalid() const { return is_overflow_ && is_underflow_; }
+  constexpr bool IsOverflow() const { return is_overflow_ && !is_underflow_; }
+  constexpr bool IsUnderflow() const { return !is_overflow_ && is_underflow_; }
+  constexpr bool IsOverflowFlagSet() const { return is_overflow_; }
+  constexpr bool IsUnderflowFlagSet() const { return is_underflow_; }
+  constexpr bool operator==(const RangeCheck rhs) const {
+    return is_underflow_ == rhs.is_underflow_ &&
+           is_overflow_ == rhs.is_overflow_;
+  }
+  constexpr bool operator!=(const RangeCheck rhs) const {
+    return !(*this == rhs);
+  }
+
+ private:
+  // Do not change the order of these member variables. The integral conversion
+  // optimization depends on this exact order.
+  const bool is_underflow_;
+  const bool is_overflow_;
+};
+
+// The following helper template addresses a corner case in range checks for
+// conversion from a floating-point type to an integral type of smaller range
+// but larger precision (e.g. float -> unsigned). The problem is as follows:
+//   1. Integral maximum is always one less than a power of two, so it must be
+//      truncated to fit the mantissa of the floating point. The direction of
+//      rounding is implementation defined, but by default it's always IEEE
+//      floats, which round to nearest and thus result in a value of larger
+//      magnitude than the integral value.
+//      Example: float f = UINT_MAX; // f is 4294967296f but UINT_MAX
+//                                   // is 4294967295u.
+//   2. If the floating point value is equal to the promoted integral maximum
+//      value, a range check will erroneously pass.
+//      Example: (4294967296f <= 4294967295u) // This is true due to a precision
+//                                            // loss in rounding up to float.
+//   3. When the floating point value is then converted to an integral, the
+//      resulting value is out of range for the target integral type and
+//      thus is implementation defined.
+//      Example: unsigned u = (float)INT_MAX; // u will typically overflow to 0.
+// To fix this bug we manually truncate the maximum value when the destination
+// type is an integral of larger precision than the source floating-point type,
+// such that the resulting maximum is represented exactly as a floating point.
+template <typename Dst, typename Src, template <typename> class Bounds>
+struct NarrowingRange {
+  using SrcLimits = std::numeric_limits<Src>;
+  using DstLimits = typename std::numeric_limits<Dst>;
+
+  // Computes the mask required to make an accurate comparison between types.
+  static const int kShift =
+      (MaxExponent<Src>::value > MaxExponent<Dst>::value &&
+       SrcLimits::digits < DstLimits::digits)
+          ? (DstLimits::digits - SrcLimits::digits)
+          : 0;
+  template <
+      typename T,
+      typename std::enable_if<std::is_integral<T>::value>::type* = nullptr>
+
+  // Masks out the integer bits that are beyond the precision of the
+  // intermediate type used for comparison.
+  static constexpr T Adjust(T value) {
+    static_assert(std::is_same<T, Dst>::value, "");
+    static_assert(kShift < DstLimits::digits, "");
+    return static_cast<T>(
+        ConditionalNegate(SafeUnsignedAbs(value) & ~((T(1) << kShift) - T(1)),
+                          IsValueNegative(value)));
+  }
+
+  template <typename T,
+            typename std::enable_if<std::is_floating_point<T>::value>::type* =
+                nullptr>
+  static constexpr T Adjust(T value) {
+    static_assert(std::is_same<T, Dst>::value, "");
+    static_assert(kShift == 0, "");
+    return value;
+  }
+
+  static constexpr Dst max() { return Adjust(Bounds<Dst>::max()); }
+  static constexpr Dst lowest() { return Adjust(Bounds<Dst>::lowest()); }
+};
+
+template <typename Dst,
+          typename Src,
+          template <typename> class Bounds,
+          IntegerRepresentation DstSign = std::is_signed<Dst>::value
+                                              ? INTEGER_REPRESENTATION_SIGNED
+                                              : INTEGER_REPRESENTATION_UNSIGNED,
+          IntegerRepresentation SrcSign = std::is_signed<Src>::value
+                                              ? INTEGER_REPRESENTATION_SIGNED
+                                              : INTEGER_REPRESENTATION_UNSIGNED,
+          NumericRangeRepresentation DstRange =
+              StaticDstRangeRelationToSrcRange<Dst, Src>::value>
+struct DstRangeRelationToSrcRangeImpl;
+
+// The following templates are for ranges that must be verified at runtime. We
+// split it into checks based on signedness to avoid confusing casts and
+// compiler warnings on signed an unsigned comparisons.
+
+// Same sign narrowing: The range is contained for normal limits.
+template <typename Dst,
+          typename Src,
+          template <typename> class Bounds,
+          IntegerRepresentation DstSign,
+          IntegerRepresentation SrcSign>
+struct DstRangeRelationToSrcRangeImpl<Dst,
+                                      Src,
+                                      Bounds,
+                                      DstSign,
+                                      SrcSign,
+                                      NUMERIC_RANGE_CONTAINED> {
+  static constexpr RangeCheck Check(Src value) {
+    using SrcLimits = std::numeric_limits<Src>;
+    using DstLimits = NarrowingRange<Dst, Src, Bounds>;
+    return RangeCheck(
+        static_cast<Dst>(SrcLimits::lowest()) >= DstLimits::lowest() ||
+            static_cast<Dst>(value) >= DstLimits::lowest(),
+        static_cast<Dst>(SrcLimits::max()) <= DstLimits::max() ||
+            static_cast<Dst>(value) <= DstLimits::max());
+  }
+};
+
+// Signed to signed narrowing: Both the upper and lower boundaries may be
+// exceeded for standard limits.
+template <typename Dst, typename Src, template <typename> class Bounds>
+struct DstRangeRelationToSrcRangeImpl<Dst,
+                                      Src,
+                                      Bounds,
+                                      INTEGER_REPRESENTATION_SIGNED,
+                                      INTEGER_REPRESENTATION_SIGNED,
+                                      NUMERIC_RANGE_NOT_CONTAINED> {
+  static constexpr RangeCheck Check(Src value) {
+    using DstLimits = NarrowingRange<Dst, Src, Bounds>;
+    return RangeCheck(value >= DstLimits::lowest(), value <= DstLimits::max());
+  }
+};
+
+// Unsigned to unsigned narrowing: Only the upper bound can be exceeded for
+// standard limits.
+template <typename Dst, typename Src, template <typename> class Bounds>
+struct DstRangeRelationToSrcRangeImpl<Dst,
+                                      Src,
+                                      Bounds,
+                                      INTEGER_REPRESENTATION_UNSIGNED,
+                                      INTEGER_REPRESENTATION_UNSIGNED,
+                                      NUMERIC_RANGE_NOT_CONTAINED> {
+  static constexpr RangeCheck Check(Src value) {
+    using DstLimits = NarrowingRange<Dst, Src, Bounds>;
+    return RangeCheck(
+        DstLimits::lowest() == Dst(0) || value >= DstLimits::lowest(),
+        value <= DstLimits::max());
+  }
+};
+
+// Unsigned to signed: Only the upper bound can be exceeded for standard limits.
+template <typename Dst, typename Src, template <typename> class Bounds>
+struct DstRangeRelationToSrcRangeImpl<Dst,
+                                      Src,
+                                      Bounds,
+                                      INTEGER_REPRESENTATION_SIGNED,
+                                      INTEGER_REPRESENTATION_UNSIGNED,
+                                      NUMERIC_RANGE_NOT_CONTAINED> {
+  static constexpr RangeCheck Check(Src value) {
+    using DstLimits = NarrowingRange<Dst, Src, Bounds>;
+    using Promotion = decltype(Src() + Dst());
+    return RangeCheck(DstLimits::lowest() <= Dst(0) ||
+                          static_cast<Promotion>(value) >=
+                              static_cast<Promotion>(DstLimits::lowest()),
+                      static_cast<Promotion>(value) <=
+                          static_cast<Promotion>(DstLimits::max()));
+  }
+};
+
+// Signed to unsigned: The upper boundary may be exceeded for a narrower Dst,
+// and any negative value exceeds the lower boundary for standard limits.
+template <typename Dst, typename Src, template <typename> class Bounds>
+struct DstRangeRelationToSrcRangeImpl<Dst,
+                                      Src,
+                                      Bounds,
+                                      INTEGER_REPRESENTATION_UNSIGNED,
+                                      INTEGER_REPRESENTATION_SIGNED,
+                                      NUMERIC_RANGE_NOT_CONTAINED> {
+  static constexpr RangeCheck Check(Src value) {
+    using SrcLimits = std::numeric_limits<Src>;
+    using DstLimits = NarrowingRange<Dst, Src, Bounds>;
+    using Promotion = decltype(Src() + Dst());
+    return RangeCheck(
+        value >= Src(0) && (DstLimits::lowest() == 0 ||
+                            static_cast<Dst>(value) >= DstLimits::lowest()),
+        static_cast<Promotion>(SrcLimits::max()) <=
+                static_cast<Promotion>(DstLimits::max()) ||
+            static_cast<Promotion>(value) <=
+                static_cast<Promotion>(DstLimits::max()));
+  }
+};
+
+// Simple wrapper for statically checking if a type's range is contained.
+template <typename Dst, typename Src>
+struct IsTypeInRangeForNumericType {
+  static const bool value = StaticDstRangeRelationToSrcRange<Dst, Src>::value ==
+                            NUMERIC_RANGE_CONTAINED;
+};
+
+template <typename Dst,
+          template <typename> class Bounds = std::numeric_limits,
+          typename Src>
+constexpr RangeCheck DstRangeRelationToSrcRange(Src value) {
+  static_assert(std::is_arithmetic<Src>::value, "Argument must be numeric.");
+  static_assert(std::is_arithmetic<Dst>::value, "Result must be numeric.");
+  static_assert(Bounds<Dst>::lowest() < Bounds<Dst>::max(), "");
+  return DstRangeRelationToSrcRangeImpl<Dst, Src, Bounds>::Check(value);
+}
+
+// Integer promotion templates used by the portable checked integer arithmetic.
+template <size_t Size, bool IsSigned>
+struct IntegerForDigitsAndSign;
+
+#define INTEGER_FOR_DIGITS_AND_SIGN(I)                          \
+  template <>                                                   \
+  struct IntegerForDigitsAndSign<IntegerBitsPlusSign<I>::value, \
+                                 std::is_signed<I>::value> {    \
+    using type = I;                                             \
+  }
+
+INTEGER_FOR_DIGITS_AND_SIGN(int8_t);
+INTEGER_FOR_DIGITS_AND_SIGN(uint8_t);
+INTEGER_FOR_DIGITS_AND_SIGN(int16_t);
+INTEGER_FOR_DIGITS_AND_SIGN(uint16_t);
+INTEGER_FOR_DIGITS_AND_SIGN(int32_t);
+INTEGER_FOR_DIGITS_AND_SIGN(uint32_t);
+INTEGER_FOR_DIGITS_AND_SIGN(int64_t);
+INTEGER_FOR_DIGITS_AND_SIGN(uint64_t);
+#undef INTEGER_FOR_DIGITS_AND_SIGN
+
+// WARNING: We have no IntegerForSizeAndSign<16, *>. If we ever add one to
+// support 128-bit math, then the ArithmeticPromotion template below will need
+// to be updated (or more likely replaced with a decltype expression).
+static_assert(IntegerBitsPlusSign<intmax_t>::value == 64,
+              "Max integer size not supported for this toolchain.");
+
+template <typename Integer, bool IsSigned = std::is_signed<Integer>::value>
+struct TwiceWiderInteger {
+  using type =
+      typename IntegerForDigitsAndSign<IntegerBitsPlusSign<Integer>::value * 2,
+                                       IsSigned>::type;
+};
+
+enum ArithmeticPromotionCategory {
+  LEFT_PROMOTION,  // Use the type of the left-hand argument.
+  RIGHT_PROMOTION  // Use the type of the right-hand argument.
+};
+
+// Determines the type that can represent the largest positive value.
+template <typename Lhs,
+          typename Rhs,
+          ArithmeticPromotionCategory Promotion =
+              (MaxExponent<Lhs>::value > MaxExponent<Rhs>::value)
+                  ? LEFT_PROMOTION
+                  : RIGHT_PROMOTION>
+struct MaxExponentPromotion;
+
+template <typename Lhs, typename Rhs>
+struct MaxExponentPromotion<Lhs, Rhs, LEFT_PROMOTION> {
+  using type = Lhs;
+};
+
+template <typename Lhs, typename Rhs>
+struct MaxExponentPromotion<Lhs, Rhs, RIGHT_PROMOTION> {
+  using type = Rhs;
+};
+
+// Determines the type that can represent the lowest arithmetic value.
+template <typename Lhs,
+          typename Rhs,
+          ArithmeticPromotionCategory Promotion =
+              std::is_signed<Lhs>::value
+                  ? (std::is_signed<Rhs>::value
+                         ? (MaxExponent<Lhs>::value > MaxExponent<Rhs>::value
+                                ? LEFT_PROMOTION
+                                : RIGHT_PROMOTION)
+                         : LEFT_PROMOTION)
+                  : (std::is_signed<Rhs>::value
+                         ? RIGHT_PROMOTION
+                         : (MaxExponent<Lhs>::value < MaxExponent<Rhs>::value
+                                ? LEFT_PROMOTION
+                                : RIGHT_PROMOTION))>
+struct LowestValuePromotion;
+
+template <typename Lhs, typename Rhs>
+struct LowestValuePromotion<Lhs, Rhs, LEFT_PROMOTION> {
+  using type = Lhs;
+};
+
+template <typename Lhs, typename Rhs>
+struct LowestValuePromotion<Lhs, Rhs, RIGHT_PROMOTION> {
+  using type = Rhs;
+};
+
+// Determines the type that is best able to represent an arithmetic result.
+template <
+    typename Lhs,
+    typename Rhs = Lhs,
+    bool is_intmax_type =
+        std::is_integral<typename MaxExponentPromotion<Lhs, Rhs>::type>::value&&
+            IntegerBitsPlusSign<typename MaxExponentPromotion<Lhs, Rhs>::type>::
+                value == IntegerBitsPlusSign<intmax_t>::value,
+    bool is_max_exponent =
+        StaticDstRangeRelationToSrcRange<
+            typename MaxExponentPromotion<Lhs, Rhs>::type,
+            Lhs>::value ==
+        NUMERIC_RANGE_CONTAINED&& StaticDstRangeRelationToSrcRange<
+            typename MaxExponentPromotion<Lhs, Rhs>::type,
+            Rhs>::value == NUMERIC_RANGE_CONTAINED>
+struct BigEnoughPromotion;
+
+// The side with the max exponent is big enough.
+template <typename Lhs, typename Rhs, bool is_intmax_type>
+struct BigEnoughPromotion<Lhs, Rhs, is_intmax_type, true> {
+  using type = typename MaxExponentPromotion<Lhs, Rhs>::type;
+  static const bool is_contained = true;
+};
+
+// We can use a twice wider type to fit.
+template <typename Lhs, typename Rhs>
+struct BigEnoughPromotion<Lhs, Rhs, false, false> {
+  using type =
+      typename TwiceWiderInteger<typename MaxExponentPromotion<Lhs, Rhs>::type,
+                                 std::is_signed<Lhs>::value ||
+                                     std::is_signed<Rhs>::value>::type;
+  static const bool is_contained = true;
+};
+
+// No type is large enough.
+template <typename Lhs, typename Rhs>
+struct BigEnoughPromotion<Lhs, Rhs, true, false> {
+  using type = typename MaxExponentPromotion<Lhs, Rhs>::type;
+  static const bool is_contained = false;
+};
+
+// We can statically check if operations on the provided types can wrap, so we
+// can skip the checked operations if they're not needed. So, for an integer we
+// care if the destination type preserves the sign and is twice the width of
+// the source.
+template <typename T, typename Lhs, typename Rhs = Lhs>
+struct IsIntegerArithmeticSafe {
+  static const bool value =
+      !std::is_floating_point<T>::value &&
+      !std::is_floating_point<Lhs>::value &&
+      !std::is_floating_point<Rhs>::value &&
+      std::is_signed<T>::value >= std::is_signed<Lhs>::value &&
+      IntegerBitsPlusSign<T>::value >= (2 * IntegerBitsPlusSign<Lhs>::value) &&
+      std::is_signed<T>::value >= std::is_signed<Rhs>::value &&
+      IntegerBitsPlusSign<T>::value >= (2 * IntegerBitsPlusSign<Rhs>::value);
+};
+
+// Promotes to a type that can represent any possible result of a binary
+// arithmetic operation with the source types.
+template <typename Lhs,
+          typename Rhs,
+          bool is_promotion_possible = IsIntegerArithmeticSafe<
+              typename std::conditional<std::is_signed<Lhs>::value ||
+                                            std::is_signed<Rhs>::value,
+                                        intmax_t,
+                                        uintmax_t>::type,
+              typename MaxExponentPromotion<Lhs, Rhs>::type>::value>
+struct FastIntegerArithmeticPromotion;
+
+template <typename Lhs, typename Rhs>
+struct FastIntegerArithmeticPromotion<Lhs, Rhs, true> {
+  using type =
+      typename TwiceWiderInteger<typename MaxExponentPromotion<Lhs, Rhs>::type,
+                                 std::is_signed<Lhs>::value ||
+                                     std::is_signed<Rhs>::value>::type;
+  static_assert(IsIntegerArithmeticSafe<type, Lhs, Rhs>::value, "");
+  static const bool is_contained = true;
+};
+
+template <typename Lhs, typename Rhs>
+struct FastIntegerArithmeticPromotion<Lhs, Rhs, false> {
+  using type = typename BigEnoughPromotion<Lhs, Rhs>::type;
+  static const bool is_contained = false;
+};
+
+// Extracts the underlying type from an enum.
+template <typename T, bool is_enum = std::is_enum<T>::value>
+struct ArithmeticOrUnderlyingEnum;
+
+template <typename T>
+struct ArithmeticOrUnderlyingEnum<T, true> {
+  using type = typename std::underlying_type<T>::type;
+  static const bool value = std::is_arithmetic<type>::value;
+};
+
+template <typename T>
+struct ArithmeticOrUnderlyingEnum<T, false> {
+  using type = T;
+  static const bool value = std::is_arithmetic<type>::value;
+};
+
+// The following are helper templates used in the CheckedNumeric class.
+template <typename T>
+class CheckedNumeric;
+
+template <typename T>
+class ClampedNumeric;
+
+template <typename T>
+class StrictNumeric;
+
+// Used to treat CheckedNumeric and arithmetic underlying types the same.
+template <typename T>
+struct UnderlyingType {
+  using type = typename ArithmeticOrUnderlyingEnum<T>::type;
+  static const bool is_numeric = std::is_arithmetic<type>::value;
+  static const bool is_checked = false;
+  static const bool is_clamped = false;
+  static const bool is_strict = false;
+};
+
+template <typename T>
+struct UnderlyingType<CheckedNumeric<T>> {
+  using type = T;
+  static const bool is_numeric = true;
+  static const bool is_checked = true;
+  static const bool is_clamped = false;
+  static const bool is_strict = false;
+};
+
+template <typename T>
+struct UnderlyingType<ClampedNumeric<T>> {
+  using type = T;
+  static const bool is_numeric = true;
+  static const bool is_checked = false;
+  static const bool is_clamped = true;
+  static const bool is_strict = false;
+};
+
+template <typename T>
+struct UnderlyingType<StrictNumeric<T>> {
+  using type = T;
+  static const bool is_numeric = true;
+  static const bool is_checked = false;
+  static const bool is_clamped = false;
+  static const bool is_strict = true;
+};
+
+template <typename L, typename R>
+struct IsCheckedOp {
+  static const bool value =
+      UnderlyingType<L>::is_numeric && UnderlyingType<R>::is_numeric &&
+      (UnderlyingType<L>::is_checked || UnderlyingType<R>::is_checked);
+};
+
+template <typename L, typename R>
+struct IsClampedOp {
+  static const bool value =
+      UnderlyingType<L>::is_numeric && UnderlyingType<R>::is_numeric &&
+      (UnderlyingType<L>::is_clamped || UnderlyingType<R>::is_clamped) &&
+      !(UnderlyingType<L>::is_checked || UnderlyingType<R>::is_checked);
+};
+
+template <typename L, typename R>
+struct IsStrictOp {
+  static const bool value =
+      UnderlyingType<L>::is_numeric && UnderlyingType<R>::is_numeric &&
+      (UnderlyingType<L>::is_strict || UnderlyingType<R>::is_strict) &&
+      !(UnderlyingType<L>::is_checked || UnderlyingType<R>::is_checked) &&
+      !(UnderlyingType<L>::is_clamped || UnderlyingType<R>::is_clamped);
+};
+
+// as_signed<> returns the supplied integral value (or integral castable
+// Numeric template) cast as a signed integral of equivalent precision.
+// I.e. it's mostly an alias for: static_cast<std::make_signed<T>::type>(t)
+template <typename Src>
+constexpr typename std::make_signed<
+    typename base::internal::UnderlyingType<Src>::type>::type
+as_signed(const Src value) {
+  static_assert(std::is_integral<decltype(as_signed(value))>::value,
+                "Argument must be a signed or unsigned integer type.");
+  return static_cast<decltype(as_signed(value))>(value);
+}
+
+// as_unsigned<> returns the supplied integral value (or integral castable
+// Numeric template) cast as an unsigned integral of equivalent precision.
+// I.e. it's mostly an alias for: static_cast<std::make_unsigned<T>::type>(t)
+template <typename Src>
+constexpr typename std::make_unsigned<
+    typename base::internal::UnderlyingType<Src>::type>::type
+as_unsigned(const Src value) {
+  static_assert(std::is_integral<decltype(as_unsigned(value))>::value,
+                "Argument must be a signed or unsigned integer type.");
+  return static_cast<decltype(as_unsigned(value))>(value);
+}
+
+template <typename L, typename R>
+constexpr bool IsLessImpl(const L lhs,
+                          const R rhs,
+                          const RangeCheck l_range,
+                          const RangeCheck r_range) {
+  return l_range.IsUnderflow() || r_range.IsOverflow() ||
+         (l_range == r_range &&
+          static_cast<decltype(lhs + rhs)>(lhs) <
+              static_cast<decltype(lhs + rhs)>(rhs));
+}
+
+template <typename L, typename R>
+struct IsLess {
+  static_assert(std::is_arithmetic<L>::value && std::is_arithmetic<R>::value,
+                "Types must be numeric.");
+  static constexpr bool Test(const L lhs, const R rhs) {
+    return IsLessImpl(lhs, rhs, DstRangeRelationToSrcRange<R>(lhs),
+                      DstRangeRelationToSrcRange<L>(rhs));
+  }
+};
+
+template <typename L, typename R>
+constexpr bool IsLessOrEqualImpl(const L lhs,
+                                 const R rhs,
+                                 const RangeCheck l_range,
+                                 const RangeCheck r_range) {
+  return l_range.IsUnderflow() || r_range.IsOverflow() ||
+         (l_range == r_range &&
+          static_cast<decltype(lhs + rhs)>(lhs) <=
+              static_cast<decltype(lhs + rhs)>(rhs));
+}
+
+template <typename L, typename R>
+struct IsLessOrEqual {
+  static_assert(std::is_arithmetic<L>::value && std::is_arithmetic<R>::value,
+                "Types must be numeric.");
+  static constexpr bool Test(const L lhs, const R rhs) {
+    return IsLessOrEqualImpl(lhs, rhs, DstRangeRelationToSrcRange<R>(lhs),
+                             DstRangeRelationToSrcRange<L>(rhs));
+  }
+};
+
+template <typename L, typename R>
+constexpr bool IsGreaterImpl(const L lhs,
+                             const R rhs,
+                             const RangeCheck l_range,
+                             const RangeCheck r_range) {
+  return l_range.IsOverflow() || r_range.IsUnderflow() ||
+         (l_range == r_range &&
+          static_cast<decltype(lhs + rhs)>(lhs) >
+              static_cast<decltype(lhs + rhs)>(rhs));
+}
+
+template <typename L, typename R>
+struct IsGreater {
+  static_assert(std::is_arithmetic<L>::value && std::is_arithmetic<R>::value,
+                "Types must be numeric.");
+  static constexpr bool Test(const L lhs, const R rhs) {
+    return IsGreaterImpl(lhs, rhs, DstRangeRelationToSrcRange<R>(lhs),
+                         DstRangeRelationToSrcRange<L>(rhs));
+  }
+};
+
+template <typename L, typename R>
+constexpr bool IsGreaterOrEqualImpl(const L lhs,
+                                    const R rhs,
+                                    const RangeCheck l_range,
+                                    const RangeCheck r_range) {
+  return l_range.IsOverflow() || r_range.IsUnderflow() ||
+         (l_range == r_range &&
+          static_cast<decltype(lhs + rhs)>(lhs) >=
+              static_cast<decltype(lhs + rhs)>(rhs));
+}
+
+template <typename L, typename R>
+struct IsGreaterOrEqual {
+  static_assert(std::is_arithmetic<L>::value && std::is_arithmetic<R>::value,
+                "Types must be numeric.");
+  static constexpr bool Test(const L lhs, const R rhs) {
+    return IsGreaterOrEqualImpl(lhs, rhs, DstRangeRelationToSrcRange<R>(lhs),
+                                DstRangeRelationToSrcRange<L>(rhs));
+  }
+};
+
+template <typename L, typename R>
+struct IsEqual {
+  static_assert(std::is_arithmetic<L>::value && std::is_arithmetic<R>::value,
+                "Types must be numeric.");
+  static constexpr bool Test(const L lhs, const R rhs) {
+    return DstRangeRelationToSrcRange<R>(lhs) ==
+               DstRangeRelationToSrcRange<L>(rhs) &&
+           static_cast<decltype(lhs + rhs)>(lhs) ==
+               static_cast<decltype(lhs + rhs)>(rhs);
+  }
+};
+
+template <typename L, typename R>
+struct IsNotEqual {
+  static_assert(std::is_arithmetic<L>::value && std::is_arithmetic<R>::value,
+                "Types must be numeric.");
+  static constexpr bool Test(const L lhs, const R rhs) {
+    return DstRangeRelationToSrcRange<R>(lhs) !=
+               DstRangeRelationToSrcRange<L>(rhs) ||
+           static_cast<decltype(lhs + rhs)>(lhs) !=
+               static_cast<decltype(lhs + rhs)>(rhs);
+  }
+};
+
+// These perform the actual math operations on the CheckedNumerics.
+// Binary arithmetic operations.
+template <template <typename, typename> class C, typename L, typename R>
+constexpr bool SafeCompare(const L lhs, const R rhs) {
+  static_assert(std::is_arithmetic<L>::value && std::is_arithmetic<R>::value,
+                "Types must be numeric.");
+  using Promotion = BigEnoughPromotion<L, R>;
+  using BigType = typename Promotion::type;
+  return Promotion::is_contained
+             // Force to a larger type for speed if both are contained.
+             ? C<BigType, BigType>::Test(
+                   static_cast<BigType>(static_cast<L>(lhs)),
+                   static_cast<BigType>(static_cast<R>(rhs)))
+             // Let the template functions figure it out for mixed types.
+             : C<L, R>::Test(lhs, rhs);
+}
+
+template <typename Dst, typename Src>
+constexpr bool IsMaxInRangeForNumericType() {
+  return IsGreaterOrEqual<Dst, Src>::Test(std::numeric_limits<Dst>::max(),
+                                          std::numeric_limits<Src>::max());
+}
+
+template <typename Dst, typename Src>
+constexpr bool IsMinInRangeForNumericType() {
+  return IsLessOrEqual<Dst, Src>::Test(std::numeric_limits<Dst>::lowest(),
+                                       std::numeric_limits<Src>::lowest());
+}
+
+template <typename Dst, typename Src>
+constexpr Dst CommonMax() {
+  return !IsMaxInRangeForNumericType<Dst, Src>()
+             ? Dst(std::numeric_limits<Dst>::max())
+             : Dst(std::numeric_limits<Src>::max());
+}
+
+template <typename Dst, typename Src>
+constexpr Dst CommonMin() {
+  return !IsMinInRangeForNumericType<Dst, Src>()
+             ? Dst(std::numeric_limits<Dst>::lowest())
+             : Dst(std::numeric_limits<Src>::lowest());
+}
+
+// This is a wrapper to generate return the max or min for a supplied type.
+// If the argument is false, the returned value is the maximum. If true the
+// returned value is the minimum.
+template <typename Dst, typename Src = Dst>
+constexpr Dst CommonMaxOrMin(bool is_min) {
+  return is_min ? CommonMin<Dst, Src>() : CommonMax<Dst, Src>();
+}
+
+}  // namespace internal
+}  // namespace base
+
+#endif  // BASE_NUMERICS_SAFE_CONVERSIONS_IMPL_H_

dep/chromium/base/numerics/safe_math.h

@@ -0,0 +1,12 @@
+// Copyright 2017 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#ifndef BASE_NUMERICS_SAFE_MATH_H_
+#define BASE_NUMERICS_SAFE_MATH_H_
+
+#include "base/numerics/checked_math.h"
+#include "base/numerics/clamped_math.h"
+#include "base/numerics/safe_conversions.h"
+
+#endif  // BASE_NUMERICS_SAFE_MATH_H_

dep/chromium/base/numerics/safe_math_arm_impl.h

@@ -0,0 +1,122 @@
+// Copyright 2017 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#ifndef BASE_NUMERICS_SAFE_MATH_ARM_IMPL_H_
+#define BASE_NUMERICS_SAFE_MATH_ARM_IMPL_H_
+
+#include <cassert>
+#include <limits>
+#include <type_traits>
+
+#include "base/numerics/safe_conversions.h"
+
+namespace base {
+namespace internal {
+
+template <typename T, typename U>
+struct CheckedMulFastAsmOp {
+  static const bool is_supported =
+      FastIntegerArithmeticPromotion<T, U>::is_contained;
+
+  // The following is much more efficient than the Clang and GCC builtins for
+  // performing overflow-checked multiplication when a twice wider type is
+  // available. The below compiles down to 2-3 instructions, depending on the
+  // width of the types in use.
+  // As an example, an int32_t multiply compiles to:
+  //    smull   r0, r1, r0, r1
+  //    cmp     r1, r1, asr #31
+  // And an int16_t multiply compiles to:
+  //    smulbb  r1, r1, r0
+  //    asr     r2, r1, #16
+  //    cmp     r2, r1, asr #15
+  template <typename V>
+  __attribute__((always_inline)) static bool Do(T x, U y, V* result) {
+    using Promotion = typename FastIntegerArithmeticPromotion<T, U>::type;
+    Promotion presult;
+
+    presult = static_cast<Promotion>(x) * static_cast<Promotion>(y);
+    *result = static_cast<V>(presult);
+    return IsValueInRangeForNumericType<V>(presult);
+  }
+};
+
+template <typename T, typename U>
+struct ClampedAddFastAsmOp {
+  static const bool is_supported =
+      BigEnoughPromotion<T, U>::is_contained &&
+      IsTypeInRangeForNumericType<
+          int32_t,
+          typename BigEnoughPromotion<T, U>::type>::value;
+
+  template <typename V>
+  __attribute__((always_inline)) static V Do(T x, U y) {
+    // This will get promoted to an int, so let the compiler do whatever is
+    // clever and rely on the saturated cast to bounds check.
+    if (IsIntegerArithmeticSafe<int, T, U>::value)
+      return saturated_cast<V>(x + y);
+
+    int32_t result;
+    int32_t x_i32 = checked_cast<int32_t>(x);
+    int32_t y_i32 = checked_cast<int32_t>(y);
+
+    asm("qadd %[result], %[first], %[second]"
+        : [result] "=r"(result)
+        : [first] "r"(x_i32), [second] "r"(y_i32));
+    return saturated_cast<V>(result);
+  }
+};
+
+template <typename T, typename U>
+struct ClampedSubFastAsmOp {
+  static const bool is_supported =
+      BigEnoughPromotion<T, U>::is_contained &&
+      IsTypeInRangeForNumericType<
+          int32_t,
+          typename BigEnoughPromotion<T, U>::type>::value;
+
+  template <typename V>
+  __attribute__((always_inline)) static V Do(T x, U y) {
+    // This will get promoted to an int, so let the compiler do whatever is
+    // clever and rely on the saturated cast to bounds check.
+    if (IsIntegerArithmeticSafe<int, T, U>::value)
+      return saturated_cast<V>(x - y);
+
+    int32_t result;
+    int32_t x_i32 = checked_cast<int32_t>(x);
+    int32_t y_i32 = checked_cast<int32_t>(y);
+
+    asm("qsub %[result], %[first], %[second]"
+        : [result] "=r"(result)
+        : [first] "r"(x_i32), [second] "r"(y_i32));
+    return saturated_cast<V>(result);
+  }
+};
+
+template <typename T, typename U>
+struct ClampedMulFastAsmOp {
+  static const bool is_supported = CheckedMulFastAsmOp<T, U>::is_supported;
+
+  template <typename V>
+  __attribute__((always_inline)) static V Do(T x, U y) {
+    // Use the CheckedMulFastAsmOp for full-width 32-bit values, because
+    // it's fewer instructions than promoting and then saturating.
+    if (!IsIntegerArithmeticSafe<int32_t, T, U>::value &&
+        !IsIntegerArithmeticSafe<uint32_t, T, U>::value) {
+      V result;
+      if (CheckedMulFastAsmOp<T, U>::Do(x, y, &result))
+        return result;
+      return CommonMaxOrMin<V>(IsValueNegative(x) ^ IsValueNegative(y));
+    }
+
+    assert((FastIntegerArithmeticPromotion<T, U>::is_contained));
+    using Promotion = typename FastIntegerArithmeticPromotion<T, U>::type;
+    return saturated_cast<V>(static_cast<Promotion>(x) *
+                             static_cast<Promotion>(y));
+  }
+};
+
+}  // namespace internal
+}  // namespace base
+
+#endif  // BASE_NUMERICS_SAFE_MATH_ARM_IMPL_H_

dep/chromium/base/numerics/safe_math_clang_gcc_impl.h

@@ -0,0 +1,157 @@
+// Copyright 2017 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#ifndef BASE_NUMERICS_SAFE_MATH_CLANG_GCC_IMPL_H_
+#define BASE_NUMERICS_SAFE_MATH_CLANG_GCC_IMPL_H_
+
+#include <cassert>
+#include <limits>
+#include <type_traits>
+
+#include "base/numerics/safe_conversions.h"
+
+#if !defined(__native_client__) && (defined(__ARMEL__) || defined(__arch64__))
+#include "base/numerics/safe_math_arm_impl.h"
+#define BASE_HAS_ASSEMBLER_SAFE_MATH (1)
+#else
+#define BASE_HAS_ASSEMBLER_SAFE_MATH (0)
+#endif
+
+namespace base {
+namespace internal {
+
+// These are the non-functioning boilerplate implementations of the optimized
+// safe math routines.
+#if !BASE_HAS_ASSEMBLER_SAFE_MATH
+template <typename T, typename U>
+struct CheckedMulFastAsmOp {
+  static const bool is_supported = false;
+  template <typename V>
+  static constexpr bool Do(T, U, V*) {
+    // Force a compile failure if instantiated.
+    return CheckOnFailure::template HandleFailure<bool>();
+  }
+};
+
+template <typename T, typename U>
+struct ClampedAddFastAsmOp {
+  static const bool is_supported = false;
+  template <typename V>
+  static constexpr V Do(T, U) {
+    // Force a compile failure if instantiated.
+    return CheckOnFailure::template HandleFailure<V>();
+  }
+};
+
+template <typename T, typename U>
+struct ClampedSubFastAsmOp {
+  static const bool is_supported = false;
+  template <typename V>
+  static constexpr V Do(T, U) {
+    // Force a compile failure if instantiated.
+    return CheckOnFailure::template HandleFailure<V>();
+  }
+};
+
+template <typename T, typename U>
+struct ClampedMulFastAsmOp {
+  static const bool is_supported = false;
+  template <typename V>
+  static constexpr V Do(T, U) {
+    // Force a compile failure if instantiated.
+    return CheckOnFailure::template HandleFailure<V>();
+  }
+};
+#endif  // BASE_HAS_ASSEMBLER_SAFE_MATH
+#undef BASE_HAS_ASSEMBLER_SAFE_MATH
+
+template <typename T, typename U>
+struct CheckedAddFastOp {
+  static const bool is_supported = true;
+  template <typename V>
+  __attribute__((always_inline)) static constexpr bool Do(T x, U y, V* result) {
+    return !__builtin_add_overflow(x, y, result);
+  }
+};
+
+template <typename T, typename U>
+struct CheckedSubFastOp {
+  static const bool is_supported = true;
+  template <typename V>
+  __attribute__((always_inline)) static constexpr bool Do(T x, U y, V* result) {
+    return !__builtin_sub_overflow(x, y, result);
+  }
+};
+
+template <typename T, typename U>
+struct CheckedMulFastOp {
+#if defined(__clang__)
+  // TODO(jschuh): Get the Clang runtime library issues sorted out so we can
+  // support full-width, mixed-sign multiply builtins.
+  // https://crbug.com/613003
+  // We can support intptr_t, uintptr_t, or a smaller common type.
+  static const bool is_supported =
+      (IsTypeInRangeForNumericType<intptr_t, T>::value &&
+       IsTypeInRangeForNumericType<intptr_t, U>::value) ||
+      (IsTypeInRangeForNumericType<uintptr_t, T>::value &&
+       IsTypeInRangeForNumericType<uintptr_t, U>::value);
+#else
+  static const bool is_supported = true;
+#endif
+  template <typename V>
+  __attribute__((always_inline)) static constexpr bool Do(T x, U y, V* result) {
+    return CheckedMulFastAsmOp<T, U>::is_supported
+               ? CheckedMulFastAsmOp<T, U>::Do(x, y, result)
+               : !__builtin_mul_overflow(x, y, result);
+  }
+};
+
+template <typename T, typename U>
+struct ClampedAddFastOp {
+  static const bool is_supported = ClampedAddFastAsmOp<T, U>::is_supported;
+  template <typename V>
+  __attribute__((always_inline)) static V Do(T x, U y) {
+    return ClampedAddFastAsmOp<T, U>::template Do<V>(x, y);
+  }
+};
+
+template <typename T, typename U>
+struct ClampedSubFastOp {
+  static const bool is_supported = ClampedSubFastAsmOp<T, U>::is_supported;
+  template <typename V>
+  __attribute__((always_inline)) static V Do(T x, U y) {
+    return ClampedSubFastAsmOp<T, U>::template Do<V>(x, y);
+  }
+};
+
+template <typename T, typename U>
+struct ClampedMulFastOp {
+  static const bool is_supported = ClampedMulFastAsmOp<T, U>::is_supported;
+  template <typename V>
+  __attribute__((always_inline)) static V Do(T x, U y) {
+    return ClampedMulFastAsmOp<T, U>::template Do<V>(x, y);
+  }
+};
+
+template <typename T>
+struct ClampedNegFastOp {
+  static const bool is_supported = std::is_signed<T>::value;
+  __attribute__((always_inline)) static T Do(T value) {
+    // Use this when there is no assembler path available.
+    if (!ClampedSubFastAsmOp<T, T>::is_supported) {
+      T result;
+      return !__builtin_sub_overflow(T(0), value, &result)
+                 ? result
+                 : std::numeric_limits<T>::max();
+    }
+
+    // Fallback to the normal subtraction path.
+    return ClampedSubFastOp<T, T>::template Do<T>(T(0), value);
+  }
+};
+
+}  // namespace internal
+}  // namespace base
+
+#endif  // BASE_NUMERICS_SAFE_MATH_CLANG_GCC_IMPL_H_

dep/chromium/base/numerics/safe_math_shared_impl.h

@@ -0,0 +1,240 @@
+// Copyright 2017 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#ifndef BASE_NUMERICS_SAFE_MATH_SHARED_IMPL_H_
+#define BASE_NUMERICS_SAFE_MATH_SHARED_IMPL_H_
+
+#include <stddef.h>
+#include <stdint.h>
+
+#include <cassert>
+#include <climits>
+#include <cmath>
+#include <cstdlib>
+#include <limits>
+#include <type_traits>
+
+#include "base/numerics/safe_conversions.h"
+
+#ifdef __asmjs__
+// Optimized safe math instructions are incompatible with asmjs.
+#define BASE_HAS_OPTIMIZED_SAFE_MATH (0)
+// Where available use builtin math overflow support on Clang and GCC.
+#elif !defined(__native_client__) &&                         \
+      ((defined(__clang__) &&                                \
+        ((__clang_major__ > 3) ||                            \
+         (__clang_major__ == 3 && __clang_minor__ >= 4))) || \
+       (defined(__GNUC__) && __GNUC__ >= 5))
+#include "base/numerics/safe_math_clang_gcc_impl.h"
+#define BASE_HAS_OPTIMIZED_SAFE_MATH (1)
+#else
+#define BASE_HAS_OPTIMIZED_SAFE_MATH (0)
+#endif
+
+namespace base {
+namespace internal {
+
+// These are the non-functioning boilerplate implementations of the optimized
+// safe math routines.
+#if !BASE_HAS_OPTIMIZED_SAFE_MATH
+template <typename T, typename U>
+struct CheckedAddFastOp {
+  static const bool is_supported = false;
+  template <typename V>
+  static constexpr bool Do(T, U, V*) {
+    // Force a compile failure if instantiated.
+    return CheckOnFailure::template HandleFailure<bool>();
+  }
+};
+
+template <typename T, typename U>
+struct CheckedSubFastOp {
+  static const bool is_supported = false;
+  template <typename V>
+  static constexpr bool Do(T, U, V*) {
+    // Force a compile failure if instantiated.
+    return CheckOnFailure::template HandleFailure<bool>();
+  }
+};
+
+template <typename T, typename U>
+struct CheckedMulFastOp {
+  static const bool is_supported = false;
+  template <typename V>
+  static constexpr bool Do(T, U, V*) {
+    // Force a compile failure if instantiated.
+    return CheckOnFailure::template HandleFailure<bool>();
+  }
+};
+
+template <typename T, typename U>
+struct ClampedAddFastOp {
+  static const bool is_supported = false;
+  template <typename V>
+  static constexpr V Do(T, U) {
+    // Force a compile failure if instantiated.
+    return CheckOnFailure::template HandleFailure<V>();
+  }
+};
+
+template <typename T, typename U>
+struct ClampedSubFastOp {
+  static const bool is_supported = false;
+  template <typename V>
+  static constexpr V Do(T, U) {
+    // Force a compile failure if instantiated.
+    return CheckOnFailure::template HandleFailure<V>();
+  }
+};
+
+template <typename T, typename U>
+struct ClampedMulFastOp {
+  static const bool is_supported = false;
+  template <typename V>
+  static constexpr V Do(T, U) {
+    // Force a compile failure if instantiated.
+    return CheckOnFailure::template HandleFailure<V>();
+  }
+};
+
+template <typename T>
+struct ClampedNegFastOp {
+  static const bool is_supported = false;
+  static constexpr T Do(T) {
+    // Force a compile failure if instantiated.
+    return CheckOnFailure::template HandleFailure<T>();
+  }
+};
+#endif  // BASE_HAS_OPTIMIZED_SAFE_MATH
+#undef BASE_HAS_OPTIMIZED_SAFE_MATH
+
+// This is used for UnsignedAbs, where we need to support floating-point
+// template instantiations even though we don't actually support the operations.
+// However, there is no corresponding implementation of e.g. SafeUnsignedAbs,
+// so the float versions will not compile.
+template <typename Numeric,
+          bool IsInteger = std::is_integral<Numeric>::value,
+          bool IsFloat = std::is_floating_point<Numeric>::value>
+struct UnsignedOrFloatForSize;
+
+template <typename Numeric>
+struct UnsignedOrFloatForSize<Numeric, true, false> {
+  using type = typename std::make_unsigned<Numeric>::type;
+};
+
+template <typename Numeric>
+struct UnsignedOrFloatForSize<Numeric, false, true> {
+  using type = Numeric;
+};
+
+// Wrap the unary operations to allow SFINAE when instantiating integrals versus
+// floating points. These don't perform any overflow checking. Rather, they
+// exhibit well-defined overflow semantics and rely on the caller to detect
+// if an overflow occured.
+
+template <typename T,
+          typename std::enable_if<std::is_integral<T>::value>::type* = nullptr>
+constexpr T NegateWrapper(T value) {
+  using UnsignedT = typename std::make_unsigned<T>::type;
+  // This will compile to a NEG on Intel, and is normal negation on ARM.
+  return static_cast<T>(UnsignedT(0) - static_cast<UnsignedT>(value));
+}
+
+template <
+    typename T,
+    typename std::enable_if<std::is_floating_point<T>::value>::type* = nullptr>
+constexpr T NegateWrapper(T value) {
+  return -value;
+}
+
+template <typename T,
+          typename std::enable_if<std::is_integral<T>::value>::type* = nullptr>
+constexpr typename std::make_unsigned<T>::type InvertWrapper(T value) {
+  return ~value;
+}
+
+template <typename T,
+          typename std::enable_if<std::is_integral<T>::value>::type* = nullptr>
+constexpr T AbsWrapper(T value) {
+  return static_cast<T>(SafeUnsignedAbs(value));
+}
+
+template <
+    typename T,
+    typename std::enable_if<std::is_floating_point<T>::value>::type* = nullptr>
+constexpr T AbsWrapper(T value) {
+  return value < 0 ? -value : value;
+}
+
+template <template <typename, typename, typename> class M,
+          typename L,
+          typename R>
+struct MathWrapper {
+  using math = M<typename UnderlyingType<L>::type,
+                 typename UnderlyingType<R>::type,
+                 void>;
+  using type = typename math::result_type;
+};
+
+// These variadic templates work out the return types.
+// TODO(jschuh): Rip all this out once we have C++14 non-trailing auto support.
+template <template <typename, typename, typename> class M,
+          typename L,
+          typename R,
+          typename... Args>
+struct ResultType;
+
+template <template <typename, typename, typename> class M,
+          typename L,
+          typename R>
+struct ResultType<M, L, R> {
+  using type = typename MathWrapper<M, L, R>::type;
+};
+
+template <template <typename, typename, typename> class M,
+          typename L,
+          typename R,
+          typename... Args>
+struct ResultType {
+  using type =
+      typename ResultType<M, typename ResultType<M, L, R>::type, Args...>::type;
+};
+
+// The following macros are just boilerplate for the standard arithmetic
+// operator overloads and variadic function templates. A macro isn't the nicest
+// solution, but it beats rewriting these over and over again.
+#define BASE_NUMERIC_ARITHMETIC_VARIADIC(CLASS, CL_ABBR, OP_NAME)       \
+  template <typename L, typename R, typename... Args>                   \
+  constexpr CLASS##Numeric<                                             \
+      typename ResultType<CLASS##OP_NAME##Op, L, R, Args...>::type>     \
+      CL_ABBR##OP_NAME(const L lhs, const R rhs, const Args... args) {  \
+    return CL_ABBR##MathOp<CLASS##OP_NAME##Op, L, R, Args...>(lhs, rhs, \
+                                                              args...); \
+  }
+
+#define BASE_NUMERIC_ARITHMETIC_OPERATORS(CLASS, CL_ABBR, OP_NAME, OP, CMP_OP) \
+  /* Binary arithmetic operator for all CLASS##Numeric operations. */          \
+  template <typename L, typename R,                                            \
+            typename std::enable_if<Is##CLASS##Op<L, R>::value>::type* =       \
+                nullptr>                                                       \
+  constexpr CLASS##Numeric<                                                    \
+      typename MathWrapper<CLASS##OP_NAME##Op, L, R>::type>                    \
+  operator OP(const L lhs, const R rhs) {                                      \
+    return decltype(lhs OP rhs)::template MathOp<CLASS##OP_NAME##Op>(lhs,      \
+                                                                     rhs);     \
+  }                                                                            \
+  /* Assignment arithmetic operator implementation from CLASS##Numeric. */     \
+  template <typename L>                                                        \
+  template <typename R>                                                        \
+  constexpr CLASS##Numeric<L>& CLASS##Numeric<L>::operator CMP_OP(             \
+      const R rhs) {                                                           \
+    return MathOp<CLASS##OP_NAME##Op>(rhs);                                    \
+  }                                                                            \
+  /* Variadic arithmetic functions that return CLASS##Numeric. */              \
+  BASE_NUMERIC_ARITHMETIC_VARIADIC(CLASS, CL_ABBR, OP_NAME)
+
+}  // namespace internal
+}  // namespace base
+
+#endif  // BASE_NUMERICS_SAFE_MATH_SHARED_IMPL_H_

dep/chromium/cgmanifest.json

@@ -0,0 +1,13 @@
+{"Registrations":[ 
+    {
+      "component": { 
+       "type": "git", 
+       "git": { 
+         "repositoryUrl": "https://github.com/chromium/chromium", 
+         "commitHash": "d8710dd959da8e3be56f20af8cc94fbf560fbb6b" 
+         }
+       }
+    }
+ ],
+ "Version": 1
+}

dep/jsoncpp/README.md

@@ -0,0 +1,7 @@
+# jsoncpp
+
+[Amalgamated](https://github.com/open-source-parsers/jsoncpp/wiki/Amalgamated)
+from source commit
+[ddabf50](https://github.com/open-source-parsers/jsoncpp/commit/ddabf50f72cf369bf652a95c4d9fe31a1865a781),
+release 1.8.4.
+

dep/jsoncpp/json/json-forwards.h

@@ -0,0 +1,333 @@
+/// Json-cpp amalgamated forward header (http://jsoncpp.sourceforge.net/).
+/// It is intended to be used with #include "json/json-forwards.h"
+/// This header provides forward declaration for all JsonCpp types.
+
+// //////////////////////////////////////////////////////////////////////
+// Beginning of content of file: LICENSE
+// //////////////////////////////////////////////////////////////////////
+
+/*
+The JsonCpp library's source code, including accompanying documentation, 
+tests and demonstration applications, are licensed under the following
+conditions...
+
+Baptiste Lepilleur and The JsonCpp Authors explicitly disclaim copyright in all 
+jurisdictions which recognize such a disclaimer. In such jurisdictions, 
+this software is released into the Public Domain.
+
+In jurisdictions which do not recognize Public Domain property (e.g. Germany as of
+2010), this software is Copyright (c) 2007-2010 by Baptiste Lepilleur and
+The JsonCpp Authors, and is released under the terms of the MIT License (see below).
+
+In jurisdictions which recognize Public Domain property, the user of this 
+software may choose to accept it either as 1) Public Domain, 2) under the 
+conditions of the MIT License (see below), or 3) under the terms of dual 
+Public Domain/MIT License conditions described here, as they choose.
+
+The MIT License is about as close to Public Domain as a license can get, and is
+described in clear, concise terms at:
+
+   http://en.wikipedia.org/wiki/MIT_License
+   
+The full text of the MIT License follows:
+
+========================================================================
+Copyright (c) 2007-2010 Baptiste Lepilleur and The JsonCpp Authors
+
+Permission is hereby granted, free of charge, to any person
+obtaining a copy of this software and associated documentation
+files (the "Software"), to deal in the Software without
+restriction, including without limitation the rights to use, copy,
+modify, merge, publish, distribute, sublicense, and/or sell copies
+of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be
+included in all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
+========================================================================
+(END LICENSE TEXT)
+
+The MIT license is compatible with both the GPL and commercial
+software, affording one all of the rights of Public Domain with the
+minor nuisance of being required to keep the above copyright notice
+and license text in the source code. Note also that by accepting the
+Public Domain "license" you can re-license your copy using whatever
+license you like.
+
+*/
+
+// //////////////////////////////////////////////////////////////////////
+// End of content of file: LICENSE
+// //////////////////////////////////////////////////////////////////////
+
+
+
+
+
+#ifndef JSON_FORWARD_AMALGAMATED_H_INCLUDED
+# define JSON_FORWARD_AMALGAMATED_H_INCLUDED
+/// If defined, indicates that the source file is amalgamated
+/// to prevent private header inclusion.
+#define JSON_IS_AMALGAMATION
+
+// //////////////////////////////////////////////////////////////////////
+// Beginning of content of file: include/json/config.h
+// //////////////////////////////////////////////////////////////////////
+
+// Copyright 2007-2010 Baptiste Lepilleur and The JsonCpp Authors
+// Distributed under MIT license, or public domain if desired and
+// recognized in your jurisdiction.
+// See file LICENSE for detail or copy at http://jsoncpp.sourceforge.net/LICENSE
+
+#ifndef JSON_CONFIG_H_INCLUDED
+#define JSON_CONFIG_H_INCLUDED
+#include <stddef.h>
+#include <string> //typedef String
+#include <stdint.h> //typedef int64_t, uint64_t
+
+/// If defined, indicates that json library is embedded in CppTL library.
+//# define JSON_IN_CPPTL 1
+
+/// If defined, indicates that json may leverage CppTL library
+//#  define JSON_USE_CPPTL 1
+/// If defined, indicates that cpptl vector based map should be used instead of
+/// std::map
+/// as Value container.
+//#  define JSON_USE_CPPTL_SMALLMAP 1
+
+// If non-zero, the library uses exceptions to report bad input instead of C
+// assertion macros. The default is to use exceptions.
+#ifndef JSON_USE_EXCEPTION
+#define JSON_USE_EXCEPTION 1
+#endif
+
+/// If defined, indicates that the source file is amalgamated
+/// to prevent private header inclusion.
+/// Remarks: it is automatically defined in the generated amalgamated header.
+// #define JSON_IS_AMALGAMATION
+
+#ifdef JSON_IN_CPPTL
+#include <cpptl/config.h>
+#ifndef JSON_USE_CPPTL
+#define JSON_USE_CPPTL 1
+#endif
+#endif
+
+#ifdef JSON_IN_CPPTL
+#define JSON_API CPPTL_API
+#elif defined(JSON_DLL_BUILD)
+#if defined(_MSC_VER) || defined(__MINGW32__)
+#define JSON_API __declspec(dllexport)
+#define JSONCPP_DISABLE_DLL_INTERFACE_WARNING
+#endif // if defined(_MSC_VER)
+#elif defined(JSON_DLL)
+#if defined(_MSC_VER) || defined(__MINGW32__)
+#define JSON_API __declspec(dllimport)
+#define JSONCPP_DISABLE_DLL_INTERFACE_WARNING
+#endif // if defined(_MSC_VER)
+#endif // ifdef JSON_IN_CPPTL
+#if !defined(JSON_API)
+#define JSON_API
+#endif
+
+// If JSON_NO_INT64 is defined, then Json only support C++ "int" type for
+// integer
+// Storages, and 64 bits integer support is disabled.
+// #define JSON_NO_INT64 1
+
+#if defined(_MSC_VER) // MSVC
+#  if _MSC_VER <= 1200 // MSVC 6
+    // Microsoft Visual Studio 6 only support conversion from __int64 to double
+    // (no conversion from unsigned __int64).
+#    define JSON_USE_INT64_DOUBLE_CONVERSION 1
+    // Disable warning 4786 for VS6 caused by STL (identifier was truncated to '255'
+    // characters in the debug information)
+    // All projects I've ever seen with VS6 were using this globally (not bothering
+    // with pragma push/pop).
+#    pragma warning(disable : 4786)
+#  endif // MSVC 6
+
+#  if _MSC_VER >= 1500 // MSVC 2008
+    /// Indicates that the following function is deprecated.
+#    define JSONCPP_DEPRECATED(message) __declspec(deprecated(message))
+#  endif
+
+#endif // defined(_MSC_VER)
+
+// In c++11 the override keyword allows you to explicitly define that a function
+// is intended to override the base-class version.  This makes the code more
+// managable and fixes a set of common hard-to-find bugs.
+#if __cplusplus >= 201103L
+# define JSONCPP_OVERRIDE override
+# define JSONCPP_NOEXCEPT noexcept
+#elif defined(_MSC_VER) && _MSC_VER > 1600 && _MSC_VER < 1900
+# define JSONCPP_OVERRIDE override
+# define JSONCPP_NOEXCEPT throw()
+#elif defined(_MSC_VER) && _MSC_VER >= 1900
+# define JSONCPP_OVERRIDE override
+# define JSONCPP_NOEXCEPT noexcept
+#else
+# define JSONCPP_OVERRIDE
+# define JSONCPP_NOEXCEPT throw()
+#endif
+
+#ifndef JSON_HAS_RVALUE_REFERENCES
+
+#if defined(_MSC_VER) && _MSC_VER >= 1600 // MSVC >= 2010
+#define JSON_HAS_RVALUE_REFERENCES 1
+#endif // MSVC >= 2010
+
+#ifdef __clang__
+#if __has_feature(cxx_rvalue_references)
+#define JSON_HAS_RVALUE_REFERENCES 1
+#endif  // has_feature
+
+#elif defined __GNUC__ // not clang (gcc comes later since clang emulates gcc)
+#if defined(__GXX_EXPERIMENTAL_CXX0X__) || (__cplusplus >= 201103L)
+#define JSON_HAS_RVALUE_REFERENCES 1
+#endif  // GXX_EXPERIMENTAL
+
+#endif // __clang__ || __GNUC__
+
+#endif // not defined JSON_HAS_RVALUE_REFERENCES
+
+#ifndef JSON_HAS_RVALUE_REFERENCES
+#define JSON_HAS_RVALUE_REFERENCES 0
+#endif
+
+#ifdef __clang__
+#  if __has_extension(attribute_deprecated_with_message)
+#    define JSONCPP_DEPRECATED(message)  __attribute__ ((deprecated(message)))
+#  endif
+#elif defined __GNUC__ // not clang (gcc comes later since clang emulates gcc)
+#  if (__GNUC__ > 4 || (__GNUC__ == 4 && __GNUC_MINOR__ >= 5))
+#    define JSONCPP_DEPRECATED(message)  __attribute__ ((deprecated(message)))
+#  elif (__GNUC__ > 3 || (__GNUC__ == 3 && __GNUC_MINOR__ >= 1))
+#    define JSONCPP_DEPRECATED(message)  __attribute__((__deprecated__))
+#  endif  // GNUC version
+#endif // __clang__ || __GNUC__
+
+#if !defined(JSONCPP_DEPRECATED)
+#define JSONCPP_DEPRECATED(message)
+#endif // if !defined(JSONCPP_DEPRECATED)
+
+#if __GNUC__ >= 6
+#  define JSON_USE_INT64_DOUBLE_CONVERSION 1
+#endif
+
+#if !defined(JSON_IS_AMALGAMATION)
+
+# include "version.h"
+
+# if JSONCPP_USING_SECURE_MEMORY
+#  include "allocator.h" //typedef Allocator
+# endif
+
+#endif // if !defined(JSON_IS_AMALGAMATION)
+
+namespace Json {
+typedef int Int;
+typedef unsigned int UInt;
+#if defined(JSON_NO_INT64)
+typedef int LargestInt;
+typedef unsigned int LargestUInt;
+#undef JSON_HAS_INT64
+#else                 // if defined(JSON_NO_INT64)
+// For Microsoft Visual use specific types as long long is not supported
+#if defined(_MSC_VER) // Microsoft Visual Studio
+typedef __int64 Int64;
+typedef unsigned __int64 UInt64;
+#else                 // if defined(_MSC_VER) // Other platforms, use long long
+typedef int64_t Int64;
+typedef uint64_t UInt64;
+#endif // if defined(_MSC_VER)
+typedef Int64 LargestInt;
+typedef UInt64 LargestUInt;
+#define JSON_HAS_INT64
+#endif // if defined(JSON_NO_INT64)
+#if JSONCPP_USING_SECURE_MEMORY
+#define JSONCPP_STRING        std::basic_string<char, std::char_traits<char>, Json::SecureAllocator<char> >
+#define JSONCPP_OSTRINGSTREAM std::basic_ostringstream<char, std::char_traits<char>, Json::SecureAllocator<char> >
+#define JSONCPP_OSTREAM       std::basic_ostream<char, std::char_traits<char>>
+#define JSONCPP_ISTRINGSTREAM std::basic_istringstream<char, std::char_traits<char>, Json::SecureAllocator<char> >
+#define JSONCPP_ISTREAM       std::istream
+#else
+#define JSONCPP_STRING        std::string
+#define JSONCPP_OSTRINGSTREAM std::ostringstream
+#define JSONCPP_OSTREAM       std::ostream
+#define JSONCPP_ISTRINGSTREAM std::istringstream
+#define JSONCPP_ISTREAM       std::istream
+#endif // if JSONCPP_USING_SECURE_MEMORY
+} // end namespace Json
+
+#endif // JSON_CONFIG_H_INCLUDED
+
+// //////////////////////////////////////////////////////////////////////
+// End of content of file: include/json/config.h
+// //////////////////////////////////////////////////////////////////////
+
+
+
+
+
+
+// //////////////////////////////////////////////////////////////////////
+// Beginning of content of file: include/json/forwards.h
+// //////////////////////////////////////////////////////////////////////
+
+// Copyright 2007-2010 Baptiste Lepilleur and The JsonCpp Authors
+// Distributed under MIT license, or public domain if desired and
+// recognized in your jurisdiction.
+// See file LICENSE for detail or copy at http://jsoncpp.sourceforge.net/LICENSE
+
+#ifndef JSON_FORWARDS_H_INCLUDED
+#define JSON_FORWARDS_H_INCLUDED
+
+#if !defined(JSON_IS_AMALGAMATION)
+#include "config.h"
+#endif // if !defined(JSON_IS_AMALGAMATION)
+
+namespace Json {
+
+// writer.h
+class FastWriter;
+class StyledWriter;
+
+// reader.h
+class Reader;
+
+// features.h
+class Features;
+
+// value.h
+typedef unsigned int ArrayIndex;
+class StaticString;
+class Path;
+class PathArgument;
+class Value;
+class ValueIteratorBase;
+class ValueIterator;
+class ValueConstIterator;
+
+} // namespace Json
+
+#endif // JSON_FORWARDS_H_INCLUDED
+
+// //////////////////////////////////////////////////////////////////////
+// End of content of file: include/json/forwards.h
+// //////////////////////////////////////////////////////////////////////
+
+
+
+
+
+#endif //ifndef JSON_FORWARD_AMALGAMATED_H_INCLUDED

dep/packages/README.md

@@ -1,7 +0,0 @@
-These packages are redistributed inside this folder because they are not yet available on a public NuGet feed.
-
-## Microsoft.UI.XAML
-This package is a custom development build fork to help us light up tab support. It will eventually go onto the same public feed as the existing `Microsoft.UI.XAML` package that's currently available on NuGet.org
-
-## TAEF.Redist.WLK
-This package is vetted for public redistribution and release, but the TAEF team hasn't set up a public feed to consume it yet. If/when they do, we'll move to that.

doc/AddASetting.md

@@ -13,7 +13,7 @@
 2. Add matching fields to Settings.hpp
     - add getters, setters, the whole drill.
 
-3. Add to the propsheet.
+3. Add to the propsheet
     - We need to add it to *reading and writing* the registry from the propsheet, and *reading* the link from the propsheet. Yes, that's weird, but the propsheet is smart enough to re-use ShortcutSerialization::s_SetLinkValues, but not smart enough to do the same with RegistrySerialization.
     - `src/propsheet/registry.cpp`
         -  `propsheet/registry.cpp@InitRegistryValues` should initialize the default value for the property.
@@ -28,7 +28,7 @@
 6. Add the setting to `Menu::s_GetConsoleState`, and `Menu::s_PropertiesUpdate`
 Now, your new setting should be stored just like all the other properties.
 
-7. Update the feature test properties to get add the setting as well.
+7. Update the feature test properties to get add the setting as well
     - `ft_uia/Common/NativeMethods.cs@WinConP`:
         - `Wtypes.PROPERTYKEY PKEY_Console_`
         - `NT_CONSOLE_PROPS`
@@ -36,5 +36,5 @@ Now, your new setting should be stored just like all the other properties.
 8. Add the default value for the setting to `win32k-settings.man`
     - If the setting shouldn't default to 0 or `nullptr`, then you'll need to set the default value of the setting in `win32k-settings.man`.
 
-9. Update `Settings::InitFromStateInfo` and `Settings::CreateConsoleStateInfo` to get/set the value in a CONSOLE_STATE_INFO appropriately.
+9. Update `Settings::InitFromStateInfo` and `Settings::CreateConsoleStateInfo` to get/set the value in a CONSOLE_STATE_INFO appropriately
 

doc/ConsoleCtrlEvent.md

@@ -2,7 +2,7 @@
 
 ## Generation
 
-conhost requests that user32 inject a thread into the attached console application.
+conhost requests that user32 injects a thread into the attached console application.
 See ntuser's exitwin.c for `CreateCtrlThread`.
 
 ## Timeouts

doc/ConsoleHostSettings.md

@@ -1,6 +1,6 @@
 # Understanding Console Host Settings
 
-Settings in the Windows Console Host can be a bit tricky to understand. This is mostly because the settings system evolved over the course of decades. Before we dig into the details of how settings are persisted, it's probably worth a quick look at what these settings are.
+Settings in the Windows Console Host can be a bit tricky to understand. This is mostly because the settings system evolved over the course of decades. Before we dig into the details of how settings are persisted, it's probably worth taking a quick look at what these settings are.
 
 ## Settings Description
 
@@ -8,9 +8,9 @@ Settings in the Windows Console Host can be a bit tricky to understand. This is
 |---------------------------|-----------------------|--------------------------------------|
 |`FontSize`                 |Coordinate (REG_DWORD) |Size of font in pixels                |
 |`FontFamily`               |REG_DWORD              |GDI Font family                       |
-|`ScreenBufferSize`         |Coordinate (REG_DWORD) |Size of the screen buffer in WxH characters |
+|`ScreenBufferSize`         |Coordinate (REG_DWORD) |Size of the screen buffer in WxH characters\*\* |
 |`CursorSize`               |REG_DWORD              |Cursor height as percentage of a single character |
-|`WindowSize`               |Coordinate (REG_DWORD) |Initial size of the window in WxH characters |
+|`WindowSize`               |Coordinate (REG_DWORD) |Initial size of the window in WxH characters\*\* |
 |`WindowPosition`           |Coordinate (REG_DWORD) |Initial position of the window in WxH pixels (if not set, use auto-positioning) |
 |`WindowAlpha`              |REG_DWORD              |Opacity of the window (valid range: 0x4D-0xFF) |
 |`ScreenColors`             |REG_DWORD              |Default foreground and background colors |
@@ -39,6 +39,10 @@ Settings in the Windows Console Host can be a bit tricky to understand. This is
 
 *: Only applies to the improved version of the Windows Console Host
 
+**: WxH stands for Width by Height, it's the fact that things like a Window size
+store the Width and Height values in the high and low word in the registry's
+double word values.
+
 ## The Settings Hierarchy
 
 Settings are persisted to a variety of locations depending on how they are modified and how the Windows Console Host was invoked:

doc/Debugging.md

@@ -6,5 +6,32 @@ This file contains notes about debugging various items in the repository.
 
 If you want to debug code in the Cascadia package via Visual Studio, your breakpoints will not be hit by default. A tweak is required to the *CascadiaPackage* project in order to enable this.
 
-1. Right-click on *CascadiaPackage* in Solution Explorer and select Properties
-2. Change the *Application process* type from *Mixed (Managed and Native)* to *Native Only*.
+1. Right-click on *CascadiaPackage* in Solution Explorer and select Properties.
+2. Change the *Application process* type from *Mixed (Managed and Native)* to *Native Only*.
+
+## Popping into the Debugger from Running Code
+
+Sometimes you will encounter a scenario where you need to break into the console or terminal code under the debugger but you cannot, for whatever reason, do so by launching it from the beginning under the debugger. This can be especially useful for debugging tests with TAEF which usually launch through several child processes and modules before hitting your code. 
+
+To accomplish this, add a `DebugBreak()` statement somewhere in the code and ensure you have a Post-Mortem debugger set.
+
+**NOTE:** `conhost.exe` already has a provision for a conditional `DebugBreak()` very early in the startup code if it was built in debug mode. Set `HKCU\Console` with `DebugLaunch` as a `REG_DWORD` with the value of `1`.
+
+### Setting Visual Studio as Post Mortem Debugger
+
+Go to `Tools > Options` and then make sure that `Native` is checked as the `Just-In-Time Debugging` provider. (Checking the box, if it is not checked, will require that Visual Studio is launched as Administrator.)
+
+![image](https://user-images.githubusercontent.com/18221333/72091481-1b870100-32c5-11ea-8235-cebb9a383c32.png)
+
+Then when you run something with `DebugBreak()` in it, you will see this: 
+![image](https://user-images.githubusercontent.com/18221333/72091543-42453780-32c5-11ea-8b4b-83a362eb73df.png)
+
+The top ones will be new instances of the Visual Studios installed on your system. The bottom ones will be the running instances of Visual Studio. You can see in the image that one is open already. If you choose the bottom one, VS will attach straight up as if you F5'd from the solution at the point from the `DebugBreak()`. Step up to get out of the break and back into the code.
+
+### Setting WinDBG as Post Mortem Debugger
+
+From an elevated context (a command prompt or whatnot...), run `windbg /I`. This will install the debugger as Post Mortem.
+
+Then run the thing and it will pop straight into a new WinDBG session. Step up to get out of the break and back into the code.
+
+**Caveat:** If you are on an x64 system, you may need to do `windbg /I` with both the x64 and x86 versions of the debugger to catch all circumstances (like if you're trying to run x86 code.) 

doc/EXCEPTIONS.md

@@ -9,12 +9,12 @@ sometimes it's significantly simpler to use them. Given that, we have a set of r
 exception use.
 
 ## Rules
-1. **DO NOT** allow exceptions to leak out of new code into new code
-1. **DO** use NTSTATUS or HRESULT as return values as appropriate
-1. **DO** Encapsulate all exception behaviors within implementing classes
+1. **DO NOT** allow exceptions to leak out of new code into old code
+1. **DO** use `NTSTATUS` or `HRESULT` as return values as appropriate (`HRESULT` is preferred)
+1. **DO** encapsulate all exception behaviors within implementing classes
 1. **DO NOT** introduce modern exception throwing code into old code. Instead, refactor as needed to allow encapsulation or
    use non-exception based code
-1. **DO** use WIL as an alternative for non-throwing modern facilities (e.g. wil::unique_ptr<>)
+1. **DO** use WIL as an alternative for non-throwing modern facilities (e.g. `wil::unique_ptr<>`)
 
 ## Examples
 
@@ -33,4 +33,4 @@ exception use.
 
 ### Using WIL for non-throwing modern facilities
 
-### 
+TODO

doc/Niksa.md

@@ -0,0 +1,181 @@
+# Niksa's explanations
+
+Sometimes @miniksa will write a big, long explanatory comment in an issue thread that turns out to be a decent bit of reference material.
+This document serves as a storage point for those posts.
+
+- [Why do we avoid changing CMD.exe?](#cmd)
+- [Why is typing-to-screen performance better than every other app?](#screenPerf)
+- [How are the Windows graphics/messaging stack assembled?](#gfxMsgStack)
+- [Output Processing between "Far East" and "Western"](#fesb)
+- [Why do we not backport things?](#backport)
+- [Why can't we have mixed elevated and non-elevated tabs in the Terminal?](#elevation)
+
+## <a name="cmd"></a>Why do we avoid changing CMD.exe?
+`setlocal` doesn't behave the same way as an environment variable. It's a thing that would have to be put in at the top of the batch script that is `somefile.cmd` as one of its first commands to adjust the way that one specific batch file is processed by the `cmd.exe` engine. That's probably not suitable for your needs, but that's the way we have to go.
+
+I don't think anyone is disagreeing with you, @mikemaccana, that this would be a five minute development change to read that environment variable and change the behavior of `cmd.exe`. It absolutely would be a tiny development time. 
+
+It's just that from our experience, we know there's going to be a 3-24 month bug tail here where we get massive investigation callbacks by some billion dollar enterprise customer who for whatever reason was already using the environment variable we pick for another purpose. Their script that they give their rank-and-file folks will tell them to press Ctrl+C at some point in the batch script to do whatever happens, it will do something different, those people will notice the script doesn't match the computer anymore. They will then halt the production line and tell their supervisor. The supervisor tells some director. Their director comes screaming at their Microsoft enterprise support contract person that we've introduced a change to the OS that is costing them millions if not billions of dollars in shipments per month. Our directors at Microsoft then come bashing down our doors angry with us and make us fix it ASAP or revert it, we don't get to go home at 5pm to our families or friends because we're fixing it, we get stressed the heck out, we have to spin up servicing potentially for already shipped operating systems which is expensive and headache-causing...etc.
+
+We can see this story coming a million miles away because it has happened before with other 'tiny' change we've been asked to make to `cmd.exe` in the past few years.
+
+I would just ask you to understand that `cmd.exe` is very, very much in a maintenance mode and I just want to set expectations here. We maintain it, yes. We have a renewed interest in command-line development, yes. But our focuses are revolving around improving the terminal and platform itself and bringing modern, supported shells to be the best they can be on Windows. Paul will put this on the backlog of things that people want in `cmd.exe`, yes. But it will sink to the bottom of the backlog because changing `cmd.exe` is our worst nightmare as its compatibility story is among the heaviest of any piece of the operating system.
+
+I would highly recommend that Gulp convert to using PowerShell scripts and that if such an issue exists with PowerShell, that we get their modern, supported, and better-engineered platform to support the scenario. I don't want you to sit around waiting for `cmd.exe` to change this because it's really not going to happen faster than that script could be converted to `ps1` and it fixed in PowerShell Core (if that's even a problem in that world.)
+
+Original Source: https://github.com/microsoft/terminal/issues/217#issuecomment-404240443
+
+## <a name="screenPerf"></a>Why is typing-to-screen performance better than every other app?
+
+I really do not mind when someone comes by and decides to tell us that we're doing a good job at something. We hear so many complaints every day that a post like this is a breath of fresh air. Thanks for your thanks!
+
+Also, I'm happy to discuss this with you until you're utterly sick of reading it. Please ask any follow-ons you want. I thrive on blathering about my work. :P
+
+If I had to take an educated guess as to what is making us faster than pretty much any other application on Windows at putting your text on the screen... I would say it is because that is literally our only job! Also probably because we are using darn near the oldest and lowest level APIs that Windows has to accomplish this work.
+
+Pretty much everything else you've listed has some sort of layer or framework involved, or many, many layers and frameworks, when you start talking about Electron and JavaScript. We don't. 
+
+We have one bare, super un-special window with no additional controls attached to it. We get our keys fed into us from just barely above the kernel given that we're processing them from window messages and not from some sort of eventing framework common to pretty much any other more complicated UI framework than ours (WPF, WinForms, UWP, Electron). And we dump our text straight onto the window surface using GDI's [PolyTextOut](https://docs.microsoft.com/en-us/windows/desktop/api/wingdi/nf-wingdi-polytextoutw) with no frills. 
+
+Even `notepad.exe` has multiple controls on its window at the very least and is probably (I haven't looked) using some sort of library framework in the edit control to figure out its text layout (which probably is using another library framework for internationalization support...)
+
+Of course this also means that we have trade offs. We don't support fully international text like pretty much every other application will. RTL? No go zone right now. Surrogate pairs and emoji? We're getting there but not there yet. Indic scripts? Nope. 
+
+Why are we like this? For one, `conhost.exe` is old as dirt. It has to use the bare metal bottom layer of everything because it was created before most of those other frameworks were created. And also it maintains as low/bottom level as possible because it is pretty much the first thing that one needs to bring up when bringing up a new operating system edition or device before you have all the nice things like frameworks or what those frameworks require to operate. Also it's written in C/C++ which is about as low and bare metal as we can get.
+
+Will this UI enhancement come to other apps on Windows? Almost certainly not. They have too much going on which is both a good and a bad thing. I'm jealous of their ability to just call one method and layout text in an uncomplicated manner in any language without manually calculating pixels or caring about what styles apply to their font. But my manual pixel calculations, dirty region math, scroll region madness, and more makes it so we go faster than them. I'm also jealous that when someone says "hey can you add a status bar to the bottom of your window" that they can pretty much click and drag that into place with their UI Framework and it will just work where as for us, it's been a backlog item forever and gives me heartburn to think about implementing. 
+
+Will we try to keep it from regressing? Yes! Right now it's sort of a manual process. We identify that something is getting slow and then we go haul out [WPR](https://docs.microsoft.com/en-us/windows-hardware/test/wpt/windows-performance-recorder) and start taking traces. We stare down the hot paths and try to reason out what is going on and then improve them. For instance, in the last cycle or two, we focused on heap allocations as a major area where we could improve our end-to-end performance, changing a ton of our code to use stack-constructed iterator-like facades over the underlying request buffer instead of translating and allocating it into a new heap space for each level of processing. 
+
+As an aside, @bitcrazed wants us to automate performance tests in some conhost specific way, but I haven't quite figured out a controlled environment to do this in yet. The Windows Engineering System runs performance tests each night that give us a coarse grained way of knowing if we messed something up for the whole operating system, and they technically offer a fine grained way for us to insert our own performance tests... but I just haven't got around to that yet. If you have an idea for a way for us to do this in an automated fashion, I'm all ears.
+
+If there's anything else you'd like to know, let me know. I could go on all day. I deleted like 15 tangents from this reply before posting it.... 
+
+Original Source: https://github.com/microsoft/terminal/issues/327#issuecomment-447391705
+
+## <a name="gfxMsgStack"></a>How are the Windows graphics/messaging stack assembled?
+
+@stakx, I am referring to USER32 and GDI32. 
+
+I'll give you a cursory overview of what I know off the top of my head without spending hours confirming the details. As such, some of this is subject to handwaving and could be mildly incorrect but is probably in the right direction. Consider every statement to be my personal knowledge on how the world works and subject to opinion or error.
+
+For the graphics part of the pipeline (GDI32), the user-mode portions of GDI are pretty far down. The app calls GDI32, some work is done in that DLL on the user-mode side, then a kernel call jumps over to the kernel and drawing occurs. 
+
+The portion that you're thinking of regarding "silently converted to sit on top of other stuff" is probably that once we hit the kernel calls, a bunch of the kernel GDI stuff tends to be re-platformed on top of the same stuff as DirectX when it is actually handled by the NVIDIA/AMD/Intel/etc. graphics driver and the GPU at the bottom of the stack. I think this happened with the graphics driver re-architecture that came as a part of WDDM for Windows Vista. There's a document out there somewhere about what calls are still really fast in GDI and which are slower as a result of the re-platforming. Last time I found that document and checked, we were using the fast ones.
+
+On top of GDI, I believe there are things like Common Controls or comctl32.dll which provided folks reusable sets of buttons and elements to make their UIs before we had nicer declarative frameworks. We don't use those in the console really (except in the property sheet off the right click menu).
+
+As for DirectWrite and D2D and D3D and DXGI themselves, they're a separate set of commands and paths that are completely off to the side from GDI at all both in user and kernel mode. They're not really related other than that there's some interoperability provisions between the two. Most of our other UI frameworks tend to be built on top of the DirectX stack though. XAML is for sure. I think WPF is. Not sure about WinForms. And I believe the composition stack and the window manager are using DirectX as well.
+
+As for the input/interaction part of the pipeline (USER32), I tend to find most other newer things (at least for desktop PCs) are built on top of what is already there. USER32's major concept is windows and window handles and everything is sent to a window handle. As long as you're on a desktop machine (or a laptop or whatever... I mean a classic-style Windows-powered machine), there's a window handle involved and messages floating around and that means we're talking USER32. 
+
+The window message queue is just a straight up FIFO (more or less) of whatever input has occurred relevant to that window while it's in the foreground + whatever has been sent to the window by other components in the system. 
+
+The newer technologies and the frameworks like XAML and WPF and WinForms tend to receive the messages from the window message queue one way or another and process them and turn them into event callbacks to various objects that they've provisioned within their world. 
+
+However, the newer technologies that also work on other non-desktop platforms like XAML tend to have the ability to process stuff off of a completely different non-USER32 stack as well. There's a separate parallel stack to USER32 with all of our new innovations and realizations on how input and interaction should occur that doesn't exactly deal with classic messaging queues and window handles the same way. This is the whole Core* family of things like CoreWindow and CoreMessaging. They also have a different concept of "what is a user" that isn't so centric around your butt in rolling chair in front of a screen with a keyboard and mouse on the desk.
+
+Now, if you're on XAML or one of the other Frameworks... all this intricacy is handled for you. XAML figures out how to draw on DirectX for you and negotiates with the compositor and window manager for cool effects on your behalf. It figures out whether to get your input events from USER32 or Core* or whatever transparently depending on your platform and the input stacks can handle pen, touch, keyboard, mouse, and so on in a unified manner. It has provisions inside it embedded to do all the sorts of globalization, accessibility, input interaction, etc. stuff that make your life easy. But you could choose to go directly to the low-level and handle it yourself or skip handling what you don't care about.
+
+The trick is that GDI32 and USER32 were designed for a limited world with a limited set of commands. Desktop PCs were the only thing that existed, single user at the keyboard and mouse, simple graphics output to a VGA monitor. So using them directly at the "low level" like conhost does is pretty easy. The new platforms could be used at the "low level" but they're orders of magnitude more complicated because they now account for everything that has happened with personal computing in 20+ years like different form factors, multiple active users, multiple graphics adapters, and on and on and on and on. So you tend to use a framework when using the new stuff so your head doesn't explode. They handle it for you, but they handle more than they ever did before so they're slower to some degree.
+
+So are GDI32 and USER32 "lower" than the new stuff? Sort of. 
+Can you get that low with the newer stuff? Mostly yes, but you probably shouldn't and don't want to.
+Does new live on top of old or is old replatformed on the new? Sometimes and/or partially.
+Basically... it's like the answer to anything software... "it's an unmitigated disaster and if we all stepped back a moment, we should be astounded that it works at all." :P
+
+Anyway, that's enough ramble for one morning. Hopefully that somewhat answered your questions and gave you a bit more insight.
+
+Original Source: https://github.com/microsoft/terminal/issues/327#issuecomment-447926388
+
+## <a name="fesb"></a>Output Processing between "Far East" and "Western"
+
+> 
+> ```
+> if (WI_IsFlagSet(CharType, C1_CNTRL))
+> ```
+
+In short, this is probably fine to fix.
+
+However, I would personally feed a few characters through `WriteCharsLegacy` under the debugger and assert that your theory is correct first (that multiple flags coming back are what the problem is) before making the change.
+
+I am mildly terrified, less than Dustin, because it is freaking `WriteCharsLegacy` which is the spawn of hell and I fear some sort of regression in it.
+
+In long, why is it fine to fix?
+
+For reference, this particular segment of code https://github.com/microsoft/terminal/blob/9b92986b49bed8cc41fde4d6ef080921c41e6d9e/src/host/_stream.cpp#L514-L539 appears to only be used when the codepoint is < 0x20 or == 0x7F https://github.com/microsoft/terminal/blob/9b92986b49bed8cc41fde4d6ef080921c41e6d9e/src/host/_stream.cpp#L408 and ENABLE_PROCESSED_OUTPUT is off. https://github.com/microsoft/terminal/blob/9b92986b49bed8cc41fde4d6ef080921c41e6d9e/src/host/_stream.cpp#L320
+
+I looked back at the console v1 code and this particular section had a divergence for "Western" countries and "Far East" countries (a geopolitically-charged term, but what it was, nonetheless.)
+
+For "Western" countries, we would unconditionally run all the characters through `MultiByteToWideChar` with `MB_USEGLYPHCHARS` without the `C1_CNTRL` test and move the result into the buffer.
+
+For "Eastern" countries, we did the `C1_CNTRL` test and then if true, we would run through `MultiByteToWideChar` with `MB_USEGLYPHCHARS`. Otherwise, we would just move the original character into the buffer and call it a day.
+
+Note in both of these, there is a little bit of indirection before `MultiByteToWideChar` is called through some other helper methods like `ConvertOutputToUnicode`, but that's the effective conversion point, as far as I can tell. And that's where the control characters would turn into acceptable low ASCII symbols.
+
+When we took over the console codebase, this variation between "Western" and "Eastern" countries was especially painful because `conhost.exe` would choose which one it was in based on the `Codepage for Non-Unicode Applications` set in the Control Panel's Regional > Administrative panel and it could only be changed with a reboot. It wouldn't even change properly when you `chcp` to a different codepage. Heck, `chcp` would deny you from switching into many codepages. There was a block in place to prevent going to an "Eastern" codepage if you booted up in a "Western" codepage. There was also a block preventing you from going between "Eastern" codepages, if I recall correctly.
+
+In modernizing, I decided a few things:
+1. What's good for the "Far East" should be good for the rest of the world. CJK languages that encompassed the "Far East" code have to be able to handle "Western" text as well even if the reverse wasn't true. 
+2. We need to scrub all usages of "Far East" from the code. Someone already started that and replaced them with "East Asia" except then they left behind the shorthand of "FE" prefixing dozens of functions which made it hard to follow the code. It took us months to realize "FE" and "East Asia" were the same thing.
+3. It's obnoxious that the way this was handled was to literally double-define every output function in the code base to have two definitions, compile them both into the conhost, then choose to run down the SB_ versions or the FE_ versions depending on the startup Non-Unicode codepage. It was a massive pile of complex pre-compilation `#ifdef` and `#else`s that would sometimes surround individual lines in the function bodies. Gross.
+4. The fact that the FE_ versions of the functions were way slower than the SB_ ones was unacceptable even for the same output of Latin-character text.
+5. Anyone should be free to switch between any codepage they want at any time and restricting it based on a value from OS startup or region/locale is not acceptable in the modern world.
+6. I concluded by all of the above that I was going to tank/delete/remove the SB_ versions of everything and force the entire world to use the FE_ versions as truth. I would fix the FE_ versions to handle everything correctly, I would fix the performance characteristics of the FE_ versions so they were only slower when things were legitimately more complicated and never otherwise, I would banish all usage of "Far East", "East Asia", "FE_", and "SB_" from the codebase, and codepages would be freely switchable.
+7. Oh. Also, the conhost used to rewrite its entire backing buffer into whatever your current codepage was whenever you switched codepages. I changed that to always hold it as UTF-16.
+
+Now, after that backstory. This is where the problem comes in. It looks like the code you're pointing to that didn't check flags and instead checked direct equality... is the way that it was ALWAYS done for the "Eastern" copy of the code. So it was ALWAYS broken for the "Eastern" codepages and country variants of the OS.
+
+I don't know why the "Eastern" copy was checking `C1_CNTRL` at all in the first place. There is no documentation. I presume it has to do with Shift-JIS or GB-2312 or Unified Hangul or something having a conflict < 0x20 || == 0x7F. Or alternatively, it's because someone wrote the code naively thinking it was a good idea in a hurry and never tested it. Very possible and even probable.
+
+Presuming CJK codepages have no conflict in this range for their DBCS codepages... we could probably remove the check with `GetStringTypeW` entirely and always run everything through `ConvertOutputToUnicode`. More risky than just the flag test change... but theoretically an option as well.
+
+
+Original Source: https://github.com/microsoft/terminal/issues/166#issuecomment-510953359
+
+## <a name="backport"></a>Why do we not backport things?
+
+Someone has to prove that this is costing millions to billions of dollars of lost productivity or revenue to outweigh the risks of shipping the fix to hundreds of millions of Windows machines and potentially breaking something. 
+
+Our team generally finds it pretty hard to prove that against the developer audience given that they're only a small portion of the total installed market of Windows machines. 
+
+Our only backport successes really come from corporations with massive addressable market (like OEMs shipping PCs) who complain that this is fouling up their manufacturing line (or something of that ilk). Otherwise, our management typically says that the risks don't outweigh the benefits.
+
+It's also costly in terms of time, effort, and testing for us to validate a modification to a released OS. We have a mindbogglingly massive amount of automated machinery dedicated to processing and validating the things that we check in while developing the current OS builds. But it's a special costly ask to spin up some to all of those activities to validate backported fixes. We do it all the time for Patch Tuesday, but in those patches, they only pass through the minimum number of fixes required to maximize the restoration of productivity/security/revenue/etc. because every additional fix adds additional complexity and additional risk. 
+
+So from our little team working hard to make developers happy, we virtually never make the cut for servicing. We're sorry, but we hope you can understand. It's just the reality of the situation to say "nope" when people ask for a backport. In our team's ideal world, you would all be running the latest console bits everywhere everytime we make a change. But that's just not how it is today.
+
+Original Source: https://github.com/microsoft/terminal/issues/279#issuecomment-439179675
+
+## <a name="elevation"></a>Why can't we have mixed elevated and non-elevated tabs in the Terminal?
+
+_guest speaker @DHowett-MSFT_
+
+[1] It is trivial when you are _hosting traditional windows_ with traditional window handles. That works very well in the conemu case, or in the tabbed shell case, where you can take over a window in an elevated session and re-parent it under a window in a non-elevated session.
+
+When you do that, there's a few security features that I'll touch on in [2]. Because of those, you can parent it but you can't really force it to do anything.
+
+There's a problem, though. The Terminal isn't architected as a collection of re-parentable windows. For example, it's not running a console host and moving its window into a tab. It was designed to support a "connection" -- something that can read and write text. It's a lower-level primitive than a window. We realized the error of our ways and decided that the UNIX model was right the entire time, and pipes and text and streams are _where it's at._
+
+Given that we're using Xaml islands to host a modern UI and stitching a DirectX surface into it, we're far beyond the world of standard window handles anyway. Xaml islands are fully composed into a single HWND, much like Chrome and Firefox and the gamut of DirectX/OpenGL/SDL games. We don't **have** components that can be run in one process (elevated) and hosted in another (non-elevated) that aren't the aforementioned "connections".
+
+Now, the obvious followup question is _"why can't you have one elevated connection in a tab next to a non-elevated connection?"_ This is where @sba923 should pick up reading (:smile:). I'm probably going to cover some things that you (@robomac) know already.
+
+[2] When you have two windows on the same desktop in the same window station, they can communicate with eachother. I can use `SendKeys` easily through `WScript.Shell` to send keyboard input to any window that the shell can see.
+
+Running a process elevated _severs_ that connection. The shell can't see the elevated window. No other program at the same integrity level as the shell can see the elevated window. Even if it has its window handle, it can't really interact with it. This is also why you can't drag/drop from explorer into notepad if notepad is running elevated. Only another elevated process can interact with another elevated window.
+
+That "security" feature (call it what you like, it was probably intended to be a security feature at one point) only exists for a few session-global object types. Windows are one of them. Pipes aren't really one of them.
+
+Because of that, it's trivial to break that security. Take the terminal as an example of that. If we start an elevated connection and host it in a _non-elevated_ window, we've suddenly created a conduit through that security boundary. The elevated thing on the other end isn't a window, it's just a text-mode application. It immediately does the bidding of the non-elevated host.
+
+Anybody that can _control_ the non-elevated host (like `WScript.Shell::SendKeys`) _also_ gets an instant conduit through the elevation boundary. Suddenly, any medium integrity application on your system can control a high-integrity process. This could be your browser, or the bitcoin miner that got installed with the `left-pad` package from NPM, or really any number of things.
+
+It's a small risk, but it _is_ a risk.
+
+---
+
+Other platforms have accepted that risk in preference for user convenience. They aren't wrong to do so, but I think Microsoft gets less of a "pass" on things like "accepting risk for user convenience". Windows 9x was an unmitigated security disaster, and limited user accounts and elevation prompts and kernel-level security for window management were the answer to those things. They're not locks to be loosened lightly.
+
+Original Source: https://github.com/microsoft/terminal/issues/632#issuecomment-519375707
+

doc/ORGANIZATION.md

@@ -4,7 +4,7 @@
 
 - **Follow the pattern of what you already see in the code**
 - Try to package new ideas/components into libraries that have nicely defined interfaces
-- Package new ideas into classes or refactor existing ideas into a class as you extend. 
+- Package new ideas into classes or refactor existing ideas into a class as you extend
 - Each project should have a Unit test in a ut_ folder in its subdirectory (like `ut_host`)
 - Functional tests should be in ft_ subdirectories (like `ft_api`)
 - Build scripts are generally in subdirectories with their type of output (like `/dll` or `/exe`)
@@ -24,6 +24,16 @@
 * `/ipch` – not checked in is where intellisense data will be generated if you use Visual Studio 2015
 * `/obj` – not checked in is where objects will be generated by the MSBuild system
 * `/src` – This is the fun one. In the root is common build system data.
+	* `/src/cascadia` - This directory contains all the code specific to the Windows Terminal
+		* `/src/cascadia/TerminalConnection` - This DLL is responsible for the various different ways a terminal instance can communicate with different terminal backends. Examples include the `ConptyConnection` (for communicating with Windows Console processes), or the `AzureCloudShellConnection` for communicating with Azure.
+		* `/src/cascadia/TerminalSettings` - This is the DLL responsible for abstracting the settings for both the TerminalCore and the TerminalControl. This provides consumers of the TerminalControl a common interface for supplying settings to the Terminal.
+		* `/src/cascadia/TerminalCore` - This LIB is responsible for the core implementation of a terminal instance. This defines one important class `Terminal` which is a complete terminal instance, with buffer, colors table, VT parsing, input handling, etc. It does _not_ prescribe any sort of UI implementation - it should be connected to code that can handle rendering its contents, and provide input to it.
+		* `/src/cascadia/TerminalControl` - This DLL provides the UWP-XAML implementation of a `TermControl`, which can be embedded within an application to provide a terminal instance within the application. It contains a DX renderer for drawing text to the screen, and translates input to send to the core Terminal. It also receives settings to apply to both itself and the core Terminal.
+		* `/src/cascadia/TerminalApp` - This DLL represents the implementation of the Windows Terminal application. This includes parsing settings, hosting tabs & panes with Terminals in them, and displaying other UI elements. This DLL is almost entirely UWP-like code, and shouldn't be doing any Win32-like UI work.
+		* `/src/cascadia/WindowsTerminal` - This EXE provides Win32 hosting for the TerminalApp. It will set up XAML islands, and is responsible for drawing the window, either as a standard window or with content in the titlebar (non-client area).
+		* `/src/cascadia/CascadiaPackage` - This is a project for packaging the Windows Terminal and its dependencies into an .appx/.msix for deploying to the machine.
+		* `/src/cascadia/PublicTerminalCore` - This is a DLL wrapper for the TerminalCore and Renderer, similar to `TermControl`, which exposes some exported functions that so the Terminal can be used from C#.
+		* `/src/cascadia/WpfTerminalControl` - A DLL implementing a WPF version of the Terminal Control.
 	* `/src/host` – The meat of the windows console host. This includes buffer, input, output, windowing, server management, clipboard, and most interactions with the console host window that aren’t stated anywhere else. We’re trying to pull things out that are reusable into other libraries, but it’s a work in progress
 		* `/src/host/lib` – Builds the reusable LIB copy of the host
 		* `/src/host/dll` – Packages LIB into conhostv2.dll to be put into the OS C:\windows\system32\
@@ -36,16 +46,16 @@
 		* `/src/host/ft_resize` – Special test for resizing/reflowing the buffer window
 		* `/src/host/ft_uia` – Currently disabled (for not being very reliable) UI Automation tests that we are looking to re-enable and expand to do UI Automation coverage of various human interactions
 		* `/src/host/...` - The files I’ll list out below
-		* `/src/inc` – Include files that are shared between the host and some of the other libraries. This is only some of them. The include story is kind of a mess right now, but we’d like to clean it up at some point
-		* `/src/propslib` – Library shared between console host and the OS shell “right click a shortcut file and modify console properties” page to read/write user settings to and from the registry and embedded within shortcut LNK data
-		* `/src/renderer` – Refactored extraction of all activities related to rendering the text in the buffers onto the screen
-			* `/src/renderer/base` – Base interface layer providing non-engine-specific rendering things like choosing the data from the console buffer, deciding how to lay out or transform that data, then dispatching commands to a specific final display engine
-			* `/src/renderer/gdi` – The GDI implementation of rendering to the screen. Takes commands to “draw a line” or “fill the background” or “select a region” from the base and turns them into GDI calls to the screen. Extracted from original console host code.
-			* `/src/renderer/inc – Interface definitions for all renderer communication
-		* `/src/terminal` – Virtual terminal support for the console. This is the sequences that are found in-band with other text on STDIN/STDOUT that command the display to do things. This is the *nix way of controlling a console.
-			* `/src/terminal/parser` – This contains a state machine and sorting engine for feeding in individual characters from STDOUT or STDIN and decoding them into the appropriate verbs that should be performed
-			* `/src/terminal/adapter` – This converts the verbs from the interface into calls on the console API. It doesn’t actually call through the API (for performance reasons since it lives inside the same binary), but it tries to remain as close to an API call as possible. There are some private extensions to the API for behaviors that didn’t exist before this was written that we’ve not made public. We don’t know if we will yet or force people to use VT to get at them.
-		* `/src/tsf` – Text Services Foundation. This provides IME input services to the console. This was historically used for only Chinese, Japanese, and Korean IMEs specifically on OS installations with those as the primary language. It was in the summer of 2016 unrestricted to be able to be used on any OS installation with any IME (whether or not it will display correctly is a different story). It also was unrestricted to allow things like Pen and Touch input (which are routed via IME messages) to display properly inside the console from the TabTip window (the little popup that helps you insert pen/touch writing/keyboard candidates into an application)
+	* `/src/inc` – Include files that are shared between the host and some of the other libraries. This is only some of them. The include story is kind of a mess right now, but we’d like to clean it up at some point
+	* `/src/propslib` – Library shared between console host and the OS shell “right click a shortcut file and modify console properties” page to read/write user settings to and from the registry and embedded within shortcut LNK data
+	* `/src/renderer` – Refactored extraction of all activities related to rendering the text in the buffers onto the screen
+		* `/src/renderer/base` – Base interface layer providing non-engine-specific rendering things like choosing the data from the console buffer, deciding how to lay out or transform that data, then dispatching commands to a specific final display engine
+		* `/src/renderer/gdi` – The GDI implementation of rendering to the screen. Takes commands to “draw a line” or “fill the background” or “select a region” from the base and turns them into GDI calls to the screen. Extracted from original console host code.
+		* `/src/renderer/inc` – Interface definitions for all renderer communication
+	* `/src/terminal` – Virtual terminal support for the console. This is the sequences that are found in-band with other text on STDIN/STDOUT that command the display to do things. This is the \*nix way of controlling a console.
+		* `/src/terminal/parser` – This contains a state machine and sorting engine for feeding in individual characters from STDOUT or STDIN and decoding them into the appropriate verbs that should be performed
+		* `/src/terminal/adapter` – This converts the verbs from the interface into calls on the console API. It doesn’t actually call through the API (for performance reasons since it lives inside the same binary), but it tries to remain as close to an API call as possible. There are some private extensions to the API for behaviors that didn’t exist before this was written that we’ve not made public. We don’t know if we will yet or force people to use VT to get at them.
+	* `/src/tsf` – Text Services Foundation. This provides IME input services to the console. This was historically used for only Chinese, Japanese, and Korean IMEs specifically on OS installations with those as the primary language. It was in the summer of 2016 unrestricted to be able to be used on any OS installation with any IME (whether or not it will display correctly is a different story). It also was unrestricted to allow things like Pen and Touch input (which are routed via IME messages) to display properly inside the console from the TabTip window (the little popup that helps you insert pen/touch writing/keyboard candidates into an application)
 
 ## Host File Overview
 
@@ -90,7 +100,7 @@
 * Assorted utilities and stuff
 	* `Misc.cpp` (left for us by previous eras of random console devs)
 	* `Util.cpp` (created in our era)
-* Custom zeroing and non-throwing allocator 
+* Custom zeroing and non-throwing allocator
 	* `Newdelete.cpp`
 * Related to inserting text into the TextInfo buffer
 	* `Output.cpp`

doc/STYLE.md

@@ -3,5 +3,6 @@
 ## Philosophy
 1. If it's inserting something into the existing classes/functions, try to follow the existing style as closely as possible.
 1. If it's brand new code or refactoring a complete class or area of the code, please follow as Modern C++ of a style as you can and reference the [C++ Core Guidelines](https://github.com/isocpp/CppCoreGuidelines) as much as you possibly can.
-1. When working with any Win32 or NT API, please try to use the [Windows Internal Library](./WIL.md) smart pointers and result handlers.
+1. When working with any Win32 or NT API, please try to use the [Windows Implementation Library](./WIL.md) smart pointers and result handlers.
 1. The use of NTSTATUS as a result code is discouraged, HRESULT or exceptions are preferred. Functions should not return a status code if they would always return a successful status code. Any function that returns a status code should be marked `noexcept` and have the `nodiscard` attribute.
+1. When contributing code in `TerminalApp`, be mindful to appropriately use C++/WinRT [strong and weak references](https://docs.microsoft.com/en-us/windows/uwp/cpp-and-winrt-apis/weak-references), and have a good understanding of C++/WinRT [concurrency schemes](https://docs.microsoft.com/en-us/windows/uwp/cpp-and-winrt-apis/concurrency).

doc/UniversalTest.md

@@ -7,13 +7,13 @@ Universal Testing is the Microsoft framework for creating and deploying test pac
 It involves several parts:
 - TESTMD
   - These define a package unit for deployment to the test device. This usually includes the test binaries and any dependent data that it will need to execute.
-  - There can also be a hierarchy where one package can depend on another such that packages can be re-used
+  - There can also be a hierarchy where one package can depend on another such that packages can be re-used.
 
 - TESTLIST
   - This defines a batch of TESTMD packages that should be executed together.
 
 - TESTPASSES
-  - This defines a list of tests via a TESTLIST and a lab environment configuration on which the tests should be run
+  - This defines a list of tests via a TESTLIST and a lab environment configuration on which the tests should be run.
 
   These files can either include their child element as they're supposed to (TESTMDs included in TESTLISTs) or they can often include themselves to provide chain structuring (one TESTLIST can reference another TESTLIST).
 

doc/WIL.md

@@ -1,7 +1,7 @@
-# Windows Internal Library
+# Windows Implementation Library
 
 ## Overview
-Windows Internal Library, or WIL, is a header-only library created to help make working with the Windows API more predictable and (hopefully) bug free.
+[Windows Implementation Library](https://github.com/Microsoft/wil), or WIL, is a header-only library created to help make working with the Windows API more predictable and (hopefully) bug free.
 
 A majority of functions are in either the `wil::` or `wistd::` namespace. `wistd::` is used for things that have an equivalent in STL's `std::` namespace but have some special functionality like being exception-free. Everything else is in `wil::` namespace.
 
@@ -9,13 +9,13 @@ The primary usages of WIL in our code so far are...
 
 ### Smart Pointers ###
 
-Inside [wil\resource.h](..\dep\wil\resource.h) are smart pointer like classes for many Windows OS resources like file handles, socket handles, process handles, and so on. They're of the form `wil::unique_handle` and call the appropriate/matching OS function (like `CloseHandle()` in this case) when they go out of scope.
+Inside [wil\resource.h](https://github.com/microsoft/wil/blob/master/include/wil/resource.h) are smart pointer like classes for many Windows OS resources like file handles, socket handles, process handles, and so on. They're of the form `wil::unique_handle` and call the appropriate/matching OS function (like `CloseHandle()` in this case) when they go out of scope.
 
 Another useful item is `wil::make_unique_nothrow()` which is analogous to `std::make_unique` (except without the exception which might help you integrate with existing exception-free code in the console.) This will return a `wistd::unique_ptr` (vs. a `std::unique_ptr`) which can be used in a similar manner.
 
 ### Result Handling ###
 
-To manage the various types of result codes that come back from Windows APIs, the file [wil\result.h](..\dep\wil\result.h) provides a wealth of macros that can help. 
+To manage the various types of result codes that come back from Windows APIs, the file [wil\result.h](https://github.com/microsoft/wil/blob/master/include/wil/result.h) provides a wealth of macros that can help. 
 
 As an example, the method `DuplicateHandle()` returns a `BOOL` value that is `FALSE` under failure and would like you to `GetLastError()` from the operating system to find out what the actual result code is. In this circumstance, you could use the macro `RETURN_IF_WIN32_BOOL_FALSE` to wrap the call to `DuplicateHandle()` which would automatically handle this pattern for you and return the `HRESULT` equivalent on failure.
 

doc/WindowsTestPasses.md

@@ -16,7 +16,7 @@ The next step would be investigating one of these failures...
 
 A quick overview of investigation... normally you can just attempt to build and reproduce the failure locally with the `OpenConsole` project and it will happen the same way as it did on the nightly build in the lab. However, sometimes the failure will be exclusive to the lab or won't happen in the same way as it does on your local dev machine. At that point, you need to move into setting up the environment as it was during the testpass and figuring out what went wrong.
 
-You can try to do this all manually by pulling down a VM image from the release share for the nightly build, making a VM, deploying the test binaries and TAEF test runnner executables to the machine, installing the VS Remote Debugging or WinDBG tools on the VM, and then running the test and figuring out what's going wrong with the debuggers.
+You can try to do this all manually by pulling down a VM image from the release share for the nightly build, making a VM, deploying the test binaries and TAEF test runner executables to the machine, installing the VS Remote Debugging or WinDBG tools on the VM, and then running the test and figuring out what's going wrong with the debuggers.
 
 Or you can use some of the Engineering Systems tools to make this easier. I'll detail how to do that below.
 
@@ -24,12 +24,12 @@ Prerequisites:
 - Visual Studio 2017
 - Install the TDP (Test Development Platform) plug-in (see: [https://osgwiki.com/wiki/Test_Development_Platform_(TDP)]).
 
-1. Open Visual Studio 2017 and use the TDP drop-down menu to open the `Device Manager`
+1. Open Visual Studio 2017 and use the TDP drop-down menu to open the `Device Manager`.
 1. In the pane that opens to the left, choose `Add` and then `Nebula VM Device`. Nebula is a cloud provider for VMs (like Azure but a more private instance for corporate work usage).
-1. Name the machine and choose the build/branch/flavor/SKU from the drop downs at the bottom. It will find the VHD for you from the build shares. Hit `Add Device` to deploy to Nebula
+1. Name the machine and choose the build/branch/flavor/SKU from the drop downs at the bottom. It will find the VHD for you from the build shares. Hit `Add Device` to deploy to Nebula.
 1. Wait a few minutes. It took 5-10 for it to be deployed.
 1. Right click the machine name in the `Device Manager` list and choose `Launch T-Shell`. You can also use `Connect via Console` to get a "remote desktop"-like session to the KVM port on the VM.
-1. In T-shell, use `testd Microsoft.Console.TestLab.Desktop.testlist` or a command of that format with a different TESTLIST or TESTMD name from our project (see the [UniversalTest.md] documentation). The `testd` utility will automatically resolve the build/branch/flavor information, dig through the build shares for the matching TESTLIST/TESTMD metadata, and attempt to deploy all relevent packages and dependencies on the device. When it's successful, it will move onto running all the tests and giving you the results. On conclusion, the test results should pop up in the web browser or the `Hubble - Log Viewer` tool provided by the Engineering Systems team.
+1. In T-shell, use `testd Microsoft.Console.TestLab.Desktop.testlist` or a command of that format with a different TESTLIST or TESTMD name from our project (see the [UniversalTest.md] documentation). The `testd` utility will automatically resolve the build/branch/flavor information, dig through the build shares for the matching TESTLIST/TESTMD metadata, and attempt to deploy all relevant packages and dependencies on the device. When it's successful, it will move onto running all the tests and giving you the results. On conclusion, the test results should pop up in the web browser or the `Hubble - Log Viewer` tool provided by the Engineering Systems team.
 
 If some of the above things do not work, go to [https://osgwiki.com] and type them into the Search bar. For instance, if T-Shell isn't found or working, you can find out where to get it or download it on `OSGWiki`. The same goes for the other commands besides `testd` to use in T-shell and more information on what `Hubble` or `Nebula` are. 
 

doc/bot.md

@@ -0,0 +1,130 @@
+# Issue/PR Management Bot Information
+
+## Overview
+
+The goal here is to help us automate, manage, and narrow down what we actually need to focus on in this repository.
+We'll be using tags, primarily, to help us understand what needs attention, what is sitting around and turning stale, etc.
+
+### Quick-Guidance to Core Contributors
+1. Look at `Needs-Attention` as top priority
+1. Look at `Needs-Triage` during triage meetings to get a handle on what's new and sort it out
+1. Look at `Needs-Tag-Fix` when you have a few minutes to fix up things tagged improperly
+1. Manually add `Needs-Author-Feedback` when there's something we need the author to follow up on and want attention if they return it or an auto-close for inactivity if it goes stale.
+
+### Tagging/Process Details
+1. When new issues arrive, or when issues are not properly tagged... we'll mark them as `Needs-Triage` automatically.
+   - The core contributor team will then come through and mark them up as appropriate. The goal is to have a tag that fits the `Product`, `Area`, and `Issue` category.
+   - The `Needs-Triage` tag will be removed manually by the core contributor team during a triage meeting. (Exception, triage may also be done offline by senior team members during high-volume times.)
+   - An issue may or may not be assigned to a contributor during triage. It is not necessary to assign someone to complete it.
+   - We're not focusing on Projects yet.
+1. When core contributors need to ask something of the author, they will manually assign the `Needs-Author-Feedback` tag.
+   - This tag will automatically drop off when the author comes back around and applies activity to the thread.
+   - When this tag drops off, the bot will apply the `Needs-Attention` tag to get the core contribution team's attention again. If an author cares enough to be active, we will attempt to prioritize engaging with that author.
+   - If the author doesn't come back around in a while, this will become a `No-Recent-Activity` tag.
+   - If there's activity on an issue, the `No-Recent-Activity` tag will automatically drop.
+   - If the `No-Recent-Activity` stays, the issue will be closed as stale.
+1. PRs will automatically get a `Needs-Author-Feedback` tag when reviewers wait on the author
+   - This follows a similar decay strategy to issues.
+   - If the author responds, the `Needs-Author-Feedback` tag will drop.
+   - If there is no activity in a while, the `No-Recent-Activity` tag will appear.
+   - If the `No-Recent-Activity` tag exists for a while, the PR will be closed as stale.
+1. Issues manually marked as `Resolution-Duplicate` will be closed shortly after activity stops
+1. Pull requests manually marked as `AutoMerge` will permit the bot to complete the PR and do cleanup when certain conditions are met. See details below.
+
+## Rules
+
+### Triage Shorthand
+- All rules in this category apply to triaging issues. They're shorthand comments that the triage team can use in order to complete the triage process faster. 
+- Only individuals with `Write` or `Admin` privileges on the repository can use these responses.
+
+#### Duplicate Issues
+- When a comment on the thread says `/dup #<issue ID>`...
+1. Reply with a comment explaining that the issue is a duplicate and recommend that the opener and interested parties follow the issue on the listed ID number.
+1. Close the issue
+1. Remove all `Needs-*` tags
+1. Add `Resolution-Duplicate`
+
+### Issue Management
+
+#### Mark as Triage Needed
+- When an issue doesn't meet triage criteria, applies `Needs-Triage` tag. Right now, this is just when it's opened.
+
+#### Author Has Responded
+- When an issue with `Needs-Author-Feedback` gets an author response, drops that tag in favor of `Needs-Attention` to flag core contributors to drop by.
+
+#### Remove Activity Tag
+- When an issue with `No-Recent-Activity` has activity, drops this tag
+
+#### Close Stale
+- Every hour, checks if there's an issue with `Needs-Author-Feedback` and `No-Recent-Activity` for 3 days. Closes as stale.
+
+#### Tag as No Activity
+- Every hour, checks if there's been no activity in 4 days on an issue that `Needs-Author-Feedback`. If it's been 4 days, mark `No-Recent-Activity` as well.
+
+#### Close Duplicate Issues
+- Every hour, checks if there's been a day since the last activity on an issue with tag `Resolution-Duplicate` and closes it if inactive.
+
+#### Enforce tag system
+- When an issue is opened or labels are changed in any way, we will check if the tagging matches the system. If not, it will get `Needs-Tag-Fix`. The system is to have an `Area-`, `Issue-`, and `Product-` tag for all open things, and also a `Resolution-` for closed ones.
+- When the tags from appropriate categories are applied, it will auto-remove the `Needs-Tag-Fix` tag.
+- `Resolution-Duplicate` is sufficient to fix all tagging. (`Area-`, `Issue-`, and `Product-` are not needed for a duplicate.)
+
+#### Clean-up low quality issues
+- If an issue is filed with an incomplete title...
+- If an issue is filed with nothing in the body...
+- If an issue is filed matching a pattern that happens all the time (common duplicate phrase, obvious multiple-issues-in-one pattern)...
+- Then close the issue automatically informing the opener that they can resolve the problem and reopen the issue. (See Bug/Feature templates for example situations.)
+
+#### Help ask for Feedback Hub
+- When a comment on the thread says `/feedback`...
+1. Then reply to the issue with a bit of text on asking the author to send us data with Feedback Hub and give us the link.
+1. And add the `Needs-Author-Feedback` tag
+
+#### Remove Help Wanted from In PR issues
+- If an issue gets the `In-PR` tag when a new PR is created, we will remove the `Help-Wanted` tag to avoid someone trying to work on an issue where another person has already submitted a proposed fix.
+
+### PR Management
+
+#### Codeflow Link *(Disabled)*
+- Bumps a PR with a link to the Microsoft CodeFlow tool for reviewing PRs
+
+#### Marks PR as Awaiting Author Feedback
+- When a reviewer marks the PR as changes requested, the `Needs-Author-Feedback` tag will be applied
+
+#### Removes Awaiting Author Feedback
+- When the PR author updates the pull request, comments on it, or responds to a review, the `Needs-Author-Feedback` tag is removed.
+
+#### Removes No Recent Activity
+- When anyone touches the pull request, the `No-Recent-Activity` tag is removed.
+
+#### Markup stale pull requests
+- Every hour, if a pull request `Needs-Author-Feedback` and hasn't been touched in 7 days, it will get the `No-Recent-Activity` tag.
+
+#### Close stale pull requests
+- Every hour, if a pull request has `No-Recent-Activity` and hasn't been touched in a further 7 days, it will be closed.
+
+#### Auto-Merge pull requests
+- When a pull request has the `AutoMerge` label...
+  - If it has been at least 480 minutes and all the statuses pass, merge it in.
+  - Will use Squash merge strategy
+  - Will attempt to delete branch after merge, if possible
+  - Will automatically remove the `AutoMerge` label if changes are pushed by someone *without* Write Access.
+  - More information on bot-logic that can be controlled with comments is [here](https://github.com/OfficeDev/office-ui-fabric-react/wiki/Advanced-auto-merge)
+
+#### Mark issues with an active PR
+- If there is an active PR for an issue, label that issue with the `In-PR` label
+
+#### Add committed fix tag for completed PRs
+- When a PR is finished and there's no outstanding work left on a linked issue, add the `Resolution-Fix-Committed` label
+
+#### Remove Needs-Second from completed PRs
+- If a PR is closed and it has the `Needs-Second` tag, the bot will remove the tag.
+
+### Release Management
+
+When a release is created, if the PR ID number is linked inside the release description, the bot will walk through the related PR and all of its related issues and leave a message.
+- PR message: "🎉{release name} {release version} has been released which incorporates this pull request.🎉
+- Issue message: 🎉This issue was addressed in #{pull request ID}, which has now been successfully released as {release name} {release version}.🎉"
+
+## Admin Panel
+[Here](https://fabric-cp.azurewebsites.net/bot/)

doc/building.md

@@ -1,25 +1,29 @@
 
 # How to build Openconsole
 
-Openconsole can be built with Visual Studio or from the command line. There are build scripts for both cmd and powershell in /tools.
+Openconsole can be built with Visual Studio or from the command line. There are build scripts for both cmd and PowerShell in /tools.
+
+When using Visual Studio, be sure to set up the path for code formatting. This can be done in Visual Studio by going to Tools > Options > Text Editor > C++ > Formatting and checking "Use custom clang-format.exe file" and choosing the clang-format.exe in the repository at /dep/llvm/clang-format.exe by clicking "browse" right under the check box.
 
 ## Building with cmd
 
 The cmd scripts are set up to emulate a portion of the OS razzle build environment. razzle.cmd is the first script that should be run. bcz.cmd will build clean and bz.cmd should build incrementally.
 
 There are also scripts for running the tests:
-- runut.cmd - run the unit tests
-- runft.cmd - run the feature tests
-- runuia.cmd - run the UIA tests
+- `runut.cmd` - run the unit tests
+- `runft.cmd` - run the feature tests
+- `runuia.cmd` - run the UIA tests
+- `runformat` - uses clang-format to format all c++ files to match our coding style.
 
 ## Build with Powershell
 
 Openconsole.psm1 should be loaded with `Import-Module`. From there `Set-MsbuildDevEnvironment` will set up environment variables required to build. There are a few exported functions (look at their documentation for further details):
 
-- Invoke-OpenConsolebuild - builds the solution. Can be passed msbuild arguments.
-- Invoke-OpenConsoleTests - runs the various tests. Will run the unit tests by default.
-- Start-OpenConsole - starts Openconsole.exe from the output directory. x64 is run by default.
-- Debug-OpenConsole - starts Openconsole.exe and attaches it to the default debugger. x64 is run by default.
+- `Invoke-OpenConsolebuild` - builds the solution. Can be passed msbuild arguments.
+- `Invoke-OpenConsoleTests` - runs the various tests. Will run the unit tests by default.
+- `Start-OpenConsole` - starts Openconsole.exe from the output directory. x64 is run by default.
+- `Debug-OpenConsole` - starts Openconsole.exe and attaches it to the default debugger. x64 is run by default.
+- `Invoke-CodeFormat` - uses clang-format to format all c++ files to match our coding style.
 
 ## Configuration Types
 
@@ -29,4 +33,4 @@ Openconsole has three configuration types:
 - Release
 - AuditMode
 
-AuditMode is an experimental mode that enables some additional static analyis from CppCoreCheck.
+AuditMode is an experimental mode that enables some additional static analysis from CppCoreCheck.

doc/cascadia/Keybindings-spec.md

@@ -6,7 +6,7 @@
 ## Abstract
 It should be possible to configure the terminal so that it doesn't send certain keystrokes as input to the terminal, and instead triggers certain actions. Examples of these actions could be copy/pasting text, opening a new tab, or changing the font size.
 
-This spec describes a mechanism by which we could provide a common implementation of handling keyboard shortcuts like these. This common implementation could then be leveraged and extended by the UX implementation as to handle certain callbacks in the UX layer. For example, The TerminalCore doesn't have a concept of what a tab is, but the keymap abstraction could raise an event such that a WPF app could implement creating a new tab in it's idomatic way, and UWP could implement them in their own way.
+This spec describes a mechanism by which we could provide a common implementation of handling keyboard shortcuts like these. This common implementation could then be leveraged and extended by the UX implementation as to handle certain callbacks in the UX layer. For example, The TerminalCore doesn't have a concept of what a tab is, but the keymap abstraction could raise an event such that a WPF app could implement creating a new tab in its idiomatic way, and UWP could implement them in their own way.
 
 ## Terminology
 * **Key Chord**: This is any possible keystroke that a user can input
@@ -28,7 +28,7 @@ This spec describes a mechanism by which we could provide a common implementatio
 
 When the UX frontend is created, it should instantiate a `IKeyBindings` object with the keybindings mapped as it would like.
 
-When it's creating it's platform-dependent terminal component, it can pass the `IKeyBindings` object to that component. The component will then be able to pass that object to the terminal instance.
+When it's creating its platform-dependent terminal component, it can pass the `IKeyBindings` object to that component. The component will then be able to pass that object to the terminal instance.
 
 When the terminal component calls `ITerminalInput.SendKeyEvent(uint vkey, KeyModifiers modifiers)`, the terminal will use `IKeyBindings.TryKeyChord` to see if there are any bound actions to that input. If there are, the `IKeyBindings` implementation will either handle the event by interacting with the `ITerminalInput`, or it'll invoke an event that's been registered by the frontend
 
@@ -66,7 +66,7 @@ partial class Terminal
 ```
 ### Project Cascadia Sample
 
-Below is an example of how the Project Cascadia application might implement it's
+Below is an example of how the Project Cascadia application might implement its
   keybindings.
 
 ```csharp
@@ -105,4 +105,4 @@ class KeyBindings : IKeyBindings
 How does Copy/paste play into this?
 
 When Input is written to the terminal, and it tries the copy keybinding, what happens?
-The Keybindings are global to the frontend, not local to the terminal. Copy/Paste events should also be delegates that get raised, and the frontend can then determine what to do with them. It'll probably query it's active/focused Terminal Component, then Get the `ITerminalInput` from that component, and use that to CopyText / PasteText from the Terminal as needed.
+The Keybindings are global to the frontend, not local to the terminal. Copy/Paste events should also be delegates that get raised, and the frontend can then determine what to do with them. It'll probably query its active/focused Terminal Component, then Get the `ITerminalInput` from that component, and use that to CopyText / PasteText from the Terminal as needed.

doc/cascadia/SettingsSchema.md

@@ -0,0 +1,194 @@
+# Profiles.json Documentation
+
+## Globals
+Properties listed below affect the entire window, regardless of the profile settings.
+
+| Property | Necessity | Type | Default | Description |
+| -------- | --------- | ---- | ------- | ----------- |
+| `alwaysShowTabs` | _Required_ | Boolean | `true` | When set to `true`, tabs are always displayed. When set to `false` and `showTabsInTitlebar` is set to `false`, tabs only appear after typing <kbd>Ctrl</kbd> + <kbd>T</kbd>. |
+| `copyOnSelect` | Optional | Boolean | `false` | When set to `true`, a selection is immediately copied to your clipboard upon creation. When set to `false`, the selection persists and awaits further action. |
+| `defaultProfile` | _Required_ | String | PowerShell guid | Sets the default profile. Opens by typing <kbd>Ctrl</kbd> + <kbd>T</kbd> or by clicking the '+' icon. The guid of the desired default profile is used as the value. |
+| `initialCols` | _Required_ | Integer | `120` | The number of columns displayed in the window upon first load. |
+| `initialPosition` | Optional | String | `","` | The position of the top left corner of the window upon first load. On a system with multiple displays, these coordinates are relative to the top left of the primary display. If `launchMode` is set to `"maximized"`, the window will be maximized on the monitor specified by those coordinates. |
+| `initialRows` | _Required_ | Integer | `30` | The number of rows displayed in the window upon first load. |
+| `launchMode` | Optional | String | `default` | Defines whether the Terminal will launch as maximized or not. Possible values: `"default"`, `"maximized"` |
+| `rowsToScroll` | Optional | Integer | `system` | The number of rows to scroll at a time with the mouse wheel. This will override the system setting if the value is not zero or "system". |
+| `requestedTheme` | _Required_ | String | `system` | Sets the theme of the application. Possible values: `"light"`, `"dark"`, `"system"` |
+| `showTerminalTitleInTitlebar` | _Required_ | Boolean | `true` | When set to `true`, titlebar displays the title of the selected tab. When set to `false`, titlebar displays "Windows Terminal". |
+| `showTabsInTitlebar` | Optional | Boolean | `true` | When set to `true`, the tabs are moved into the titlebar and the titlebar disappears. When set to `false`, the titlebar sits above the tabs. |
+| `snapToGridOnResize` | Optional | Boolean | `false` | When set to `true`, the window will snap to the nearest character boundary on resize. When `false`, the window will resize "smoothly" |
+| `tabWidthMode` | Optional | String | `equal` | Sets the width of the tabs. Possible values: `"equal"`, `"titleLength"` |
+| `wordDelimiters` | Optional | String | <code>&nbsp;&#x2f;&#x5c;&#x28;&#x29;&#x22;&#x27;&#x2d;&#x3a;&#x2c;&#x2e;&#x3b;&#x3c;&#x3e;&#x7e;&#x21;&#x40;&#x23;&#x24;&#x25;&#x5e;&#x26;&#x2a;&#x7c;&#x2b;&#x3d;&#x5b;&#x5d;&#x7b;&#x7d;&#x7e;&#x3f;│</code><br>_(`│` is `U+2502 BOX DRAWINGS LIGHT VERTICAL`)_ | Determines the delimiters used in a double click selection. |
+| `confirmCloseAllTabs` | Optional | Boolean | `true` | When set to `true` closing a window with multiple tabs open WILL require confirmation.  When set to `false` closing a window with multiple tabs open WILL NOT require confirmation. |
+
+## Profiles
+Properties listed below are specific to each unique profile.
+
+| Property | Necessity | Type | Default | Description |
+| -------- | --------- | ---- | ------- | ----------- |
+| `guid` | _Required_ | String | | Unique identifier of the profile. Written in registry format: `"{00000000-0000-0000-0000-000000000000}"`. |
+| `name` | _Required_ | String | | Name of the profile. Displays in the dropdown menu. <br>Additionally, this value will be used as the "title" to pass to the shell on startup. Some shells (like `bash`) may choose to ignore this initial value, while others (`cmd`, `powershell`) may use this value over the lifetime of the application. This "title" behavior can be overridden by using `tabTitle`. |
+| `acrylicOpacity` | Optional | Number | `0.5` | When `useAcrylic` is set to `true`, it sets the transparency of the window for the profile. Accepts floating point values from 0-1. |
+| `background` | Optional | String | | Sets the background color of the profile. Overrides `background` set in color scheme if `colorscheme` is set. Uses hex color format: `"#rrggbb"`. |
+| `backgroundImage` | Optional | String | | Sets the file location of the Image to draw over the window background. |
+| `backgroundImageAlignment` | Optional | String | `center` | Sets how the background image aligns to the boundaries of the window. Possible values: `"center"`, `"left"`, `"top"`, `"right"`, `"bottom"`, `"topLeft"`, `"topRight"`, `"bottomLeft"`, `"bottomRight"` |
+| `backgroundImageOpacity` | Optional | Number | `1.0` | Sets the transparency of the background image. Accepts floating point values from 0-1. |
+| `backgroundImageStretchMode` | Optional | String | `uniformToFill` | Sets how the background image is resized to fill the window. Possible values: `"none"`, `"fill"`, `"uniform"`, `"uniformToFill"` |
+| `closeOnExit` | Optional | String | `graceful` | Sets how the profile reacts to termination or failure to launch. Possible values: `"graceful"` (close when `exit` is typed or the process exits normally), `"always"` (always close) and `"never"` (never close). `true` and `false` are accepted as synonyms for `"graceful"` and `"never"` respectively. |
+| `colorScheme` | Optional | String | `Campbell` | Name of the terminal color scheme to use. Color schemes are defined under `schemes`. |
+| `colorTable` | Optional | Array[String] | | Array of colors used in the profile if `colorscheme` is not set. Array follows the format defined in `schemes`. |
+| `commandline` | Optional | String | | Executable used in the profile. |
+| `cursorColor` | Optional | String | `#FFFFFF` | Sets the cursor color for the profile. Uses hex color format: `"#rrggbb"`. |
+| `cursorHeight` | Optional | Integer | | Sets the percentage height of the cursor starting from the bottom. Only works when `cursorShape` is set to `"vintage"`. Accepts values from 25-100. |
+| `cursorShape` | Optional | String | `bar` | Sets the cursor shape for the profile. Possible values: `"vintage"` ( &#x2583; ), `"bar"` ( &#x2503; ), `"underscore"` ( &#x2581; ), `"filledBox"` ( &#x2588; ), `"emptyBox"` ( &#x25AF; ) |
+| `fontFace` | Optional | String | `Consolas` | Name of the font face used in the profile. We will try to fallback to Consolas if this can't be found or is invalid. |
+| `fontSize` | Optional | Integer | `12` | Sets the font size. |
+| `foreground` | Optional | String | | Sets the foreground color of the profile. Overrides `foreground` set in color scheme if `colorscheme` is set. Uses hex color format: `#rgb` or `"#rrggbb"`. |
+| `hidden` | Optional | Boolean | `false` | If set to true, the profile will not appear in the list of profiles. This can be used to hide default profiles and dynamically generated profiles, while leaving them in your settings file. |
+| `historySize` | Optional | Integer | `9001` | The number of lines above the ones displayed in the window you can scroll back to. |
+| `icon` | Optional | String | | Image file location of the icon used in the profile. Displays within the tab and the dropdown menu. |
+| `padding` | Optional | String | `8, 8, 8, 8` | Sets the padding around the text within the window. Can have three different formats: `"#"` sets the same padding for all sides, `"#, #"` sets the same padding for left-right and top-bottom, and `"#, #, #, #"` sets the padding individually for left, top, right, and bottom. |
+| `scrollbarState` | Optional | String | | Defines the visibility of the scrollbar. Possible values: `"visible"`, `"hidden"` |
+| `selectionBackground` | Optional | String | | Sets the selection background color of the profile. Overrides `selectionBackground` set in color scheme if `colorscheme` is set. Uses hex color format: `"#rrggbb"`. |
+| `snapOnInput` | Optional | Boolean | `true` | When set to `true`, the window will scroll to the command input line when typing. When set to `false`, the window will not scroll when you start typing. |
+| `source` | Optional | String | | Stores the name of the profile generator that originated this profile. _There are no discoverable values for this field._ |
+| `startingDirectory` | Optional | String | `%USERPROFILE%` | The directory the shell starts in when it is loaded. |
+| `suppressApplicationTitle` | Optional | Boolean | | When set to `true`, `tabTitle` overrides the default title of the tab and any title change messages from the application will be suppressed. When set to `false`, `tabTitle` behaves as normal. |
+| `tabTitle` | Optional | String | | If set, will replace the `name` as the title to pass to the shell on startup. Some shells (like `bash`) may choose to ignore this initial value, while others (`cmd`, `powershell`) may use this value over the lifetime of the application.  |
+| `useAcrylic` | Optional | Boolean | `false` | When set to `true`, the window will have an acrylic background. When set to `false`, the window will have a plain, untextured background. The transparency only applies to focused windows due to OS limitation. |
+| `experimental.retroTerminalEffect` | Optional | Boolean | `false` | When set to `true`, enable retro terminal effects. This is an experimental feature, and its continued existence is not guaranteed. |
+
+## Schemes
+Properties listed below are specific to each color scheme. [ColorTool](https://github.com/microsoft/terminal/tree/master/src/tools/ColorTool) is a great tool you can use to create and explore new color schemes. All colors use hex color format.
+
+| Property | Necessity | Type | Description |
+| -------- | ---- | ----------- | ----------- |
+| `name` | _Required_ | String | Name of the color scheme. |
+| `foreground` | _Required_ | String | Sets the foreground color of the color scheme. |
+| `background` | _Required_ | String | Sets the background color of the color scheme. |
+| `selectionBackground` | Optional | String | Sets the selection background color of the color scheme. |
+| `black` | _Required_ | String | Sets the color used as ANSI black. |
+| `blue` | _Required_ | String | Sets the color used as ANSI blue. |
+| `brightBlack` | _Required_ | String | Sets the color used as ANSI bright black. |
+| `brightBlue` | _Required_ | String | Sets the color used as ANSI bright blue. |
+| `brightCyan` | _Required_ | String | Sets the color used as ANSI bright cyan. |
+| `brightGreen` | _Required_ | String | Sets the color used as ANSI bright green. |
+| `brightPurple` | _Required_ | String | Sets the color used as ANSI bright purple. |
+| `brightRed` | _Required_ | String | Sets the color used as ANSI bright red. |
+| `brightWhite` | _Required_ | String | Sets the color used as ANSI bright white. |
+| `brightYellow` | _Required_ | String | Sets the color used as ANSI bright yellow. |
+| `cyan` | _Required_ | String | Sets the color used as ANSI cyan. |
+| `green` | _Required_ | String | Sets the color used as ANSI green. |
+| `purple` | _Required_ | String | Sets the color used as ANSI purple. |
+| `red` | _Required_ | String | Sets the color used as ANSI red. |
+| `white` | _Required_ | String | Sets the color used as ANSI white. |
+| `yellow` | _Required_ | String | Sets the color used as ANSI yellow. |
+
+## Keybindings
+Properties listed below are specific to each custom key binding.
+
+| Property | Necessity | Type | Description |
+| -------- | ---- | ----------- | ----------- |
+| `command` | _Required_ | String | The command executed when the associated key bindings are pressed. |
+| `keys` | _Required_ | Array[String] | Defines the key combinations used to call the command. |
+| `action` | Optional | String | Adds additional functionality to certain commands. |
+
+### Implemented Commands and Actions
+
+Commands listed below are per the implementation in [`src/cascadia/TerminalApp/AppKeyBindingsSerialization.cpp`](https://github.com/microsoft/terminal/blob/master/src/cascadia/TerminalApp/AppKeyBindingsSerialization.cpp).
+
+Keybindings can be structured in the following manners:
+
+For commands without arguments:
+<br>
+`{ "command": "commandName", "keys": [ "modifiers+key" ] }`
+
+For commands with arguments:
+<br>
+`{ "command": { "action": "commandName", "argument": "value" }, "keys": ["modifiers+key"] }`
+
+| Command | Command Description | Action (*=required) | Action Arguments | Argument Descriptions |
+| ------- | ------------------- | ------ | ---------------- | ----------------- |
+| closePane | Close the active pane. | | | |
+| closeTab | Close the current tab. | | | |
+| closeWindow | Close the current window and all tabs within it. | | | |
+| copy | Copy the selected terminal content to your Windows Clipboard. | `trimWhitespace` | boolean | When `true`, newlines persist from the selected text. When `false`, copied content will paste on one line. |
+| decreaseFontSize | Make the text smaller by one delta. | `delta` | integer | Amount of size decrease per command invocation. |
+| duplicateTab | Make a copy and open the current tab. | | | |
+| find | Open the search dialog box. | | | |
+| increaseFontSize | Make the text larger by one delta. | `delta` | integer | Amount of size increase per command invocation. |
+| moveFocus | Focus on a different pane depending on direction. | `direction`* | `left`, `right`, `up`, `down` | Direction in which the focus will move. |
+| newTab | Create a new tab. Without any arguments, this will open the default profile in a new tab. | 1. `commandLine`<br>2. `startingDirectory`<br>3. `tabTitle`<br>4. `index`<br>5. `profile` | 1. string<br>2. string<br>3. string<br>4. integer<br>5. string | 1. Executable run within the tab.<br>2. Directory in which the tab will open.<br>3. Title of the new tab.<br>4. Profile that will open based on its position in the dropdown (starting at 0).<br>5. Profile that will open based on its GUID or name. |
+| nextTab | Open the tab to the right of the current one. | | | |
+| openNewTabDropdown | Open the dropdown menu. | | | |
+| openSettings | Open the settings file. | | | |
+| paste | Insert the content that was copied onto the clipboard. | | | |
+| prevTab | Open the tab to the left of the current one. | | | |
+| resetFontSize | Reset the text size to the default value. | | | |
+| resizePane | Change the size of the active pane. | `direction`* | `left`, `right`, `up`, `down` | Direction in which the pane will be resized. |
+| scrollDown | Move the screen down. | | | |
+| scrollUp | Move the screen up. | | | |
+| scrollUpPage | Move the screen up a whole page. | | | |
+| scrollDownPage | Move the screen down a whole page. | | | |
+| splitPane | Halve the size of the active pane and open another. Without any arguments, this will open the default profile in the new pane. | 1. `split`*<br>2. `commandLine`<br>3. `startingDirectory`<br>4. `tabTitle`<br>5. `index`<br>6. `profile` | 1. `vertical`, `horizontal`, `auto`<br>2. string<br>3. string<br>4. string<br>5. integer<br>6. string | 1. How the pane will split. `auto` will split in the direction that provides the most surface area.<br>2. Executable run within the pane.<br>3. Directory in which the pane will open.<br>4. Title of the tab when the new pane is focused.<br>5. Profile that will open based on its position in the dropdown (starting at 0).<br>6. Profile that will open based on its GUID or name. |
+| switchToTab | Open a specific tab depending on index. | `index`* | integer | Tab that will open based on its position in the tab bar (starting at 0). |
+| toggleFullscreen | Switch between fullscreen and default window sizes. | | | |
+| unbound | Unbind the associated keys from any command. | | | |
+
+### Accepted Modifiers and Keys
+
+#### Modifiers
+`Ctrl+`, `Shift+`, `Alt+`
+
+#### Keys
+| Type | Keys |
+| ---- | ---- |
+| Function and Alphanumeric Keys | `f1-f24`, `a-z`, `0-9` |
+| Symbols | ``` ` ```, `-`, `=`, `[`, `]`, `\`, `;`, `'`, `,`, `.`, `/` |
+| Arrow Keys | `down`, `left`, `right`, `up`, `pagedown`, `pageup`, `pgdn`, `pgup`, `end`, `home`, `plus` |
+| Action Keys | `tab`, `enter`, `esc`, `escape`, `space`, `backspace`, `delete`, `insert` |
+| Numpad Keys | `numpad_0-numpad_9`, `numpad0-numpad9`, `numpad_add`, `numpad_plus`, `numpad_decimal`, `numpad_period`, `numpad_divide`, `numpad_minus`, `numpad_subtract`, `numpad_multiply` |
+
+## Background Images and Icons
+Some Terminal settings allow you to specify custom background images and icons. It is recommended that custom images and icons are stored in system-provided folders and are referred to using the correct [URI Schemes](https://docs.microsoft.com/en-us/windows/uwp/app-resources/uri-schemes). URI Schemes provide a way to reference files independent of their physical paths (which may change in the future).
+
+The most useful URI schemes to remember when customizing background images and icons are:
+
+| URI Scheme | Corresponding Physical Path | Use / description |
+| --- | --- | ---|
+| `ms-appdata:///Local/` | `%localappdata%\Packages\Microsoft.WindowsTerminal_8wekyb3d8bbwe\LocalState\` | Per-machine files |
+| `ms-appdata:///Roaming/` | `%localappdata%\Packages\Microsoft.WindowsTerminal_8wekyb3d8bbwe\RoamingState\` | Common files |
+
+> ⚠ Note: Do not rely on file references using the `ms-appx` URI Scheme (i.e. icons). These files are considered an internal implementation detail and may change name/location or may be omitted in the future.
+
+### Icons
+Terminal displays icons for each of your profiles which Terminal generates for any built-in shells - PowerShell Core, PowerShell, and any installed Linux/WSL distros. Each profile refers to a stock icon via the `ms-appx` URI Scheme.
+
+> ⚠ Note: Do not rely on the files referenced by the `ms-appx` URI Scheme - they are considered an internal implementation detail and may change name/location or may be omitted in the future.
+
+You can refer to you own icons if you wish, e.g.:
+
+```json
+    "icon" : "C:\\Users\\richturn\\OneDrive\\WindowsTerminal\\icon-ubuntu-32.png",
+```
+
+> 👉 Tip: Icons should be sized to 32x32px in an appropriate raster image format (e.g. .PNG, .GIF, or .ICO) to avoid having to scale your icons during runtime (causing a noticeable delay and loss of quality.)
+
+### Custom Background Images
+You can apply a background image to each of your profiles, allowing you to configure/brand/style each of your profiles independently from one another if you wish.
+
+To do so, specify your preferred `backgroundImage`, position it using `backgroundImageAlignment`, set its opacity with `backgroundImageOpacity`, and/or specify how your image fill the available space using `backgroundImageStretchMode`.
+
+For example:
+```json
+    "backgroundImage": "C:\\Users\\richturn\\OneDrive\\WindowsTerminal\\bg-ubuntu-256.png",
+    "backgroundImageAlignment": "bottomRight",
+    "backgroundImageOpacity": 0.1,
+    "backgroundImageStretchMode": "none"
+```
+
+> 👉 Tip: You can easily roam your collection of images and icons across all your machines by storing your icons and images in OneDrive (as shown above).
+
+With these settings, your Terminal's Ubuntu profile would look similar to this:
+
+![Custom icon and background image](../images/custom-icon-and-background-image.jpg)

doc/cascadia/TerminalSettings-spec.md

@@ -33,8 +33,8 @@ This spec will outline how various terminal frontends will be able to interact w
 5. Visual Studio should be able to persist and edit settings globally, without
    the need for a globals/profiles structure.
 6. The Terminal should be able to read information from a settings structure
-   that's independant of how it's persisted / implemented by the Application
-7. The Component should be able to have it's own settings independent of the
+   that's independent of how it's persisted / implemented by the Application
+7. The Component should be able to have its own settings independent of the
    application that's embedding it, such as font size and face, scrollbar
    visibility, etc. These should be settings that are specific to the component,
    and the Terminal should logically be unaffected by these settings.
@@ -57,7 +57,7 @@ VS needs to be able to persist settings just as a simple set of global settings.
 
 When the application needs to retrieve these settings, they need to use them as a tripartite structure: frontend-component-terminal settings.
 
-Each frontend will have it's own set of settings.
+Each frontend will have its own set of settings.
 Each component implementation will also ned to have some settings that control it.
 The terminal also will have some settings specific to the terminal.
 
@@ -79,7 +79,7 @@ Shell Commandline       |
 
 ### Simple Settings
 
-An application like VS might not even care about settings profiles. They should be able to persist the settings as just a singular entity, and change those as needed, without the additional overhead. Profiles will be something that's more specifc to Project Cascadia.
+An application like VS might not even care about settings profiles. They should be able to persist the settings as just a singular entity, and change those as needed, without the additional overhead. Profiles will be something that's more specific to Project Cascadia.
 
 ### Interface Descriptions
 
@@ -106,13 +106,13 @@ public interface IApplicationSettings
 }
 ```
 
-The Application can store whatever settings it wants in it's implementation of `IApplicationSettings`. When it instantiates a Terminal Component, it will pass it's `IComponentSettings` to it.
+The Application can store whatever settings it wants in its implementation of `IApplicationSettings`. When it instantiates a Terminal Component, it will pass its `IComponentSettings` to it.
 
 The component will retrieve whatever settings it wants from that object, and then pass the `TerminalSettings` to the Terminal it creates.
 
 
-The frontend will be able to get/set it's settings from the `IApplicationSettings` implementation.
-The frontend will be able to create components using the `IComponentSettings` in it's `IApplicationSettings`.
+The frontend will be able to get/set its settings from the `IApplicationSettings` implementation.
+The frontend will be able to create components using the `IComponentSettings` in its `IApplicationSettings`.
 The Component will then create the Terminal using the `TerminalSettings`.
 
 #### Project Cascadia Settings Details
@@ -228,6 +228,6 @@ I don't like that - if we change the font size, we should just recalculate how m
 ## Questions / TODO
 * How does this interplay with setting properties of the terminal component in XAML?
     * I would think that the component would load the XAML properties first, and if the controlling application calls `UpdateSettings` on the component, then those in-XAML properties would likely get overwritten.
-    * It's not necessary to create the component with a `IComponentSettings`, nor is it necessary to call `UpdateSettings`. If you wanted to create a trivial settings-less terminal component entriely in XAML, go right ahead.
+    * It's not necessary to create the component with a `IComponentSettings`, nor is it necessary to call `UpdateSettings`. If you wanted to create a trivial settings-less terminal component entirely in XAML, go right ahead.
     * Any settings that *are* exposed through XAML properties *should* also be exposed in the component's settings implementation as well.
         * Can that be enforced any way? I doubt it.

doc/cascadia/Unittesting-CppWinRT-Xaml.md

@@ -0,0 +1,527 @@
+# Getting TAEF unittests to work with a C++/WinRT XAML Islands application
+
+* __Author__: Mike Griese @zadjii-msft
+* __Created on__: 2019-06-06
+
+
+So you've built a Win32 application that uses XAML Islands to display it's UI
+with C++/WinRT. How do you go about adding unittests to this application? I'm
+going to cover the steps that I took to get the Windows Terminal updated to be
+able to test not only our C++/WinRT components, but also pure c++ classes that
+were used in the application, and components that used XAML UI elements.
+
+## Prerequisites
+
+Make sure you're using at least the 2.0.190605.7 version of the CppWinRT nuget
+package. Prior to this version, there are some bugs with C++/WinRT's detection
+of static lib dependencies. You might be able to get your build working with
+Visual Studio on earlier versions, but not straight from MsBuild.
+
+Also, if you're going to be running your tests in a CI build of some sort, make
+sure that your tests are running on a machine running at least Windows 18362. If
+your CI isn't running that version, then this doesn't matter at all.
+
+Furthermore, you may need an updated TAEF package as well. Our CI uses the TAEF
+VsTest adapter to allow ADO to run TAEF tests in CI. However, there's a bug in
+the tests adapter that prevents it from running tests in a UAP context. The
+`10.38.190605002` TAEF is the most recent release at the time of writing,
+however, that doesn't have the fix necessary. Fortunately, the TAEF team was
+kind enough to prototype a fix for us, which is the version
+`10.38.190610001-uapadmin`, which we're using in this repo until an official
+release with the fix is available.
+
+## Move the C++/WinRT implementation to a static lib
+
+By default, most (newly authored) C++/WinRT components are authored as a dll
+that can be used to activate your types. However, you might have other classes
+in that binary that you want to be able to test, which aren't winrt types. If
+the implementation is sitting in a DLL, it'll be hard to write a TAEF unittest
+dll that can call the pure c++ types you've defined.
+
+The first thing you're going to need to do is move the implementation of your
+winrt component from a dll to a static lib. Once you have the static lib, we'll
+be able to link it into the dll you were previously producing, as well as being
+able to link it into the dll we'll be using to test the types. Once this is
+complete, your dll project will exist as little more than some extra packaging
+for your new lib, as all your code will be built by the lib.
+
+To aid in this description, I'll be referring to the projects that we changed.
+The dll project we changed to a lib was the `TerminalApp` project. From it, we
+created a new `TerminalAppLib` project, and changed `TerminalApp` to create a
+dll by linking the lib `TerminalAppLib` produced.
+
+### Create the static lib project
+
+We'll start by creating a new static lib project. The easiest way to do this is
+by copying your existing dll `vcxproj` file into a new file. Make sure to change
+the `ProjectGuid` and to add the new project to your `.sln` file. Then, change
+the `ConfigurationType` to `StaticLibrary`. This Lib should be responsible for
+building all of your headers, `.cpp` files, `.idl`s for your winrt types, and
+any `.xaml` files you might have.
+
+You'll likely need to place this new file into a separate directory from the
+existing dll project, as C++/WinRT uses the project directory as the root of the
+intermediate build tree. Each directory should only have one `.vcxproj` file in
+it. For the Terminal project, we created a subdirectory `lib/` underneath
+`TerminalApp/`, and updated the `Include` paths to properly point at the
+original files. You could alternatively put all the source in one directory, and
+have separate `dll/` and `lib/` subdirectories from the source that are solely
+responsible for building their binary.
+
+At this point, you might face some difficulty including the right winmd
+references, especially from other C++/WinRT dependencies for this project that
+exist in your solution. I don't know why, but I had a fair amount of difficulty
+using a `ProjectReference` from a C++/WinRT StaticLibrary to another C++/WinRT
+project in my solution. If you're referring to any other projects, you'll need
+to set up a reference to their built `.winmd`'s manually.
+
+As an example, here's how we've added a reference to the `TerminalSettings`
+project from our `TerminalAppLib` project:
+
+```xml
+  <ItemGroup>
+    <!-- Manually add references to each of our dependent winmds. Mark them as
+    private=false and CopyLocalSatelliteAssemblies=false, so that we don't
+    propagate them upwards (which can make referencing this project result in
+    duplicate type definitions)-->
+
+    <Reference Include="Microsoft.Terminal.Settings">
+      <HintPath>$(SolutionDir)$(Platform)\$(Configuration)\TerminalSettings\Microsoft.Terminal.Settings.winmd</HintPath>
+      <IsWinMDFile>true</IsWinMDFile>
+      <Private>false</Private>
+      <CopyLocalSatelliteAssemblies>false</CopyLocalSatelliteAssemblies>
+    </Reference>
+  </ItemGroup>
+```
+
+The `HintPath` may be different depending on your project structure - verify
+locally the right path to the `.winmd` file you're looking for.
+
+Notably, you'll also need to put a `pch.h` and `pch.cpp` in the new lib's
+directory, and use them instead of the `pch.h` used by the dll. C++/WinRT will be
+very angry with you if you try to use a `pch.h` in another directory. Since
+we're putting all the code into the static lib project, take your existing
+`pch.h` and move it to the lib project's directory and create an empty `pch.h`
+in the dll project's directory.
+
+### Update the dll project
+
+Now that we have a lib that builds all your code, we can go ahead and tear out
+most of the dead code from the old dll project. Remove all the source files from
+the dll's `.vcxproj` file, save for the `pch.h` and `pch.cpp` files. You _may_
+need to leave the headers for any C++/WinRT types you've authored in this project
+- I'm not totally sure it's necessary.
+
+Now, to link the static lib we've created. For whatever reason, adding a
+`ProjectReference` to the static lib doesn't work. So, we'll need to manually
+link the lib from the lib project. You can do that by adding the lib's output
+dir to your `AdditionalLibraryDirectories`, and adding the lib to your
+`AdditionalDependencies`, like so:
+
+```xml
+  <ItemDefinitionGroup>
+    <Link>
+      <!-- Manually link with the TerminalAppLib.lib we've built. -->
+      <AdditionalLibraryDirectories>$(SolutionDir)\$(Platform)\$(Configuration)\TerminalAppLib;%(AdditionalLibraryDirectories)</AdditionalLibraryDirectories>
+
+      <AdditionalDependencies>TerminalAppLib.lib;%(AdditionalDependencies)</AdditionalDependencies>
+    </Link>
+  </ItemDefinitionGroup>
+```
+
+We are NOT adding a reference to the static lib project's .winmd here. As of the
+2.0.190605.7 CppWinRT nuget package, this is enough for MsBuild and Visual
+Studio to be able to determine that the static lib's  `.winmd` should be included
+in this package.
+
+At this point, you might have some mdmerge errors, which complain about
+duplicate types in one of your dependencies. This might especially happen if one
+of your dependencies (ex `A.dll`) is also a dependency for one of your _other_
+dependencies (ex `B.dll`). In this example, your final output project `C.dll`
+depends on both `A.dll` and `B.dll`, and `B.dll` _also_ depends on `A.dll`. If
+you're seeing this, I recommend adding `Private=false` and
+`CopyLocalSatelliteAssemblies=false` to your dependent dlls. In this example,
+add similar code to `B.dll`:
+
+```xml
+    <ProjectReference Include="$(SolutionDir)src\cascadia\TerminalSettings\TerminalSettings.vcxproj">
+      <Private>false</Private>
+      <CopyLocalSatelliteAssemblies>false</CopyLocalSatelliteAssemblies>
+    </ProjectReference>
+```
+
+where `TerminalSettings` is your `A.dll`, which is included by both `B` and `C`.
+
+We additionally had an `.exe` project that was including our `TerminalApp`
+project, and all its `.xbf` and `.pri` files. If you have a similar project
+aggregating all your resources, you might need to update the paths to point to
+the new static lib project.
+
+At this point, you should be able to rebuild your solution, and everything
+should be working just the same as before.
+
+
+## Add TAEF Tests
+
+Now that you have a static library project, you can start building your unittest
+dll. Start by creating a new directory for your unittest code, and creating a
+`.vcxproj` for a TAEF unittest dll. For the Terminal solution, we use the TAEF
+nuget package `Taef.Redist.Wlk`.
+
+### Referencing your C++/WinRT static lib
+
+This step is the easiest. Add a `ProjectReference` to your static lib project,
+and your lib will be linked into your unittest dll.
+
+```xml
+    <ProjectReference Include="$(SolutionDir)\src\cascadia\TerminalApp\lib\TerminalAppLib.vcxproj" />
+```
+
+Congratulations, you can now instantiate the pure c++ types you've authored in
+your static lib. But what if you want to test your C++/WinRT types too?
+
+### Using your C++/WinRT types
+
+To be able to instantiate your C++/WinRT types in a TAEF unittest, you'll need
+to rely on a new feature to Windows in version 1903 which enables unpackaged
+activation of WinRT types. To do this, we'll need to author a SxS manifest that
+lists each of our types, and include it in the dll, and also activate it
+manually from TAEF.
+
+#### Creating the manifest
+
+First, you need to create a manifest file that lists each dll your test depends
+upon, and each of the types in that dll. For example, here's an excerpt from the
+Terminal's manifest:
+
+```xml
+<?xml version="1.0" encoding="utf-8" standalone="yes"?>
+<assembly xmlns="urn:schemas-microsoft-com:asm.v1" manifestVersion="1.0">
+  <file name="TerminalSettings.dll" hashalg="SHA1">
+    <activatableClass name="Microsoft.Terminal.Settings.KeyChord" threadingModel="both" xmlns="urn:schemas-microsoft-com:winrt.v1"></activatableClass>
+    <activatableClass name="Microsoft.Terminal.Settings.TerminalSettings" threadingModel="both" xmlns="urn:schemas-microsoft-com:winrt.v1"></activatableClass>
+  </file>
+  <file name="TerminalApp.dll" hashalg="SHA1">
+    <activatableClass name="TerminalApp.App" threadingModel="both" xmlns="urn:schemas-microsoft-com:winrt.v1"></activatableClass>
+    <activatableClass name="TerminalApp.AppKeyBindings" threadingModel="both" xmlns="urn:schemas-microsoft-com:winrt.v1"></activatableClass>
+    <activatableClass name="TerminalApp.XamlmetaDataProvider" threadingModel="both" xmlns="urn:schemas-microsoft-com:winrt.v1"></activatableClass>
+  </file>
+</assembly>
+```
+
+Here we have two dlls that we depend upon, `TerminalSettings.dll` and
+`TerminalApp.dll`. `TerminalSettings` implements two types,
+`Microsoft.Terminal.Settings.KeyChord` and
+`Microsoft.Terminal.Settings.TerminalSettings`.
+
+#### Linking the manifest to the test dll
+
+Now that we have a manifest file, we need to embed it in your unittest dll. This
+is done with the following properties in your `vcxproj` file:
+
+```xml
+  <PropertyGroup>
+    <GenerateManifest>true</GenerateManifest>
+    <EmbedManifest>true</EmbedManifest>
+  </PropertyGroup>
+  <ItemGroup>
+    <Manifest Include="TerminalApp.Unit.Tests.manifest" />
+  </ItemGroup>
+```
+where `TerminalApp.Unit.Tests.manifest` is the name of your manifest file.
+
+Additionally, you'll need to binplace the manifest _adjacent to your test
+binary_, so TAEF can find it at runtime. I've done this in the following way,
+though I'm sure there's a better way:
+
+```xml
+  <ItemDefinitionGroup>
+    <PostBuildEvent>
+      <!-- Manually copy the manifest to our outdir, because the test will need
+      to find it adjacent to us. -->
+      <Command>
+        (xcopy /Y &quot;$(OpenConsoleDir)src\cascadia\ut_app\TerminalApp.Unit.Tests.manifest&quot; &quot;$(OutDir)\TerminalApp.Unit.Tests.manifest*&quot; )
+      </Command>
+    </PostBuildEvent>
+  </ItemDefinitionGroup>
+```
+
+#### Copying your dependencies
+
+Additionally, any dlls that implement any types your test is dependent upon will
+also need to be in the output directory for the test. Manually copy those DLLs
+to the tests' output directory too. The updated `PostBuildEvent` looks like
+this:
+
+```xml
+  <ItemDefinitionGroup>
+    <PostBuildEvent>
+      <Command>
+        echo OutDir=$(OutDir)
+        (xcopy /Y &quot;$(SolutionDir)src\cascadia\ut_app\TerminalApp.Unit.Tests.manifest&quot; &quot;$(OutDir)\TerminalApp.Unit.Tests.manifest*&quot; )
+
+        (xcopy /Y &quot;$(SolutionDir)$(Platform)\$(Configuration)\TerminalConnection\TerminalConnection.dll&quot; &quot;$(OutDir)\TerminalConnection.dll*&quot; )
+        (xcopy /Y &quot;$(SolutionDir)$(Platform)\$(Configuration)\TerminalSettings\TerminalSettings.dll&quot; &quot;$(OutDir)\TerminalSettings.dll*&quot; )
+        (xcopy /Y &quot;$(SolutionDir)$(Platform)\$(Configuration)\TerminalControl\TerminalControl.dll&quot; &quot;$(OutDir)\TerminalControl.dll*&quot; )
+      </Command>
+    </PostBuildEvent>
+  </ItemDefinitionGroup>
+```
+
+Again, verify the correct paths to your dependant C++/WinRT dlls, as they may be
+different than the above
+
+#### Activating the manifest from TAEF
+
+Now that the manifest lives adjacent to your test dll, and all your dependent
+dlls are also adjacent to the unittest dll, there's only one thing left to do.
+TAEF will not use your dll's manifest by default, so you'll need to add a
+property to your test class/method to tell TAEF to do so. You can do this with
+the following:
+
+```c++
+class SettingsTests
+{
+    // Use a custom manifest to ensure that we can activate winrt types from
+    // our test. This property will tell taef to manually use this as the
+    // sxs manifest during this test class. It includes all the C++/WinRT
+    // types we've defined, so if your test is crashing for an unknown
+    // reason, make sure it's included in that file.
+    BEGIN_TEST_CLASS(SettingsTests)
+        TEST_CLASS_PROPERTY(L"ActivationContext", L"TerminalApp.Unit.Tests.manifest")
+    END_TEST_CLASS()
+
+    // Other Test code here
+}
+```
+
+Now, if you try to add any test methods that instantiate WinRT types you've
+authored, they'll work. That is of course, so long as they don't use XAML. If
+you want to use any XAML types, then you'll have to keep reading.
+
+### Using Xaml Types (with XAML Islands)
+
+To be able to instantiate XAML types in your unittest, we'll need to make use of
+the [XAML Hosting
+API](https://docs.microsoft.com/en-us/windows/apps/desktop/modernize/using-the-xaml-hosting-api)
+(Xaml Islands). This enables you to use XAML APIs from a Win32 context.
+
+
+#### Adding XAML Hosting code
+
+First and foremost, you'll need to add the following to your test's `precomp.h`:
+
+```c++
+#include <winrt/Windows.system.h>
+#include <winrt/Windows.Foundation.Collections.h>
+#include <winrt/Windows.UI.Xaml.Hosting.h>
+#include <windows.ui.xaml.hosting.desktopwindowxamlsource.h>
+```
+
+If you hit a compile warning that refers to `GetCurrentTime`, you'll probably
+also need the following, after you've `#include`'d `Windows.h`:
+
+```c++
+#ifdef GetCurrentTime
+#undef GetCurrentTime
+#endif
+```
+
+Then, somewhere in your test code, you'll need to start up Xaml Islands. I've done this in my `TEST_CLASS_SETUP`, so that I only create it once, and re-use it for each method.
+
+```c++
+
+class TabTests
+{
+    TEST_CLASS_SETUP(ClassSetup)
+    {
+        winrt::init_apartment(winrt::apartment_type::single_threaded);
+        // Initialize the Xaml Hosting Manager
+        _manager = winrt::Windows::UI::Xaml::Hosting::WindowsXamlManager::InitializeForCurrentThread();
+        _source = winrt::Windows::UI::Xaml::Hosting::DesktopWindowXamlSource{};
+
+        return true;
+    }
+
+private:
+    winrt::Windows::UI::Xaml::Hosting::WindowsXamlManager _manager{ nullptr };
+    winrt::Windows::UI::Xaml::Hosting::DesktopWindowXamlSource _source{ nullptr };
+```
+
+#### Authoring your test's `AppxManifest.xml`
+
+This alone however is not enough to get XAML Islands to work. There was a fairly
+substantial change to the XAML Hosting API around Windows build 18295, so it
+explicitly requires that you have your executable's manifest set
+`maxversiontested` to higher than that version. However, because TAEF's `te.exe`
+is not so manifested, we can't just use our SxS manifest from before to set that
+version. Instead, you'll need to make TAEF run your test binary in a packaged
+content, with our own appxmanifest.
+
+To do this, we'll need to author an `Appxmanifest.xml` to use with the test, and
+associate that manifest with the test.
+
+Here's the AppxManifest we're using:
+
+```xml
+<?xml version="1.0" encoding="utf-8"?>
+<Package xmlns:rescap="http://schemas.microsoft.com/appx/manifest/foundation/windows10/restrictedcapabilities" xmlns="http://schemas.microsoft.com/appx/manifest/foundation/windows10" xmlns:uap="http://schemas.microsoft.com/appx/manifest/uap/windows10" IgnorableNamespaces="uap">
+
+  <Identity Name="TerminalApp.Unit.Tests.Package"
+          ProcessorArchitecture="neutral"
+          Publisher="CN=Microsoft Corporation, O=Microsoft Corporation, L=Redmond, S=Washington, C=US"
+          Version="1.0.0.0"
+          ResourceId="en-us" />
+  <Properties>
+    <DisplayName>TerminalApp.Unit.Tests.Package Host Process</DisplayName>
+    <PublisherDisplayName>Microsoft Corp.</PublisherDisplayName>
+    <Logo>taef.png</Logo>
+    <Description>TAEF Packaged Cwa FullTrust Application Host Process</Description>
+  </Properties>
+
+  <Dependencies>
+    <TargetDeviceFamily Name="Windows.Universal" MinVersion="10.0.18362.0" MaxVersionTested="10.0.18362.0" />
+    <PackageDependency Name="Microsoft.VCLibs.140.00.Debug" MinVersion="14.0.27023.1" Publisher="CN=Microsoft Corporation, O=Microsoft Corporation, L=Redmond, S=Washington, C=US" />
+    <PackageDependency Name="Microsoft.VCLibs.140.00.Debug.UWPDesktop" MinVersion="14.0.27027.1" Publisher="CN=Microsoft Corporation, O=Microsoft Corporation, L=Redmond, S=Washington, C=US" />
+  </Dependencies>
+
+  <Resources>
+    <Resource Language="en-us" />
+  </Resources>
+
+  <Applications>
+    <Application Id="TE.ProcessHost" Executable="TE.ProcessHost.exe" EntryPoint="Windows.FullTrustApplication">
+      <uap:VisualElements DisplayName="TAEF Packaged Cwa FullTrust Application Host Process" Square150x150Logo="taef.png" Square44x44Logo="taef.png" Description="TAEF Packaged Cwa Application Host Process" BackgroundColor="#222222">
+        <uap:SplashScreen Image="taef.png" />
+      </uap:VisualElements>
+    </Application>
+  </Applications>
+
+  <Capabilities>
+    <rescap:Capability Name="runFullTrust"/>
+  </Capabilities>
+
+  <Extensions>
+    <Extension Category="windows.activatableClass.inProcessServer">
+      <InProcessServer>
+        <Path>TerminalSettings.dll</Path>
+        <ActivatableClass ActivatableClassId="Microsoft.Terminal.Settings.TerminalSettings" ThreadingModel="both" />
+        <ActivatableClass ActivatableClassId="Microsoft.Terminal.Settings.KeyChord" ThreadingModel="both" />
+      </InProcessServer>
+    </Extension>
+    <!-- More extensions here -->
+  </Extensions>
+</Package>
+```
+
+Change the `Identity.Name` and `Properties.DisplayName` to be more appropriate
+for your test, as well as other properties if you feel the need. TAEF will
+deploy the test package and remove it from your machine during testing, so it
+doesn't terribly matter what these values are.
+
+MAKE SURE that `MaxVersionTested` is higher than `10.0.18295.0`. If it isn't,
+XAML islands will still prevent you from activating it.
+
+UNDER NO CIRCUMSTANCE should you change the `<Application Id="TE.ProcessHost"
+Executable="TE.ProcessHost.exe" EntryPoint="Windows.FullTrustApplication">`
+line. This is how TAEF activates the TAEF host for your test binary. You might
+get a warning about `TE.ProcessHost.exe` being deprecated in favor of
+`TE.ProcessHost.UAP.exe`, but I haven't had success with the UAP version.
+
+Lower in the file, you'll see the `Extensions` block. In here you'll put each of
+the winrt dependencies that your test needs, much like we did for the previous
+manifest. Note that the syntax is _not_ exactly the same as the SxS manifest.
+
+#### Copy the AppxManifest to your `$(OutDir)`
+
+Again, we'll need to copy this appxmanifest adjacent to the test binary so we
+can load it from the test. We'll do this similar to how we did the SxS manifest
+before. The complete `PostBuildEvent` now looks like this:
+
+```xml
+  <ItemDefinitionGroup>
+    <PostBuildEvent>
+      <Command>
+        (xcopy /Y &quot;$(SolutionDir)src\cascadia\ut_app\TerminalApp.Unit.Tests.manifest&quot; &quot;$(OutDir)\TerminalApp.Unit.Tests.manifest*&quot; )
+
+        (xcopy /Y &quot;$(SolutionDir)src\cascadia\ut_app\TerminalApp.Unit.Tests.AppxManifest.xml&quot; &quot;$(OutDir)\TerminalApp.Unit.Tests.AppxManifest.xml*&quot; )
+
+        (xcopy /Y &quot;$(SolutionDir)$(Platform)\$(Configuration)\TerminalConnection\TerminalConnection.dll&quot; &quot;$(OutDir)\TerminalConnection.dll*&quot; )
+        (xcopy /Y &quot;$(SolutionDir)$(Platform)\$(Configuration)\TerminalSettings\TerminalSettings.dll&quot; &quot;$(OutDir)\TerminalSettings.dll*&quot; )
+        (xcopy /Y &quot;$(SolutionDir)$(Platform)\$(Configuration)\TerminalControl\TerminalControl.dll&quot; &quot;$(OutDir)\TerminalControl.dll*&quot; )
+      </Command>
+    </PostBuildEvent>
+  </ItemDefinitionGroup>
+```
+
+The new line here is the line referencing
+`TerminalApp.Unit.Tests.AppxManifest.xml`. You can only have one
+`PostBuildEvent` per project, so don't go re-defining it for each additional
+step - MsBuild will only use the last one. Again, this is probably not the best
+way of copying these files over, but it works.
+
+#### Use the AppxManifest in the test code
+
+Now that we have the AppxManifest being binplaced next to our test, we can
+finally reference it in the test. Instead of using the `ActivationContext` from
+before, we'll use two new properties to tell TAEF to run this test as a package,
+and to use our manifest as the AppxManifest for the package.
+
+```c++
+    BEGIN_TEST_CLASS(TabTests)
+        TEST_CLASS_PROPERTY(L"RunAs", L"UAP")
+        TEST_CLASS_PROPERTY(L"UAP:AppXManifest", L"TerminalApp.Unit.Tests.AppxManifest.xml")
+    END_TEST_CLASS()
+```
+
+The complete Xaml Hosting test now looks like this:
+
+```c++
+    class TabTests
+    {
+        BEGIN_TEST_CLASS(TabTests)
+            TEST_CLASS_PROPERTY(L"RunAs", L"UAP")
+            TEST_CLASS_PROPERTY(L"UAP:AppXManifest", L"TerminalApp.Unit.Tests.AppxManifest.xml")
+        END_TEST_CLASS()
+
+        TEST_METHOD(TryCreateXamlObjects);
+
+        TEST_CLASS_SETUP(ClassSetup)
+        {
+            winrt::init_apartment(winrt::apartment_type::single_threaded);
+            // Initialize the Xaml Hosting Manager
+            _manager = winrt::Windows::UI::Xaml::Hosting::WindowsXamlManager::InitializeForCurrentThread();
+            _source = winrt::Windows::UI::Xaml::Hosting::DesktopWindowXamlSource{};
+
+            return true;
+        }
+
+    private:
+        winrt::Windows::UI::Xaml::Hosting::WindowsXamlManager _manager{ nullptr };
+        winrt::Windows::UI::Xaml::Hosting::DesktopWindowXamlSource _source{ nullptr };
+    };
+    void TabTests::TryCreateXamlObjects(){ ... }
+```
+
+Congratulations, you can now use XAML types from your unittest.
+
+### Using types from `Microsoft.UI.Xaml`
+
+Let's say you're extra crazy and you're using the `Microsoft.UI.Xaml` nuget
+package. If you've followed all the steps above exactly, you're probably already
+fine! You've already put the types in your appxmanifest (there are a lot of
+them). You should be able to call the `Microsoft.UI.Xaml` types without any
+problems.
+
+This is because of a few key lines we already put in the appxmanifest:
+
+```xml
+  <Dependencies>
+    <TargetDeviceFamily Name="Windows.Universal" MinVersion="10.0.18362.0" MaxVersionTested="10.0.18362.0" />
+    <PackageDependency Name="Microsoft.VCLibs.140.00.Debug" MinVersion="14.0.27023.1" Publisher="CN=Microsoft Corporation, O=Microsoft Corporation, L=Redmond, S=Washington, C=US" />
+    <PackageDependency Name="Microsoft.VCLibs.140.00.Debug.UWPDesktop" MinVersion="14.0.27027.1" Publisher="CN=Microsoft Corporation, O=Microsoft Corporation, L=Redmond, S=Washington, C=US" />
+  </Dependencies>
+```
+
+Without these `PackageDependency` entries for the VCLibs, Microsoft.UI.Xaml.dll
+will not be able to load.

doc/cascadia/profiles.schema.json

@@ -0,0 +1,809 @@
+{
+  "$id": "https://github.com/microsoft/terminal/blob/master/doc/cascadia/profiles.schema.json",
+  "$schema": "http://json-schema.org/draft-07/schema#",
+  "title": "Microsoft's Windows Terminal Settings Profile Schema'",
+  "definitions": {
+    "Color": {
+      "default": "#",
+      "pattern": "^#([A-Fa-f0-9]{6}|[A-Fa-f0-9]{3})$",
+      "type": "string",
+      "format": "color"
+    },
+    "Coordinates": {
+      "pattern": "^(-?\\d+)?(,\\s?(-?\\d+)?)?$",
+      "type": "string"
+    },
+    "ProfileGuid": {
+      "default": "{}",
+      "pattern": "^\\{[a-fA-F0-9]{8}-[a-fA-F0-9]{4}-[a-fA-F0-9]{4}-[a-fA-F0-9]{4}-[a-fA-F0-9]{12}\\}$",
+      "type": "string"
+    },
+    "ShortcutActionName": {
+      "enum": [
+        "closePane",
+        "closeTab",
+        "closeWindow",
+        "copy",
+        "copyTextWithoutNewlines",
+        "decreaseFontSize",
+        "duplicateTab",
+        "increaseFontSize",
+        "moveFocus",
+        "moveFocusDown",
+        "moveFocusLeft",
+        "moveFocusRight",
+        "moveFocusUp",
+        "newTab",
+        "newTabProfile0",
+        "newTabProfile1",
+        "newTabProfile2",
+        "newTabProfile3",
+        "newTabProfile4",
+        "newTabProfile5",
+        "newTabProfile6",
+        "newTabProfile7",
+        "newTabProfile8",
+        "nextTab",
+        "openNewTabDropdown",
+        "openSettings",
+        "paste",
+        "prevTab",
+        "resetFontSize",
+        "resizePane",
+        "resizePaneDown",
+        "resizePaneLeft",
+        "resizePaneRight",
+        "resizePaneUp",
+        "scrollDown",
+        "scrollDownPage",
+        "scrollUp",
+        "scrollUpPage",
+        "splitHorizontal",
+        "splitVertical",
+        "splitPane",
+        "switchToTab",
+        "switchToTab0",
+        "switchToTab1",
+        "switchToTab2",
+        "switchToTab3",
+        "switchToTab4",
+        "switchToTab5",
+        "switchToTab6",
+        "switchToTab7",
+        "switchToTab8",
+        "toggleFullscreen",
+        "find"
+      ],
+      "type": "string"
+    },
+    "Direction": {
+      "enum": [
+        "left",
+        "right",
+        "up",
+        "down"
+      ],
+      "type": "string"
+    },
+    "SplitState": {
+      "enum": [
+        "vertical",
+        "horizontal",
+        "auto"
+      ],
+      "type": "string"
+    },
+    "NewTerminalArgs": {
+      "properties": {
+        "commandline": {
+          "description": "A commandline to use instead of the profile's",
+          "type": "string"
+        },
+        "tabTitle": {
+          "description": "An initial tabTitle to use instead of the profile's",
+          "type": "string"
+        },
+        "startingDirectory": {
+          "description": "A startingDirectory to use instead of the profile's",
+          "type": "string"
+        },
+        "profile": {
+          "description": "Either the GUID or name of a profile to use, instead of launching the default",
+          "type": "string"
+        },
+        "index": {
+          "type": "integer",
+          "description": "The index of the profile in the new tab dropdown to open"
+        }
+      },
+      "type": "object"
+    },
+    "ShortcutAction": {
+      "properties": {
+        "action": {
+          "description": "The action to execute",
+          "$ref": "#/definitions/ShortcutActionName"
+        }
+      },
+      "required": [
+        "action"
+      ],
+      "type": "object"
+    },
+    "CopyAction": {
+      "description": "Arguments corresponding to a Copy Text Action",
+      "allOf": [
+        { "$ref": "#/definitions/ShortcutAction" },
+        {
+          "properties": {
+            "action": { "type": "string", "pattern": "copy" },
+            "trimWhitespace": {
+              "type": "boolean",
+              "default": true,
+              "description": "If true, whitespace is removed and newlines are maintained. If false, newlines are removed and whitespace is maintained."
+            }
+          }
+        }
+      ]
+    },
+    "NewTabAction": {
+      "description": "Arguments corresponding to a New Tab Action",
+      "allOf": [
+        { "$ref": "#/definitions/ShortcutAction" },
+        { "$ref": "#/definitions/NewTerminalArgs" },
+        {
+          "properties": {
+            "action": { "type":"string", "pattern": "newTab" }
+          }
+        }
+      ]
+    },
+    "SwitchToTabAction": {
+      "description": "Arguments corresponding to a Switch To Tab Action",
+      "allOf": [
+        { "$ref": "#/definitions/ShortcutAction" },
+        {
+          "properties": {
+            "action": { "type": "string", "pattern": "switchToTab" },
+            "index": {
+              "type": "integer",
+              "default": 0,
+              "description": "Which tab to switch to, with the first being 0"
+            }
+          }
+        }
+      ],
+      "required": [ "index" ]
+    },
+    "MoveFocusAction": {
+      "description": "Arguments corresponding to a Move Focus Action",
+      "allOf": [
+        { "$ref": "#/definitions/ShortcutAction" },
+        {
+          "properties": {
+            "action": { "type": "string", "pattern": "moveFocus" },
+            "direction": {
+              "$ref": "#/definitions/Direction",
+              "default": "left",
+              "description": "The direction to move focus in, between panes"
+            }
+          }
+        }
+      ],
+      "required": [ "direction" ]
+    },
+    "ResizePaneAction": {
+      "description": "Arguments corresponding to a Resize Pane Action",
+      "allOf": [
+        { "$ref": "#/definitions/ShortcutAction" },
+        {
+          "properties": {
+            "action": { "type": "string", "pattern": "resizePane" },
+            "direction": {
+              "$ref": "#/definitions/Direction",
+              "default": "left",
+              "description": "The direction to move the pane separator in"
+            }
+          }
+        }
+      ],
+      "required": [ "direction" ]
+    },
+    "SplitPaneAction": {
+      "description": "Arguments corresponding to a Split Pane Action",
+      "allOf": [
+        { "$ref": "#/definitions/ShortcutAction" },
+        { "$ref": "#/definitions/NewTerminalArgs" },
+        {
+          "properties": {
+            "action": { "type": "string", "pattern": "splitPane" },
+            "split": {
+              "$ref": "#/definitions/SplitState",
+              "default": "auto",
+              "description": "The orientation to split the pane in, either vertical (think [|]), horizontal (think [-]), or auto (splits pane based on remaining space)"
+            }
+          }
+        }
+      ],
+      "required": [ "split" ]
+    },
+    "Keybinding": {
+      "additionalProperties": false,
+      "properties": {
+        "command": {
+          "description": "The action executed when the associated key bindings are pressed.",
+            "oneOf": [
+              { "$ref": "#/definitions/CopyAction" },
+              { "$ref": "#/definitions/ShortcutActionName" },
+              { "$ref": "#/definitions/NewTabAction" },
+              { "$ref": "#/definitions/SwitchToTabAction" },
+              { "$ref": "#/definitions/MoveFocusAction" },
+              { "$ref": "#/definitions/ResizePaneAction" },
+              { "$ref": "#/definitions/SplitPaneAction" }
+            ]
+        },
+        "keys": {
+          "description": "Defines the key combinations used to call the command.",
+          "items": {
+            "pattern": "^(?<modifier>(ctrl|alt|shift)\\+?((ctrl|alt|shift)(?<!\\2)\\+?)?((ctrl|alt|shift)(?<!\\2|\\4))?\\+?)?(?<key>[^+\\s]+?)?(?<=[^+\\s])$",
+            "type": "string"
+          },
+          "minItems": 1,
+          "type": "array"
+        }
+      },
+      "required": [
+        "command",
+        "keys"
+      ],
+      "type": "object"
+    },
+    "Globals": {
+      "additionalProperties": true,
+      "description": "Properties that affect the entire window, regardless of the profile settings.",
+      "properties": {
+        "alwaysShowTabs": {
+          "default": true,
+          "description": "When set to true, tabs are always displayed. When set to false and showTabsInTitlebar is set to false, tabs only appear after opening a new tab.",
+          "type": "boolean"
+        },
+        "copyOnSelect": {
+          "default": false,
+          "description": "When set to true, a selection is immediately copied to your clipboard upon creation. When set to false, the selection persists and awaits further action.",
+          "type": "boolean"
+        },
+        "defaultProfile": {
+          "$ref": "#/definitions/ProfileGuid",
+          "description": "Sets the default profile. Opens by clicking the '+' icon or typing the key binding assigned to 'newTab'. The guid of the desired default profile is used as the value."
+        },
+        "initialCols": {
+          "default": 120,
+          "description": "The number of columns displayed in the window upon first load.",
+          "maximum": 999,
+          "minimum": 1,
+          "type": "integer"
+        },
+        "initialPosition": {
+          "$ref": "#/definitions/Coordinates",
+          "description": "The position of the top left corner of the window upon first load. On a system with multiple displays, these coordinates are relative to the top left of the primary display. If launchMode is set to maximized, the window will be maximized on the monitor specified by those coordinates."
+        },
+        "initialRows": {
+          "default": 30,
+          "description": "The number of rows displayed in the window upon first load.",
+          "maximum": 999,
+          "minimum": 1,
+          "type": "integer"
+        },
+        "launchMode": {
+          "default": "default",
+          "description": "Defines whether the Terminal will launch as maximized or not.",
+          "enum": [
+            "maximized",
+            "default"
+          ],
+          "type": "string"
+        },
+        "rowsToScroll": {
+          "default": "system",
+          "description": "The number of rows to scroll at a time with the mouse wheel. This will override the system setting if the value is not zero or 'system'.",
+          "maximum": 999,
+          "minimum": 0,
+          "type": ["integer", "string"]
+        },
+        "keybindings": {
+          "description": "Properties are specific to each custom key binding.",
+          "items": {
+            "$ref": "#/definitions/Keybinding"
+          },
+          "type": "array"
+        },
+        "requestedTheme": {
+          "default": "system",
+          "description": "Sets the theme of the application.",
+          "enum": [
+            "light",
+            "dark",
+            "system"
+          ],
+          "type": "string"
+        },
+        "showTabsInTitlebar": {
+          "default": true,
+          "description": "When set to true, the tabs are moved into the titlebar and the titlebar disappears. When set to false, the titlebar sits above the tabs.",
+          "type": "boolean"
+        },
+        "showTerminalTitleInTitlebar": {
+          "default": true,
+          "description": "When set to true, titlebar displays the title of the selected tab. When set to false, titlebar displays 'Windows Terminal'.",
+          "type": "boolean"
+        },
+        "snapToGridOnResize": {
+          "default": false,
+          "description": "When set to true, the window will snap to the nearest character boundary on resize. When false, the window will resize 'smoothly'",
+          "type": "boolean"
+        },
+        "tabWidthMode": {
+          "default": "equal",
+          "description": "Sets the width of the tabs.",
+          "enum": [
+            "equal",
+            "titleLength"
+          ],
+          "type": "string"
+        },
+        "wordDelimiters": {
+          "default": " ./\\()\"'-:,.;<>~!@#$%^&*|+=[]{}~?│",
+          "description": "Determines the delimiters used in a double click selection.",
+          "type": "string"
+        },
+        "confirmCloseAllTabs": {
+          "default": true,
+          "description": " When set to `true` closing a window with multiple tabs open WILL require confirmation.  When set to `false` closing a window with multiple tabs open WILL NOT require confirmation.",
+          "type":"boolean"
+        }
+      },
+      "required": [
+        "defaultProfile"
+      ],
+      "type": "object"
+    },
+    "Profile": {
+      "description": "Properties specific to a unique profile.",
+      "additionalProperties": false,
+      "properties": {
+        "acrylicOpacity": {
+          "default": 0.5,
+          "description": "When useAcrylic is set to true, it sets the transparency of the window for the profile. Accepts floating point values from 0-1 (default 0.5).",
+          "maximum": 1,
+          "minimum": 0,
+          "type": "number"
+        },
+        "background": {
+          "$ref": "#/definitions/Color",
+          "default": "#0c0c0c",
+          "description": "Sets the background color of the profile. Overrides background set in color scheme if colorscheme is set. Uses hex color format: \"#rrggbb\".",
+          "type": ["string", "null"]
+        },
+        "backgroundImage": {
+          "description": "Sets the file location of the Image to draw over the window background.",
+          "type": "string"
+        },
+        "backgroundImageAlignment": {
+          "default": "center",
+          "enum": [
+            "bottom",
+            "bottomLeft",
+            "bottomRight",
+            "center",
+            "left",
+            "right",
+            "top",
+            "topLeft",
+            "topRight"
+          ],
+          "type": "string"
+        },
+        "backgroundImageOpacity": {
+          "description": "(Not in SettingsSchema.md)",
+          "maximum": 1,
+          "minimum": 0,
+          "type": "number"
+        },
+        "backgroundImageStretchMode": {
+          "default": "uniformToFill",
+          "description": "Sets how the background image is resized to fill the window.",
+          "enum": [
+            "fill",
+            "none",
+            "uniform",
+            "uniformToFill"
+          ],
+          "type": "string"
+        },
+        "closeOnExit": {
+          "default": "graceful",
+          "description": "Sets how the profile reacts to termination or failure to launch. Possible values: \"graceful\" (close when exit is typed or the process exits normally), \"always\" (always close) and \"never\" (never close). true and false are accepted as synonyms for \"graceful\" and \"never\" respectively.",
+          "oneOf": [
+            {
+              "enum": [
+                "never",
+                "graceful",
+                "always"
+              ],
+              "type": "string"
+            },
+            {
+              "type": "boolean"
+            }
+          ]
+        },
+        "colorScheme": {
+          "default": "Campbell",
+          "description": "Name of the terminal color scheme to use. Color schemes are defined under \"schemes\".",
+          "type": "string"
+        },
+        "colorTable": {
+          "description": "Array of colors used in the profile if colorscheme is not set. Colors use hex color format: \"#rrggbb\". Ordering is as follows: [black, red, green, yellow, blue, magenta, cyan, white, bright black, bright red, bright green, bright yellow, bright blue, bright magenta, bright cyan, bright white]",
+          "items": {
+            "additionalProperties": false,
+            "properties": {
+              "background": {
+                "$ref": "#/definitions/Color",
+                "description": "Sets the background color of the color table."
+              },
+              "black": {
+                "$ref": "#/definitions/Color",
+                "description": "Sets the color used as ANSI black."
+              },
+              "blue": {
+                "$ref": "#/definitions/Color",
+                "description": "Sets the color used as ANSI blue."
+              },
+              "brightBlack": {
+                "$ref": "#/definitions/Color",
+                "description": "Sets the color used as ANSI bright black."
+              },
+              "brightBlue": {
+                "$ref": "#/definitions/Color",
+                "description": "Sets the color used as ANSI bright blue."
+              },
+              "brightCyan": {
+                "$ref": "#/definitions/Color",
+                "description": "Sets the color used as ANSI bright cyan."
+              },
+              "brightGreen": {
+                "$ref": "#/definitions/Color",
+                "description": "Sets the color used as ANSI bright green."
+              },
+              "brightPurple": {
+                "$ref": "#/definitions/Color",
+                "description": "Sets the color used as ANSI bright purple."
+              },
+              "brightRed": {
+                "$ref": "#/definitions/Color",
+                "description": "Sets the color used as ANSI bright red."
+              },
+              "brightWhite": {
+                "$ref": "#/definitions/Color",
+                "description": "Sets the color used as ANSI bright white."
+              },
+              "brightYellow": {
+                "$ref": "#/definitions/Color",
+                "description": "Sets the color used as ANSI bright yellow."
+              },
+              "cyan": {
+                "$ref": "#/definitions/Color",
+                "description": "Sets the color used as ANSI cyan."
+              },
+              "foreground": {
+                "$ref": "#/definitions/Color",
+                "description": "Sets the foreground color of the color table."
+              },
+              "green": {
+                "$ref": "#/definitions/Color",
+                "description": "Sets the color used as ANSI green."
+              },
+              "purple": {
+                "$ref": "#/definitions/Color",
+                "description": "Sets the color used as ANSI purple."
+              },
+              "red": {
+                "$ref": "#/definitions/Color",
+                "description": "Sets the color used as ANSI red."
+              },
+              "white": {
+                "$ref": "#/definitions/Color",
+                "description": "Sets the color used as ANSI white."
+              },
+              "yellow": {
+                "$ref": "#/definitions/Color",
+                "description": "Sets the color used as ANSI yellow."
+              }
+            },
+            "type": "object"
+          },
+          "type": "array"
+        },
+        "commandline": {
+          "description": "Executable used in the profile.",
+          "type": "string"
+        },
+        "connectionType": {
+          "$ref": "#/definitions/ProfileGuid",
+          "description": "A GUID reference to a connection type. Currently undocumented as of 0.3, this is used for Azure Cloud Shell"
+        },
+        "cursorColor": {
+          "$ref": "#/definitions/Color",
+          "default": "#FFFFFF",
+          "description": "Sets the cursor color for the profile. Uses hex color format: \"#rrggbb\"."
+        },
+        "cursorHeight": {
+          "description": "Sets the percentage height of the cursor starting from the bottom. Only works when cursorShape is set to \"vintage\". Accepts values from 25-100.",
+          "maximum": 100,
+          "minimum": 25,
+          "type": "integer"
+        },
+        "cursorShape": {
+          "default": "bar",
+          "description": "Sets the cursor shape for the profile. Possible values: \"vintage\" ( ▃ ), \"bar\" ( ┃, default ), \"underscore\" ( ▁ ), \"filledBox\" ( █ ), \"emptyBox\" ( ▯ )",
+          "enum": [
+            "bar",
+            "emptyBox",
+            "filledBox",
+            "underscore",
+            "vintage"
+          ],
+          "type": "string"
+        },
+        "experimental.retroTerminalEffect": {
+          "description": "When set to true, enable retro terminal effects. This is an experimental feature, and its continued existence is not guaranteed.",
+          "type": "boolean"
+        },
+        "fontFace": {
+          "default": "Consolas",
+          "description": "Name of the font face used in the profile.",
+          "type": "string"
+        },
+        "fontSize": {
+          "default": 12,
+          "description": "Sets the font size.",
+          "minimum": 1,
+          "type": "integer"
+        },
+        "foreground": {
+          "$ref": "#/definitions/Color",
+          "default": "#cccccc",
+          "description": "Sets the foreground color of the profile. Overrides foreground set in color scheme if colorscheme is set. Uses hex color format: \"#rrggbb\".",
+          "type": ["string", "null"]
+        },
+        "guid": {
+          "$ref": "#/definitions/ProfileGuid",
+          "description": "Unique identifier of the profile. Written in registry format: \"{00000000-0000-0000-0000-000000000000}\"."
+        },
+        "hidden": {
+          "default": false,
+          "description": "If set to true, the profile will not appear in the list of profiles. This can be used to hide default profiles and dynamically generated profiles, while leaving them in your settings file.",
+          "type": "boolean"
+        },
+        "historySize": {
+          "default": 9001,
+          "description": "The number of lines above the ones displayed in the window you can scroll back to.",
+          "minimum": -1,
+          "type": "integer"
+        },
+        "icon": {
+          "description": "Image file location of the icon used in the profile. Displays within the tab and the dropdown menu.",
+          "type": "string"
+        },
+        "name": {
+          "description": "Name of the profile. Displays in the dropdown menu.",
+          "minLength": 1,
+          "type": "string"
+        },
+        "padding": {
+          "default": "8, 8, 8, 8",
+          "description": "Sets the padding around the text within the window. Can have three different formats: \"#\" sets the same padding for all sides, \"#, #\" sets the same padding for left-right and top-bottom, and \"#, #, #, #\" sets the padding individually for left, top, right, and bottom.",
+          "pattern": "^-?[0-9]+(\\.[0-9]+)?( *, *-?[0-9]+(\\.[0-9]+)?|( *, *-?[0-9]+(\\.[0-9]+)?){3})?$",
+          "type": "string"
+        },
+        "scrollbarState": {
+          "default": "visible",
+          "description": "Defines the visibility of the scrollbar.",
+          "enum": [
+            "visible",
+            "hidden"
+          ],
+          "type": "string"
+        },
+        "selectionBackground": {
+          "$ref": "#/definitions/Color",
+          "description": "Sets the selection background color of the profile. Overrides selection background set in color scheme if colorscheme is set. Uses hex color format: \"#rrggbb\"."
+        },
+        "snapOnInput": {
+          "default": true,
+          "description": "When set to true, the window will scroll to the command input line when typing. When set to false, the window will not scroll when you start typing.",
+          "type": "boolean"
+        },
+        "source": {
+          "description": "Stores the name of the profile generator that originated this profile.",
+          "type": "string"
+        },
+        "startingDirectory": {
+          "description": "The directory the shell starts in when it is loaded.",
+          "type": "string"
+        },
+        "suppressApplicationTitle": {
+          "description": "When set to true, tabTitle overrides the default title of the tab and any title change messages from the application will be suppressed. When set to false, tabTitle behaves as normal.",
+          "type": "boolean"
+        },
+        "tabTitle": {
+          "description": "If set, will replace the name as the title to pass to the shell on startup. Some shells (like bash) may choose to ignore this initial value, while others (cmd, powershell) may use this value over the lifetime of the application.",
+          "type": "string"
+        },
+        "useAcrylic": {
+          "default": false,
+          "description": "When set to true, the window will have an acrylic background. When set to false, the window will have a plain, untextured background.",
+          "type": "boolean"
+        }
+      },
+      "type": "object"
+    },
+    "ProfileList": {
+      "description": "A list of profiles and the properties specific to each.",
+      "items": {
+        "$ref": "#/definitions/Profile",
+        "required": [
+          "guid",
+          "name"
+        ]
+      },
+      "type": "array"
+    },
+    "ProfilesObject": {
+      "description": "A list of profiles and default settings that apply to all of them",
+      "properties": {
+        "list": {
+          "$ref": "#/definitions/ProfileList"
+        },
+        "defaults": {
+          "description": "The default settings that apply to every profile.",
+          "$ref": "#/definitions/Profile"
+        }
+      },
+      "type": "object"
+    },
+    "SchemeList": {
+      "description": "Properties are specific to each color scheme. ColorTool is a great tool you can use to create and explore new color schemes. All colors use hex color format.",
+      "items": {
+        "additionalProperties": false,
+        "properties": {
+          "name": {
+            "description": "Name of the color scheme.",
+            "minLength": 1,
+            "type": "string"
+          },
+          "background": {
+            "$ref": "#/definitions/Color",
+            "description": "Sets the background color of the color scheme."
+          },
+          "black": {
+            "$ref": "#/definitions/Color",
+            "description": "Sets the color used as ANSI black."
+          },
+          "blue": {
+            "$ref": "#/definitions/Color",
+            "description": "Sets the color used as ANSI blue."
+          },
+          "brightBlack": {
+            "$ref": "#/definitions/Color",
+            "description": "Sets the color used as ANSI bright black."
+          },
+          "brightBlue": {
+            "$ref": "#/definitions/Color",
+            "description": "Sets the color used as ANSI bright blue."
+          },
+          "brightCyan": {
+            "$ref": "#/definitions/Color",
+            "description": "Sets the color used as ANSI bright cyan."
+          },
+          "brightGreen": {
+            "$ref": "#/definitions/Color",
+            "description": "Sets the color used as ANSI bright green."
+          },
+          "brightPurple": {
+            "$ref": "#/definitions/Color",
+            "description": "Sets the color used as ANSI bright purple."
+          },
+          "brightRed": {
+            "$ref": "#/definitions/Color",
+            "description": "Sets the color used as ANSI bright red."
+          },
+          "brightWhite": {
+            "$ref": "#/definitions/Color",
+            "description": "Sets the color used as ANSI bright white."
+          },
+          "brightYellow": {
+            "$ref": "#/definitions/Color",
+            "description": "Sets the color used as ANSI bright yellow."
+          },
+          "cyan": {
+            "$ref": "#/definitions/Color",
+            "description": "Sets the color used as ANSI cyan."
+          },
+          "foreground": {
+            "$ref": "#/definitions/Color",
+            "description": "Sets the foreground color of the color scheme."
+          },
+          "green": {
+            "$ref": "#/definitions/Color",
+            "description": "Sets the color used as ANSI green."
+          },
+          "purple": {
+            "$ref": "#/definitions/Color",
+            "description": "Sets the color used as ANSI purple."
+          },
+          "red": {
+            "$ref": "#/definitions/Color",
+            "description": "Sets the color used as ANSI red."
+          },
+          "selectionBackground": {
+            "$ref": "#/definitions/Color",
+            "description": "Sets the selection background color of the color scheme."
+          },
+          "white": {
+            "$ref": "#/definitions/Color",
+            "description": "Sets the color used as ANSI white."
+          },
+          "yellow": {
+            "$ref": "#/definitions/Color",
+            "description": "Sets the color used as ANSI yellow."
+          }
+        },
+        "type": "object"
+      },
+      "type": "array"
+    }
+  },
+  "oneOf": [
+    {
+      "allOf": [
+        { "$ref": "#/definitions/Globals" },
+        {
+          "additionalItems": true,
+          "properties": {
+            "profiles": {
+              "oneOf": [
+              { "$ref": "#/definitions/ProfileList" },
+              { "$ref": "#/definitions/ProfilesObject" }
+              ]
+            },
+            "schemes": { "$ref": "#/definitions/SchemeList" }
+          },
+          "required": [
+            "profiles",
+            "schemes",
+            "defaultProfile"
+          ]
+        }
+      ]
+    },
+    {
+      "additionalItems": false,
+      "properties": {
+        "globals": { "$ref": "#/definitions/Globals" },
+        "profiles": {
+          "oneOf": [
+          { "$ref": "#/definitions/ProfileList" },
+          { "$ref": "#/definitions/ProfilesObject" }
+          ]
+        },
+        "schemes": { "$ref": "#/definitions/SchemeList" }
+      },
+      "required": [
+        "profiles",
+        "schemes",
+        "globals"
+      ]
+    }
+  ]
+}

doc/reference/UTF8-torture-test.txt

@@ -0,0 +1,212 @@
+
+UTF-8 encoded sample plain-text file
+‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾
+
+Markus Kuhn [ˈmaʳkʊs kuːn] <http://www.cl.cam.ac.uk/~mgk25/> — 2002-07-25
+
+
+The ASCII compatible UTF-8 encoding used in this plain-text file
+is defined in Unicode, ISO 10646-1, and RFC 2279.
+
+
+Using Unicode/UTF-8, you can write in emails and source code things such as
+
+Mathematics and sciences:
+
+  ∮ E⋅da = Q,  n → ∞, ∑ f(i) = ∏ g(i),      ⎧⎡⎛┌─────┐⎞⎤⎫
+                                            ⎪⎢⎜│a²+b³ ⎟⎥⎪
+  ∀x∈ℝ: ⌈x⌉ = −⌊−x⌋, α ∧ ¬β = ¬(¬α ∨ β),    ⎪⎢⎜│───── ⎟⎥⎪
+                                            ⎪⎢⎜⎷ c₈   ⎟⎥⎪
+  ℕ ⊆ ℕ₀ ⊂ ℤ ⊂ ℚ ⊂ ℝ ⊂ ℂ,                   ⎨⎢⎜       ⎟⎥⎬
+                                            ⎪⎢⎜ ∞     ⎟⎥⎪
+  ⊥ < a ≠ b ≡ c ≤ d ≪ ⊤ ⇒ (⟦A⟧ ⇔ ⟪B⟫),      ⎪⎢⎜ ⎲     ⎟⎥⎪
+                                            ⎪⎢⎜ ⎳aⁱ-bⁱ⎟⎥⎪
+  2H₂ + O₂ ⇌ 2H₂O, R = 4.7 kΩ, ⌀ 200 mm     ⎩⎣⎝i=1    ⎠⎦⎭
+
+Linguistics and dictionaries:
+
+  ði ıntəˈnæʃənəl fəˈnɛtık əsoʊsiˈeıʃn
+  Y [ˈʏpsilɔn], Yen [jɛn], Yoga [ˈjoːgɑ]
+
+APL:
+
+  ((V⍳V)=⍳⍴V)/V←,V    ⌷←⍳→⍴∆∇⊃‾⍎⍕⌈
+
+Nicer typography in plain text files:
+
+  ╔══════════════════════════════════════════╗
+  ║                                          ║
+  ║   • ‘single’ and “double” quotes         ║
+  ║                                          ║
+  ║   • Curly apostrophes: “We’ve been here” ║
+  ║                                          ║
+  ║   • Latin-1 apostrophe and accents: '´`  ║
+  ║                                          ║
+  ║   • ‚deutsche‘ „Anführungszeichen“       ║
+  ║                                          ║
+  ║   • †, ‡, ‰, •, 3–4, —, −5/+5, ™, …      ║
+  ║                                          ║
+  ║   • ASCII safety test: 1lI|, 0OD, 8B     ║
+  ║                      ╭─────────╮         ║
+  ║   • the euro symbol: │ 14.95 € │         ║
+  ║                      ╰─────────╯         ║
+  ╚══════════════════════════════════════════╝
+
+Combining characters:
+
+  STARGΛ̊TE SG-1, a = v̇ = r̈, a⃑ ⊥ b⃑
+
+Greek (in Polytonic):
+
+  The Greek anthem:
+
+  Σὲ γνωρίζω ἀπὸ τὴν κόψη
+  τοῦ σπαθιοῦ τὴν τρομερή,
+  σὲ γνωρίζω ἀπὸ τὴν ὄψη
+  ποὺ μὲ βία μετράει τὴ γῆ.
+
+  ᾿Απ᾿ τὰ κόκκαλα βγαλμένη
+  τῶν ῾Ελλήνων τὰ ἱερά
+  καὶ σὰν πρῶτα ἀνδρειωμένη
+  χαῖρε, ὦ χαῖρε, ᾿Ελευθεριά!
+
+  From a speech of Demosthenes in the 4th century BC:
+
+  Οὐχὶ ταὐτὰ παρίσταταί μοι γιγνώσκειν, ὦ ἄνδρες ᾿Αθηναῖοι,
+  ὅταν τ᾿ εἰς τὰ πράγματα ἀποβλέψω καὶ ὅταν πρὸς τοὺς
+  λόγους οὓς ἀκούω· τοὺς μὲν γὰρ λόγους περὶ τοῦ
+  τιμωρήσασθαι Φίλιππον ὁρῶ γιγνομένους, τὰ δὲ πράγματ᾿
+  εἰς τοῦτο προήκοντα,  ὥσθ᾿ ὅπως μὴ πεισόμεθ᾿ αὐτοὶ
+  πρότερον κακῶς σκέψασθαι δέον. οὐδέν οὖν ἄλλο μοι δοκοῦσιν
+  οἱ τὰ τοιαῦτα λέγοντες ἢ τὴν ὑπόθεσιν, περὶ ἧς βουλεύεσθαι,
+  οὐχὶ τὴν οὖσαν παριστάντες ὑμῖν ἁμαρτάνειν. ἐγὼ δέ, ὅτι μέν
+  ποτ᾿ ἐξῆν τῇ πόλει καὶ τὰ αὑτῆς ἔχειν ἀσφαλῶς καὶ Φίλιππον
+  τιμωρήσασθαι, καὶ μάλ᾿ ἀκριβῶς οἶδα· ἐπ᾿ ἐμοῦ γάρ, οὐ πάλαι
+  γέγονεν ταῦτ᾿ ἀμφότερα· νῦν μέντοι πέπεισμαι τοῦθ᾿ ἱκανὸν
+  προλαβεῖν ἡμῖν εἶναι τὴν πρώτην, ὅπως τοὺς συμμάχους
+  σώσομεν. ἐὰν γὰρ τοῦτο βεβαίως ὑπάρξῃ, τότε καὶ περὶ τοῦ
+  τίνα τιμωρήσεταί τις καὶ ὃν τρόπον ἐξέσται σκοπεῖν· πρὶν δὲ
+  τὴν ἀρχὴν ὀρθῶς ὑποθέσθαι, μάταιον ἡγοῦμαι περὶ τῆς
+  τελευτῆς ὁντινοῦν ποιεῖσθαι λόγον.
+
+  Δημοσθένους, Γ´ ᾿Ολυνθιακὸς
+
+Georgian:
+
+  From a Unicode conference invitation:
+
+  გთხოვთ ახლავე გაიაროთ რეგისტრაცია Unicode-ის მეათე საერთაშორისო
+  კონფერენციაზე დასასწრებად, რომელიც გაიმართება 10-12 მარტს,
+  ქ. მაინცში, გერმანიაში. კონფერენცია შეჰკრებს ერთად მსოფლიოს
+  ექსპერტებს ისეთ დარგებში როგორიცაა ინტერნეტი და Unicode-ი,
+  ინტერნაციონალიზაცია და ლოკალიზაცია, Unicode-ის გამოყენება
+  ოპერაციულ სისტემებსა, და გამოყენებით პროგრამებში, შრიფტებში,
+  ტექსტების დამუშავებასა და მრავალენოვან კომპიუტერულ სისტემებში.
+
+Russian:
+
+  From a Unicode conference invitation:
+
+  Зарегистрируйтесь сейчас на Десятую Международную Конференцию по
+  Unicode, которая состоится 10-12 марта 1997 года в Майнце в Германии.
+  Конференция соберет широкий круг экспертов по  вопросам глобального
+  Интернета и Unicode, локализации и интернационализации, воплощению и
+  применению Unicode в различных операционных системах и программных
+  приложениях, шрифтах, верстке и многоязычных компьютерных системах.
+
+Thai (UCS Level 2):
+
+  Excerpt from a poetry on The Romance of The Three Kingdoms (a Chinese
+  classic 'San Gua'):
+
+  [----------------------------|------------------------]
+    ๏ แผ่นดินฮั่นเสื่อมโทรมแสนสังเวช  พระปกเกศกองบู๊กู้ขึ้นใหม่
+  สิบสองกษัตริย์ก่อนหน้าแลถัดไป       สององค์ไซร้โง่เขลาเบาปัญญา
+    ทรงนับถือขันทีเป็นที่พึ่ง           บ้านเมืองจึงวิปริตเป็นนักหนา
+  โฮจิ๋นเรียกทัพทั่วหัวเมืองมา         หมายจะฆ่ามดชั่วตัวสำคัญ
+    เหมือนขับไสไล่เสือจากเคหา      รับหมาป่าเข้ามาเลยอาสัญ
+  ฝ่ายอ้องอุ้นยุแยกให้แตกกัน          ใช้สาวนั้นเป็นชนวนชื่นชวนใจ
+    พลันลิฉุยกุยกีกลับก่อเหตุ          ช่างอาเพศจริงหนาฟ้าร้องไห้
+  ต้องรบราฆ่าฟันจนบรรลัย           ฤๅหาใครค้ำชูกู้บรรลังก์ ฯ
+
+  (The above is a two-column text. If combining characters are handled
+  correctly, the lines of the second column should be aligned with the
+  | character above.)
+
+Ethiopian:
+
+  Proverbs in the Amharic language:
+
+  ሰማይ አይታረስ ንጉሥ አይከሰስ።
+  ብላ ካለኝ እንደአባቴ በቆመጠኝ።
+  ጌጥ ያለቤቱ ቁምጥና ነው።
+  ደሀ በሕልሙ ቅቤ ባይጠጣ ንጣት በገደለው።
+  የአፍ ወለምታ በቅቤ አይታሽም።
+  አይጥ በበላ ዳዋ ተመታ።
+  ሲተረጉሙ ይደረግሙ።
+  ቀስ በቀስ፥ ዕንቁላል በእግሩ ይሄዳል።
+  ድር ቢያብር አንበሳ ያስር።
+  ሰው እንደቤቱ እንጅ እንደ ጉረቤቱ አይተዳደርም።
+  እግዜር የከፈተውን ጉሮሮ ሳይዘጋው አይድርም።
+  የጎረቤት ሌባ፥ ቢያዩት ይስቅ ባያዩት ያጠልቅ።
+  ሥራ ከመፍታት ልጄን ላፋታት።
+  ዓባይ ማደሪያ የለው፥ ግንድ ይዞ ይዞራል።
+  የእስላም አገሩ መካ የአሞራ አገሩ ዋርካ።
+  ተንጋሎ ቢተፉ ተመልሶ ባፉ።
+  ወዳጅህ ማር ቢሆን ጨርስህ አትላሰው።
+  እግርህን በፍራሽህ ልክ ዘርጋ።
+
+Runes:
+
+  ᚻᛖ ᚳᚹᚫᚦ ᚦᚫᛏ ᚻᛖ ᛒᚢᛞᛖ ᚩᚾ ᚦᚫᛗ ᛚᚪᚾᛞᛖ ᚾᚩᚱᚦᚹᛖᚪᚱᛞᚢᛗ ᚹᛁᚦ ᚦᚪ ᚹᛖᛥᚫ
+
+  (Old English, which transcribed into Latin reads 'He cwaeth that he
+  bude thaem lande northweardum with tha Westsae.' and means 'He said
+  that he lived in the northern land near the Western Sea.')
+
+Braille:
+
+  ⡌⠁⠧⠑ ⠼⠁⠒  ⡍⠜⠇⠑⠹⠰⠎ ⡣⠕⠌
+
+  ⡍⠜⠇⠑⠹ ⠺⠁⠎ ⠙⠑⠁⠙⠒ ⠞⠕ ⠃⠑⠛⠔ ⠺⠊⠹⠲ ⡹⠻⠑ ⠊⠎ ⠝⠕ ⠙⠳⠃⠞
+  ⠱⠁⠞⠑⠧⠻ ⠁⠃⠳⠞ ⠹⠁⠞⠲ ⡹⠑ ⠗⠑⠛⠊⠌⠻ ⠕⠋ ⠙⠊⠎ ⠃⠥⠗⠊⠁⠇ ⠺⠁⠎
+  ⠎⠊⠛⠝⠫ ⠃⠹ ⠹⠑ ⠊⠇⠻⠛⠹⠍⠁⠝⠂ ⠹⠑ ⠊⠇⠻⠅⠂ ⠹⠑ ⠥⠝⠙⠻⠞⠁⠅⠻⠂
+  ⠁⠝⠙ ⠹⠑ ⠡⠊⠑⠋ ⠍⠳⠗⠝⠻⠲ ⡎⠊⠗⠕⠕⠛⠑ ⠎⠊⠛⠝⠫ ⠊⠞⠲ ⡁⠝⠙
+  ⡎⠊⠗⠕⠕⠛⠑⠰⠎ ⠝⠁⠍⠑ ⠺⠁⠎ ⠛⠕⠕⠙ ⠥⠏⠕⠝ ⠰⡡⠁⠝⠛⠑⠂ ⠋⠕⠗ ⠁⠝⠹⠹⠔⠛ ⠙⠑
+  ⠡⠕⠎⠑ ⠞⠕ ⠏⠥⠞ ⠙⠊⠎ ⠙⠁⠝⠙ ⠞⠕⠲
+
+  ⡕⠇⠙ ⡍⠜⠇⠑⠹ ⠺⠁⠎ ⠁⠎ ⠙⠑⠁⠙ ⠁⠎ ⠁ ⠙⠕⠕⠗⠤⠝⠁⠊⠇⠲
+
+  ⡍⠔⠙⠖ ⡊ ⠙⠕⠝⠰⠞ ⠍⠑⠁⠝ ⠞⠕ ⠎⠁⠹ ⠹⠁⠞ ⡊ ⠅⠝⠪⠂ ⠕⠋ ⠍⠹
+  ⠪⠝ ⠅⠝⠪⠇⠫⠛⠑⠂ ⠱⠁⠞ ⠹⠻⠑ ⠊⠎ ⠏⠜⠞⠊⠊⠥⠇⠜⠇⠹ ⠙⠑⠁⠙ ⠁⠃⠳⠞
+  ⠁ ⠙⠕⠕⠗⠤⠝⠁⠊⠇⠲ ⡊ ⠍⠊⠣⠞ ⠙⠁⠧⠑ ⠃⠑⠲ ⠔⠊⠇⠔⠫⠂ ⠍⠹⠎⠑⠇⠋⠂ ⠞⠕
+  ⠗⠑⠛⠜⠙ ⠁ ⠊⠕⠋⠋⠔⠤⠝⠁⠊⠇ ⠁⠎ ⠹⠑ ⠙⠑⠁⠙⠑⠌ ⠏⠊⠑⠊⠑ ⠕⠋ ⠊⠗⠕⠝⠍⠕⠝⠛⠻⠹
+  ⠔ ⠹⠑ ⠞⠗⠁⠙⠑⠲ ⡃⠥⠞ ⠹⠑ ⠺⠊⠎⠙⠕⠍ ⠕⠋ ⠳⠗ ⠁⠝⠊⠑⠌⠕⠗⠎
+  ⠊⠎ ⠔ ⠹⠑ ⠎⠊⠍⠊⠇⠑⠆ ⠁⠝⠙ ⠍⠹ ⠥⠝⠙⠁⠇⠇⠪⠫ ⠙⠁⠝⠙⠎
+  ⠩⠁⠇⠇ ⠝⠕⠞ ⠙⠊⠌⠥⠗⠃ ⠊⠞⠂ ⠕⠗ ⠹⠑ ⡊⠳⠝⠞⠗⠹⠰⠎ ⠙⠕⠝⠑ ⠋⠕⠗⠲ ⡹⠳
+  ⠺⠊⠇⠇ ⠹⠻⠑⠋⠕⠗⠑ ⠏⠻⠍⠊⠞ ⠍⠑ ⠞⠕ ⠗⠑⠏⠑⠁⠞⠂ ⠑⠍⠏⠙⠁⠞⠊⠊⠁⠇⠇⠹⠂ ⠹⠁⠞
+  ⡍⠜⠇⠑⠹ ⠺⠁⠎ ⠁⠎ ⠙⠑⠁⠙ ⠁⠎ ⠁ ⠙⠕⠕⠗⠤⠝⠁⠊⠇⠲
+
+  (The first couple of paragraphs of "A Christmas Carol" by Dickens)
+
+Compact font selection example text:
+
+  ABCDEFGHIJKLMNOPQRSTUVWXYZ /0123456789
+  abcdefghijklmnopqrstuvwxyz £©µÀÆÖÞßéöÿ
+  –—‘“”„†•…‰™œŠŸž€ ΑΒΓΔΩαβγδω АБВГДабвгд
+  ∀∂∈ℝ∧∪≡∞ ↑↗↨↻⇣ ┐┼╔╘░►☺♀ fi<>⑀₂ἠḂӥẄɐː⍎אԱა
+
+Greetings in various languages:
+
+  Hello world, Καλημέρα κόσμε, コンニチハ
+
+Box drawing alignment tests:                                          █
+                                                                      ▉
+  ╔══╦══╗  ┌──┬──┐  ╭──┬──╮  ╭──┬──╮  ┏━━┳━━┓  ┎┒┏┑   ╷  ╻ ┏┯┓ ┌┰┐    ▊ ╱╲╱╲╳╳╳
+  ║┌─╨─┐║  │╔═╧═╗│  │╒═╪═╕│  │╓─╁─╖│  ┃┌─╂─┐┃  ┗╃╄┙  ╶┼╴╺╋╸┠┼┨ ┝╋┥    ▋ ╲╱╲╱╳╳╳
+  ║│╲ ╱│║  │║   ║│  ││ │ ││  │║ ┃ ║│  ┃│ ╿ │┃  ┍╅╆┓   ╵  ╹ ┗┷┛ └┸┘    ▌ ╱╲╱╲╳╳╳
+  ╠╡ ╳ ╞╣  ├╢   ╟┤  ├┼─┼─┼┤  ├╫─╂─╫┤  ┣┿╾┼╼┿┫  ┕┛┖┚     ┌┄┄┐ ╎ ┏┅┅┓ ┋ ▍ ╲╱╲╱╳╳╳
+  ║│╱ ╲│║  │║   ║│  ││ │ ││  │║ ┃ ║│  ┃│ ╽ │┃  ░░▒▒▓▓██ ┊  ┆ ╎ ╏  ┇ ┋ ▎
+  ║└─╥─┘║  │╚═╤═╝│  │╘═╪═╛│  │╙─╀─╜│  ┃└─╂─┘┃  ░░▒▒▓▓██ ┊  ┆ ╎ ╏  ┇ ┋ ▏
+  ╚══╩══╝  └──┴──┘  ╰──┴──╯  ╰──┴──╯  ┗━━┻━━┛  ▗▄▖▛▀▜   └╌╌┘ ╎ ┗╍╍┛ ┋  ▁▂▃▄▅▆▇█
+                                               ▝▀▘▙▄▟

doc/specs/#1043 - Set the initial position of the Terminal/spec.md

@@ -0,0 +1,100 @@
+---
+author: Kaiyu Wang KaiyuWang16/kawa@microsoft.com
+created on: 2019-09-03
+last updated: 2020-01-02
+issue id: #1043
+---
+
+# Set the initial position for terminal
+
+## Abstract
+
+This spec is for task #1043 “Be able to set an initial position for the terminal”. It goes over the details of a new feature that allows users to set the initial position and size of the terminal. Expected behavior and design of this feature is included. Besides, future possible follow-up works are also addressed.
+
+## Inspiration
+
+The idea is to allow users to set the initial position of the Terminal when they launch it, prevent the Terminal from appearing on unexpected position (e.g. outside of the screen bounds). We are also going to let users choose to maximize the window when they launch it.
+
+## Solution Design
+
+For now, the Terminal window is put on a default initial position. The program uses CW_USEDEFAULT in the screen coordinates for top-left corner. We have two different types of window – client window and non-client window. However, code path for window creation (WM_CREATE message is shared by the two types of windows) are almost the same for the two types of windows, except that there are some differences in calculation of the width and height of the window.
+
+Two new properties should be added in the json settings file:
+
+**initialPosition**: string. This sets the initial horizontal and vertical position of the top-left corner of the window. This property follows a structure: "X value, Y value" and has following rules:
+
+1. All spaces will be ignored.
+
+2. Both X value and Y values are optional. If anyone of them is missing, or the value is invalid, system default value will be used. Examples:
+
+       ", 1000" equals to (default, 1000)
+       "1000, " equals to (1000, default)
+       "," equals to (default, default)
+       "abc, 1000" equals to (default, 1000)
+
+**launchMode**: string. Determine the launch mode. There are two modes for now
+
+1. maximize: the window will be maximized when launch.
+2. default: the window will be initialized with system default size.
+
+The steps of this process:
+
+1. Set the top-left origin, width and height to CW_USEDEFAULT.
+2. Get the dpi of the nearest monitor; Load settings.
+3. From settings, find the user-defined initial position and launch mode. 
+4. If the user sets custom initial position, calculate the new position considering the current dpi and monitor. If not, use system default value.
+5. If the user set launch mode as "maximize", calculate the new height and width. If the user choose "default", use system default size.
+6. SetWindowPos with the new position and dimension of the window. 
+
+Step 2 to 6 should be done in `AppHost::_HandleCreateWindow`, which is consistent to the current code.
+
+In step 4, we may need to consider the dpi of the current monitor and multi-monitor scenario when calculating the initial position of the window.
+
+Edge cases:
+
+1. Multiple monitors. The user should be able to set the initial position to any monitors attached. For the monitors on the left side of the major monitor, the initial position values are negative. 
+2. If the initial position is larger than the screen resolution and the window top left corner is off-screen, we should let user be able to see and drag the window back on screen. One solution is to set the initial position to the top left corner of the nearest monitor if the top left is off-screen.
+3. If the user wants to launch maximized and provides an initial position, we should launch the maximized window on the top left corner of the monitor where the position is located.
+4. Launch the Terminal on a monitor with custom dpi. Changing the dpi of the monitor will not affect the initial position of the top left corner. So we do not need to handle this case. 
+5. Launch the Terminal on a monitor with custom resolution. Changing the resolution will change the available point for the initial position. (2) already covers this case. 
+
+## UI/UX Design
+
+Upon successful implementation, the user is able to add new properties to the json profile file, which is illustrated in the code block below:
+```json
+"initialPosition": "500,500",
+"launchMode": "default"
+```
+The rest of the UI will be the same of the current Terminal experience, except that the initial position may be different.
+
+### Accessibility
+
+Users can only set the initial position and launch mode in the Json file with keyboard. Thus, this will not affect accessibility. 
+
+### Reliability
+We need to make sure that whatever the initial position is set, the user can access the Terminal window. This is guaranteed because if the top left corner position of the Terminal Window is out of screen, we put it on the top left corner of the screen. 
+
+### Performance, Power, and Efficiency
+
+More data reading and calculation will be included in Terminal Launch process, which may inversely influence the launch time. However, the impact is trivial. 
+
+## Potential Issues
+
+We need to consider multi-monitor scenario. If the user has multiple monitors, we must guarantee that the Terminal could be initialized as expected. We can keep an eye on the feedbacks of this feature from the community.
+
+## Future considerations
+
+For now, this feature only allows the user to set initial position and choose whether to maximize the window when launch. In the future, we may consider follow-up features like:
+
+1. Save the position of the Terminal on exit, and restore the position on the next launch. This could be a true/false feature that users could choose to set.
+
+2. We may need to consider multiple Terminal windows scenario. If the user opens multiple Terminal windows, then we need to consider how to save and restore the position. 
+
+3. We may also consider more launch modes. Like full screen mode and minimized mode. 
+
+Github issue for future follow-ups: https://github.com/microsoft/terminal/issues/766
+
+## Resources
+
+Github issue:
+https://github.com/microsoft/terminal/issues/1043

doc/specs/#1142 - Keybinding Arguments.md

@@ -0,0 +1,362 @@
+---
+author: Mike Griese @zadjii-msft
+created on: 2019-06-19
+last updated: 2019-07-14
+issue id: 1142
+---
+
+# Arbitrary Keybindings Arguments
+
+## Abstract
+
+The goal of this change is to both simplify the keybindings, and also enable far
+more flexibility when editing a user's keybindings.
+
+Currently, we have many actions that are very similar in implementation - for
+example, `newTabProfile0`, `newTabProfile1`, `newTabProfile2`, etc. All these
+actions are _fundamentally_ the same function. However, we've needed to define 9
+different actions to enable the user to provide different values to the `newTab`
+function.
+
+With this change, we'll be able to remove these _essentially_ duplicated events,
+and allow the user to specify arbitrary arguments to these functions.
+
+## Inspiration
+
+Largely inspired by the keybindings in VsCode and Sublime Text. Additionally,
+much of the content regarding keybinding events being "handled" was designed as
+a solution for [#2285].
+
+## Solution Design
+
+We'll need to introduce args to some actions that we already have defined. These
+are the actions I'm thinking about when writing this spec:
+
+```csharp
+    // These events already exist like this:
+    delegate void NewTabWithProfileEventArgs(Int32 profileIndex);
+    delegate void SwitchToTabEventArgs(Int32 profileIndex);
+    delegate void ResizePaneEventArgs(Direction direction);
+    delegate void MoveFocusEventArgs(Direction direction);
+
+    // These events either exist in another form or don't exist.
+    delegate void CopyTextEventArgs(Boolean copyWhitespace);
+    delegate void ScrollEventArgs(Int32 numLines);
+    delegate void SplitProfileEventArgs(Orientation splitOrientation, Int32 profileIndex);
+```
+
+Ideally, after this change, the bindings for these actions would look something
+like the following:
+
+```js
+{ "keys": ["ctrl+shift+1"], "command": "newTabProfile", "args": { "profileIndex":0 } },
+{ "keys": ["ctrl+shift+2"], "command": "newTabProfile", "args": { "profileIndex":1 } },
+// etc...
+
+{ "keys": ["alt+1"], "command": "switchToTab", "args": { "index":0 } },
+{ "keys": ["alt+2"], "command": "switchToTab", "args": { "index":1 } },
+// etc...
+
+{ "keys": ["alt+shift+down"], "command": "resizePane", "args": { "direction":"down" } },
+{ "keys": ["alt+shift+up"], "command": "resizePane", "args": { "direction":"up" } },
+// etc...
+
+{ "keys": ["alt+down"], "command": "moveFocus", "args": { "direction":"down" } },
+{ "keys": ["alt+up"], "command": "moveFocus", "args": { "direction":"up" } },
+// etc...
+
+{ "keys": ["ctrl+c"], "command": "copy", "args": { "copyWhitespace":true } },
+{ "keys": ["ctrl+shift+c"], "command": "copy", "args": { "copyWhitespace":false } },
+
+{ "keys": ["ctrl+shift+down"], "command": "scroll", "args": { "numLines":1 } },
+{ "keys": ["ctrl+shift+up"], "command": "scroll", "args": { "numLines":-1 } },
+
+{ "keys": ["ctrl+alt+1"], "command": "splitProfile", "args": { "orientation":"vertical", "profileIndex": 0 } },
+{ "keys": ["ctrl+alt+shift+1"], "command": "splitProfile", "args": { "orientation":"horizontal", "profileIndex": 0 } },
+{ "keys": ["ctrl+alt+2"], "command": "splitProfile", "args": { "orientation":"vertical", "profileIndex": 1 } },
+{ "keys": ["ctrl+alt+shift+2"], "command": "splitProfile", "args": { "orientation":"horizontal", "profileIndex": 1 } },
+// etc...
+```
+
+Note that instead of having 9 different `newTabProfile<N>` actions, we have a
+singular `newTabProfile` action, and that action requires a `profileIndex` in
+the `args` object.
+
+Also, pay attention to the last set of keybindings, the `splitProfile` ones.
+This is a function that requires two arguments, both an `orientation` and a
+`profileIndex`. Before this change we would have needed to create 20 separate
+actions (10 profile indices * 2 directions) to handle these cases. Now it can
+be done with a single action that can be much more flexible in its
+implementation.
+
+### Parsing KeyBinding Arguments
+
+We'll add two new interfaces: `IActionArgs` and `IActionEventArgs`. Classes that
+implement `IActionArgs` will contain all the per-action args, like
+`CopyWhitespace` or `ProfileIndex`. `IActionArgs` by itself will be an empty
+interface, but all other arguments will derive from it. `IActionEventArgs` will
+have a single property `Handled`, which will be used for indicating if a
+particular event was processed or not. When parsing args, we'll build
+`IActionArgs` to contain all the parameters. When dispatching events, we'll
+build `IActionEventArgs` using the `IActionArgs` to set all the parameter values.
+
+All current keybinding events will be changed from their current types to
+`TypedEventHandler`s. These `TypedEventHandler`s second param will always be an
+instance of `IActionEventArgs`. So, for example:
+
+```csharp
+
+delegate void CopyTextEventArgs();
+delegate void NewTabEventArgs();
+delegate void NewTabWithProfileEventArgs(Int32 profileIndex);
+// ...
+
+[default_interface]
+runtimeclass AppKeyBindings : Microsoft.Terminal.Settings.IKeyBindings
+{
+    event CopyTextEventArgs CopyText;
+    event NewTabEventArgs NewTab;
+    event NewTabWithProfileEventArgs NewTabWithProfile;
+```
+
+Becomes:
+
+```csharp
+interface IActionArgs { /* Empty */ }
+
+runtimeclass ActionEventArgs
+{
+    Boolean Handled;
+    ActionArgs Args;
+}
+
+runtimeclass CopyTextArgs : IActionArgs
+{
+    Boolean CopyWhitespace;
+}
+
+runtimeclass NewTabWithProfileArgs : IActionArgs
+{
+    Int32 ProfileIndex;
+}
+runtimeclass NewTabWithProfileEventArgs : NewTabWithProfileArgs, IActionArgs { }
+
+[default_interface]
+runtimeclass AppKeyBindings : Microsoft.Terminal.Settings.IKeyBindings
+{
+    event Windows.Foundation.TypedEventHandler<AppKeyBindings, ActionEventArgs> CopyText;
+    event Windows.Foundation.TypedEventHandler<AppKeyBindings, ActionEventArgs> NewTab;
+    event Windows.Foundation.TypedEventHandler<AppKeyBindings, ActionEventArgs> NewTabWithProfile;
+```
+
+In this above example, the `CopyTextArgs` class actually contains all the
+potential arguments to the Copy action. `ActionEventArgs` is the class that
+holds any `ActionArgs`. When we parse the arguments, we'll build a
+`CopyTextArgs`, and when we're dispatching the event, we'll build a
+`ActionEventArgs` that holds a `CopyTextArgs` as its `Args` value, and dispatch
+the `ActionEventArgs` object.
+
+
+We'll also change our existing map in the `AppKeyBindings` implementation.
+Currently, it's a `std::unordered_map<KeyChord, ShortcutAction, ...>`, which
+uses the `KeyChord` to lookup the `ShortcutAction`. We'll need to introduce a
+new type `ActionAndArgs`:
+
+```csharp
+runtimeclass ActionAndArgs
+{
+    ShortcutAction Action;
+    IActionArgs Args;
+}
+```
+
+and we'll change the map in `AppKeyBindings` to a `std::unordered_map<KeyChord,
+ActionAndArgs, ...>`.
+
+When we're parsing keybindings, we'll need to construct args for each of the
+events to go with each binding. When we find some key chord bound to a given
+Action, we'll construct the `IActionArgs` for that action. For many actions,
+these args will be an empty class. However, when we do find an action that needs
+additional parsing, `AppKeyBindingsSerialization` will do the extra work to
+parse the args for that action.
+
+We'll keep a collection of functions that can be used for quickly determining
+how to parse the args for an action if necessary. This map will be a
+`std::unordered_map<ShortcutAction, function<IActionArgs(Json::Value)>>`. For
+most actions which don't require args, the function in this map will be set to
+nullptr, and we'll know that the action doesn't need to parse any more args.
+However, for actions that _do_ require args, we'll set up a global function that
+can be used to parse a json blob into an `IActionArgs`.
+
+Once the `IActionArgs` is built for the keybinding, we'll set it in
+`AppKeyBindings` with a updated `AppKeyBindings::SetKeyBinding` call.
+`SetKeyBinding`'s signature will be updated to take a `ActionAndArgs` instead.
+Should an action not need arguments, the `Args` member can be left `null` in the
+`ActionAndArgs`.
+
+### Executing KeyBinding Actions with Arguments
+
+When we're handling a keybinding in `AppKeyBindings::_DoAction`, we'll trigger
+the event handlers with the `IActionArgs` we've stored in the map with the
+`ShortcutAction`.
+
+Then, in `App`, we'll handle each of these events. We set up lambdas as event
+handlers for each event in `App::_HookupKeyBindings`. In each of those
+functions, we'll inspect the `IActionArgs` parameter, and use args from its
+implementation to call callbacks in the `App` class. We will update `App` to
+have methods defined with the actual keybinding function signatures.
+
+Instead of:
+
+```c++
+    void App::_HookupKeyBindings(TerminalApp::AppKeyBindings bindings) noexcept
+    {
+        // ...
+        bindings.NewTabWithProfile([this](const auto index) { _OpenNewTab({ index }); });
+    }
+```
+
+The code will look like:
+
+```c++
+    void App::_HookupKeyBindings(TerminalApp::AppKeyBindings bindings) noexcept
+    {
+        // ...
+        bindings.NewTabWithProfile({ this, &App::_OpenNewTab });
+    }
+    // ...
+    void App::_OpenNewTab(const TerminalApp::AppKeyBindings& sender, const NewTabEventArgs& args)
+    {
+        auto profileIndex = args.ProfileIndex();
+        args.Handled(true);
+        // ...
+    }
+```
+
+### Handling Keybinding Events
+
+Common to all implementations of `IActionArgs` is the `Handled` property. This
+will let the app indicate if it was able to actually process a keybinding event
+or not. While in the large majority of cases, the events will all be marked
+handled, there are some scenarios where the Terminal will need to know if the
+event could not be performed. For example, in the case of the `copy` event, the
+Terminal is only capable of copying text if there's actually a selection active.
+If there isn't a selection active, the `App` should make sure to not mark the
+event as not handled (it will leave `args.Handled(false)`). The App should only
+mark an event handled if it has actually dispatched the event.
+
+When an event is handled, we'll make sure to return `true` from
+`AppKeyBindings::TryKeyChord`, so that the terminal does not actually process
+that keypress. For events that were not handled by the application, the terminal
+will get another chance to dispatch the keypress.
+
+### Serializing KeyBinding Arguments
+
+Similar to how we parse arguments from the json, we'll need to update the
+`AppKeyBindingsSerialization` code to be able to serialize the arguments from a
+particular `IActionArgs`.
+
+## UI/UX Design
+
+### Keybindings in the New Tab Dropdown
+
+Small modifications will need to be made to the code responsible for the new tab
+dropdown. The new tab dropdown currently also displays the keybindings for each
+profile in the new tab dropdown. It does this by querying for the keybinding
+associated with each action. As we'll be removing the old `ShortcutAction`s that
+this dropdown uses, we'll need a new way to find which key chord corresponds to
+opening a given profile.
+
+We'll need to be able to not only lookup a keybinding by `ShortcutAction`, but
+also by a `ShortcutAction` and `IActionArgs`. We'll need to update the
+`AppKeyBindings::GetKeyBinding` method to also accept a `IActionArgs`. We'll
+also probably want each `IActionArgs` implementation to define an
+`Equals(IActionArgs)` method, so that we can easily check if two different
+`IActionArgs` are the same in this method.
+
+## Capabilities
+### Accessibility
+
+N/A
+
+### Security
+
+This should not introduce any _new_ security concerns. We're relying on the
+security of jsoncpp for parsing json. Adding new keys to the settings file
+will rely on jsoncpp's ability to securely parse those json values.
+
+### Reliability
+
+We'll need to make sure that invalid keybindings are ignored. Currently, we
+already gracefully ignore keybindings that have invalid `keys` or invalid
+`commands`. We'll need to add additional validation on invalid sets of `args`.
+When we're parsing the args from a Json blob, we'll make sure to only ever look
+for keys we're expecting and ignore everything else.
+
+If a keybinding requires certain args, but those args are not provided, we'll
+need to make sure those args each have reasonable default values to use. If for
+any reason a reasonable default can't be used for a keybinding argument, then
+we'll need to make sure to display an error dialog to the user for that
+scenario.
+
+When we're re-serializing settings, we'll only know about the keybinding arg
+keys that were successfully parsed. Other keys will be lost on re-serialization.
+
+### Compatibility
+
+This change will need to carefully be crafted to enable upgrading the legacy
+keybindings seamlessly. For most actions, the upgrade should be seamless. Since
+they already don't have args, their serializations will remain exactly the same.
+
+However, for the following actions that we'll be removing in favor of actions
+with arguments, we'll need to leave legacy deserialization in place to be able
+to find these old actions, and automatically build the correct `IActionArgs`
+for them:
+
+* `newTabProfile<n>`
+    - We'll need to make sure to build args with the right `profileIndex`
+      corresponding to the old action.
+* `switchToTab<n>`
+    - We'll need to make sure to build args with the right `index` corresponding
+      to the old action.
+* `resizePane<direction>` and `moveFocus<direction>`
+    - We'll need to make sure to build args with the right `direction`
+      corresponding to the old action.
+* `scroll<direction>`
+    - We'll need to make sure to build args with the right `amount` value
+      corresponding to the old action. `Up` will be -1, and `Down` will be 1.
+
+### Performance, Power, and Efficiency
+
+N/A
+
+## Potential Issues
+
+N/A
+
+## Future considerations
+
+* Should we support some sort of conversion from num keys to an automatic arg?
+  For example, by default, <kbd>Alt+&lt;N&gt;</kbd> to focuses the
+  Nth tab. Currently, those are 8 separate entries in the keybindings. Should we
+  enable some way for them be combined into a single binding entry, where the
+  binding automatically receives the number pressed as an arg? I couldn't find
+  any prior art of this, so it doesn't seem worth it to try and invent
+  currently. This might be something that we want to loop back on, but for the
+  time being, it remains out of scope of this PR.
+* When we inevitable support extensions, we'll need to allow extensions to also
+  be able to support their own custom keybindings and args. We'll probably want
+  to pass the settings to the extension to have the extension parse its own
+  settings. We'll want to be able to ask the extension for its own set of
+  `ActionAndArgs`<sup>[1]</sup> that it builds from the `keybindings`.  Once we
+  have that set of actions, we'll be able to store them locally, and dispatch
+  them quickly.
+  - [1] We probably won't be able to use the `ActionAndArgs` class directly,
+    since that class is specific to the actions we define. We'll need another
+    way for extensions to be able to uniquely identify their own actions.
+
+## Resources
+
+N/A
+
+[#2285]: https://github.com/microsoft/terminal/issues/2285

doc/specs/#1235 - Azure cloud shell connector/spec.md

@@ -0,0 +1,78 @@
+---
+author: Pankaj Bhojwani pankaj.d.bhoj@gmail.com
+created on: 2019-06-12
+last updated: 2019-06-12
+issue id: #1235
+---
+
+# Azure cloud shell connector
+
+## Abstract
+
+This spec goes over the details of how a feature enabling Windows Terminal users to connect to the Azure cloud shell should behave. It includes implementation and design considerations. 
+
+## Inspiration
+
+The idea is to give developers access to their Azure services smoothly within the Windows Terminal app, letting them engage with Azure technologies in a convenient manner. By integrating the Azure cloud shell into Windows Terminal, we can do just that. 
+
+## Solution Design
+
+The flowchart below shows the process by which the Azure cloud shell will be integrated into Windows Terminal. 
+
+![Sol Design](images/solDesign.png)
+
+The first three steps - authenticating the user, requesting a cloud shell and requesting a terminal - will be done via http requests. These requests will use the [cpprestsdk](https://github.com/Microsoft/cpprestsdk) library as that library is also owned by Microsoft, making it easy to resolve issues should any arise. 
+
+Authenticating the user will use [device code flow](https://github.com/AzureAD/microsoft-authentication-library-for-dotnet/wiki/Device-Code-Flow) since Windows Terminal does not support browser access (yet). As for the authentication endpoint, Azure AD v1.0 will be used because Azure AD v2.0 (also known as Microsoft Identity Platform) [does not support login to personal accounts with device code flow](https://github.com/AzureAD/microsoft-authentication-library-for-dotnet/wiki/Device-Code-Flow#constraints) at this time. Furthermore, upon successful authentication, the login/token information will be stored so that users will not need to repeatedly go through device code flow for future logins. Since this is sensitive information, the tokens will be stored with [Windows Storage](https://docs.microsoft.com/en-us/uwp/api/windows.storage) and encrypted with [Windows Security Data Protection](https://docs.microsoft.com/en-us/uwp/api/windows.security.cryptography.dataprotection.dataprotectionprovider). 
+
+The last step - connecting to the terminal - will be done via a websocket connection to allow easier communication between the app and the server. 
+
+The entire feature will be implemented in an isolated manner - i.e. it should have little to no dependency on the Windows Terminal app itself. This will allow the feature to become a plugin/extension once Windows Terminal supports plugins. More specifically, the connector will ascribe to the existing ITerminalConnection interface, making this simply another type of connection that Windows Terminal can make. 
+
+## UI/UX Design
+
+Upon successful implementation, a new profile option will appear for users as illustrated in the picture below (the profile will have its own unique icon when implemented).
+
+![Az Prof](images/azProf.png)
+
+As for the rest of the UI, the implementation will adopt the user's preferences from the Windows Terminal app. 
+
+## Capabilities
+
+### Accessibility
+
+This feature will not impact accessibility of Windows Terminal. 
+
+### Security
+
+Any feature that connects to a network introduces some security risks. However, with proper usage of Azure AD v1.0 and careful storage of tokens received from the server, these risks will be mitigated. 
+
+### Reliability
+
+This feature will not impact reliability of Windows Terminal. 
+
+### Compatibility
+
+With the implementation being mostly decoupled from the Windows Terminal app itself, no existing code/behaviors should break due to this feature. 
+
+### Performance, Power, and Efficiency
+
+This feature will not impact performance, power or efficiency of Windows Terminal.
+
+## Potential Issues
+
+1. This implementation depends on another open source project, [cpprestsdk](https://github.com/Microsoft/cpprestsdk). Thus, any issues with their code will affect this feature. However, given that cpprestsdk is a Microsoft project, we can expect a level of reliability and also solve issues internally if needed.
+2. The proposed authentication endpoint is Azure AD v1.0 instead of Azure AD v2.0 (also known as Microsoft Identity Platform). Azure AD v1.0 is still supported for now, but there is a risk of it becoming deprecated at some point in the future. However, given that it is once again another Microsoft-owned project, we can request support for it through an internal channel. In the worst case, our implementation can switch to Microsoft Identity Platform (which would only requires some minor edits to the http requests). 
+3. The Azure cloud shell API is not public, meaning that implementing this feature in an official capacity would require app permissions from the Azure cloud shell team. This brings about another dependency, but once again issues can be resolved through internal Microsoft channels. 
+
+## Future considerations
+
+This could potentially be the first plugin for Windows Terminal once the app allows for plugins/extensions! 
+
+## Resources
+
+* [Azure AD v1.0](https://docs.microsoft.com/en-us/azure/active-directory/develop/v1-overview)
+* [cpprestsdk](https://github.com/Microsoft/cpprestsdk)
+* [Device code flow](https://github.com/AzureAD/microsoft-authentication-library-for-dotnet/wiki/Device-Code-Flow)
+* [Windows Storage](https://docs.microsoft.com/en-us/uwp/api/windows.storage)
+* [Windows Security Data Protection](https://docs.microsoft.com/en-us/uwp/api/windows.security.cryptography.dataprotection.dataprotectionprovider)

doc/specs/#2325 - Default Profile Settings.md

@@ -0,0 +1,346 @@
+---
+author: Mike Griese @zadjii-msft
+created on: 2019-11-13
+last updated: 2019-12-05
+issue id: #2325
+---
+
+# Default Profile Settings
+
+## Abstract
+
+Oftentimes, users have some common settings that they'd like applied to all of
+their profiles, without needing to manually edit the settings of each of them.
+This doc will cover some of the many proposals on how to expose that
+functionality to the user in our JSON settings model. In this first document,
+we'll examine a number of proposed solutions, as well as state our finalized
+design.
+
+## Inspiration
+
+During the course of the pull request review on [#3369], the original pull
+request for this feature's implementation, it became apparent that the entire
+team has differing opinions on how this feature should be exposed to the user.
+This doc is born from that discussion.
+
+## Solution Proposals
+
+The following are a number of different proposals of different ways to achieve
+the proposed functionality:
+
+1. [`defaultSettings` Profile object in the global settings](#proposal-1-defaultsettings-profile-object-in-the-global-settings)
+2. [`__default__` Profile object in the user's profiles](#proposal-2-__default__-profile-object-in-the-users-profiles)
+3. [Change `profiles` to an object with a `list` of profiles and a `defaults`](#proposal-3-change-profiles-to-an-object-with-a-list-of-profiles-and-a-defaults-object)
+   object
+4. [`inheritFrom` in profiles](#proposal-4-inheritfrom-in-profiles)
+
+### Proposal 1: `defaultSettings` Profile object in the global settings
+
+```json
+{
+    "$schema": "https://aka.ms/terminal-profiles-schema",
+    "defaultProfile": "{61c54bbd-c2c6-5271-96e7-009a87ff44bf}",
+    "defaultSettings":
+    {
+        "useAcrylic": true,
+        "acrylicOpacity": 0.1,
+        "fontFace": "Cascadia Code",
+        "fontSize": 10
+    },
+    "requestedTheme" : "dark",
+    "showTabsInTitlebar" : true,
+    "profiles":
+    [
+        {
+            "guid": "{61c54bbd-c2c6-5271-96e7-009a87ff44bf}",
+            "name": "Windows PowerShell",
+            "commandline": "powershell.exe",
+            "hidden": false
+        },
+        {
+            "guid": "{0caa0dad-35be-5f56-a8ff-afceeeaa6101}",
+            "name": "cmd",
+            "commandline": "cmd.exe",
+            "hidden": false
+        }
+    ],
+    "schemes": [],
+    "keybindings": []
+}
+```
+#### Benefits
+
+##### Clearly encapsulates the default profile settings
+Puts all the default profiles settings in one object. It's immediately obvious
+when scanning the file where the defaults are.
+
+##### Simple to understand
+There's one object that applies to all the subsequent profiles, and that
+object is the `defaultSettings` object.
+
+#### Concerns
+
+##### What do we name this setting?
+People were concerned about the naming of this property. No one has a name that
+we're quite happy with:
+
+* `defaultSettings`: This kinda seems to conflict conceptually with
+  "defaults.json". It's different, but is that obvious?
+* `defaultProfileSettings`: Implies "settings of the default profile"
+* `defaults`: This kinda seems to conflict conceptually with "defaults.json"
+* `baseProfileSettings`: not the worst, but not terribly intuitive
+* Others considered with less enthusiasm
+    - `profiles.defaults`: people don't love the idea of a `.`, but hey, VsCode does it.
+    - `inheritedSettings`
+    - `rootSettings`
+    - `globalSettings`: again maybe conflicts a bit with other concepts/properties
+    - `profileSettings`
+    - `profilePrototype`
+
+##### Why is there this random floating profile in the global settings?
+
+Users may be confused about the purpose of this random `Profile` that's in the
+globals. What's that profile doing there? Is _it_ the default profile?
+
+### Proposal 2: `__default__` Profile object in the user's profiles
+```json
+{
+    "$schema": "https://aka.ms/terminal-profiles-schema",
+    "defaultProfile": "{61c54bbd-c2c6-5271-96e7-009a87ff44bf}",
+    "requestedTheme" : "dark",
+    "showTabsInTitlebar" : true,
+    "profiles":
+    [
+        {
+            "guid": "__default__",
+            "useAcrylic": true,
+            "acrylicOpacity": 0.1,
+            "fontFace": "Cascadia Code",
+            "fontSize": 10
+        },
+        {
+            "guid": "{61c54bbd-c2c6-5271-96e7-009a87ff44bf}",
+            "name": "Windows PowerShell",
+            "commandline": "powershell.exe",
+            "hidden": false
+        },
+        {
+            "guid": "{0caa0dad-35be-5f56-a8ff-afceeeaa6101}",
+            "name": "cmd",
+            "commandline": "cmd.exe",
+            "hidden": false
+        }
+    ],
+    "schemes": [],
+    "keybindings": []
+}
+```
+
+#### Benefits
+##### Encapsulates the default profile settings
+Puts all the default profiles settings in one object. Probably not as clear as
+proposal 1, since it could be _anywhere_ in the list of profiles.
+
+##### Groups default profile settings with profiles
+In this proposal, the default profile is grouped into the same list of objects
+as the other profiles. All the profiles, and the defaults are all under the
+`"profiles"` object. Makes sense.
+
+#### Concerns
+##### Mysterious `__defaults__` GUID
+The only way to _definitively_ identify that this profile is special is by
+giving it a constant string. This string is _not_ a guid, which again, would be
+obvious.
+
+##### Unintuitive
+Adding a profile that has a mysterious `guid` value use that profile as the
+"defaults" is _very_ unintuitive. Nothing aside from documentation would
+indicate to the user "hey, add this magic profile blob to use as defaults across
+all your profiles".
+
+##### Why does this one profile object apply to all the others
+It might be unintuitive that one profile from the list of profiles affects all
+the others.
+
+### Proposal 3: Change `profiles` to an object with a `list` of profiles and a `defaults` object
+
+```json
+{
+    "$schema": "https://aka.ms/terminal-profiles-schema",
+    "defaultProfile": "{61c54bbd-c2c6-5271-96e7-009a87ff44bf}",
+    "requestedTheme" : "dark",
+    "showTabsInTitlebar" : true,
+    "profiles":
+    {
+        "defaults": {
+            "useAcrylic": true,
+            "acrylicOpacity": 0.1,
+            "fontFace": "Cascadia Code",
+            "fontSize": 10
+        },
+        "list":[
+            {
+                "guid": "{61c54bbd-c2c6-5271-96e7-009a87ff44bf}",
+                "name": "Windows PowerShell",
+                "commandline": "powershell.exe",
+                "hidden": false
+            },
+            {
+                "guid": "{0caa0dad-35be-5f56-a8ff-afceeeaa6101}",
+                "name": "cmd",
+                "commandline": "cmd.exe",
+                "hidden": false
+            }
+        ]
+    },
+    "schemes": [],
+    "keybindings": []
+}
+```
+
+#### Benefits
+##### Groups default profile settings with profiles
+In this proposal, the default profile is grouped into the same object as the
+list of profiles. All the profiles, and the defaults are all under the
+`"profiles"` object. Makes sense.
+
+##### Backwards compatible
+Fortunately, we can add this functionality _without breaking the existing
+schema_. With Jsoncpp, we can determine at runtime if an object is an _array_ or
+an _object_. If it's an array, we can fall back to the current behavior, safe in
+our knowledge that there's no defaults object. If the object is an array
+however, we can then dig into the object to find the default profile and the
+list of profiles.
+
+#### Concerns
+##### Substantial schema change
+This is a pretty big delta to the settings schema. Instead of using `profiles`
+as a list of `Profile` objects, it instead becomes an object, with a list inside
+it.
+
+As noted above, we could gracefully upgrade this. If the `profiles` object is a
+list, then we can assume there's no `defaults`. This ensures that user's current
+settings files don't break. This is not a major problem.
+
+##### Adds another level of indentation to all profiles
+Some people just hate having things indented this much. 4 layers of indentation
+is quite a lot.
+
+### Proposal 4: `inheritFrom` in profiles
+
+```json
+{
+    "$schema": "https://aka.ms/terminal-profiles-schema",
+    "defaultProfile": "{61c54bbd-c2c6-5271-96e7-009a87ff44bf}",
+    "requestedTheme" : "dark",
+    "showTabsInTitlebar" : true,
+    "profiles":
+    [
+        {
+            "guid": "{11111111-1111-1111-1111-111111111111}",
+            "hidden": true,
+            "useAcrylic": true,
+            "acrylicOpacity": 0.1,
+            "fontFace": "Cascadia Code",
+            "fontSize": 10
+        },
+        {
+            "guid": "{61c54bbd-c2c6-5271-96e7-009a87ff44bf}",
+            "inheritFrom": "{11111111-1111-1111-1111-111111111111}",
+            "name": "Windows PowerShell",
+            "commandline": "powershell.exe",
+            "hidden": false
+        },
+        {
+            "guid": "{0caa0dad-35be-5f56-a8ff-afceeeaa6101}",
+            "inheritFrom": "{11111111-1111-1111-1111-111111111111}",
+            "name": "cmd",
+            "commandline": "cmd.exe",
+            "hidden": false
+        },
+        {
+            "guid": "{0caa0dad-ffff-5f56-a8ff-afceeeaa6101}",
+            "inheritFrom": "{0caa0dad-35be-5f56-a8ff-afceeeaa6101}",
+            "name": "This is another CMD",
+            "commandline": "cmd.exe /c myCoolScript.bat",
+            "hidden": false
+        }
+    ],
+    "schemes": [],
+    "keybindings": []
+}
+```
+
+#### Benefits
+
+##### Matches the existing settings model without major refactoring
+Simply adding a new property to `Profile` would not majorly alter the structure
+of the file.
+
+##### Property name is unique
+`inheritFrom` is very unique relative to other keys we already have.
+
+##### Powerful
+This lets the user have potentially many layers of settings grouping. These
+layers would let the user separate out common settings however they like,
+without forcing them to a single "default" profile. They could potentially have
+many "default" profiles, e.g.
+* one that's used for all their WSL profiles, with `startingDirectory` set to
+  `~` and `fontFace` set to "Ubuntu Mono"
+* One that's used for all their powershell profiles
+
+etc.
+
+#### Concerns
+
+##### GUIDs are not human friendly
+
+Using the guid in the `inheritFrom` field is the only way to be sure we're
+uniquely identifying profiles. However, guids are notoriously un-friendly. The
+above example manually uses `"{11111111-1111-1111-1111-111111111111}"` as the
+guid of the "default" profile, but inheriting from other profiles with "real"
+GUIDs would be less understandable. Consider the "This is another CMD" case,
+where it's inheriting from the "cmd" profile. That `"inheritFrom"` value does
+not mean at a quick glance "cmd".
+
+##### We have to make sure that there are no cycles as we're layering
+
+This is mostly a technical challenge, but this does make the implementation a
+bit trickier.
+
+##### How does this work with the settings UI?
+
+When the user edits settings for a profile with the UI, do we only place the
+changes in the top-most profile?
+
+How do we communicate in the UI that a profile is inheriting settings from other
+profiles?
+
+##### Harder to mentally parse
+Maybe not as easy to mentally picture how one profile inherits from another. The
+user would probably need to manually build the tree of profile inheritance in
+their own head to understand how a profile gets its settings.
+
+## Conclusions
+
+After discussion the available options, the team has settled on proposal 3. The
+major selling points being:
+* It groups the new "default profile settings" with the rest of the profile
+  settings
+* While being a schema change, it's not a _breaking_ schema change.
+* When looking at the settings, it's easy to understand how they're related
+
+We also like the idea of proposal 4, but felt that it was too heavy-handed of an
+approach for this relatively simple feature. It's been added to the backlog of
+terminal features, tracked in [#3818].
+
+## Resources
+
+* Default Profile for Common Profile Settings (the original issue) [#2325]
+* Add support for "User Default" settings (the original PR) [#3369]
+* Add support for inheriting and overriding another profile's settings [#3818]
+
+<!-- Footnotes -->
+[#2325]: https://github.com/microsoft/terminal/issues/2325
+[#3369]: https://github.com/microsoft/terminal/pull/3369
+[#3818]: https://github.com/microsoft/terminal/issues/3818

doc/specs/#2563 - closeOnExit and TerminalConnection evolution.md

@@ -0,0 +1,107 @@
+---
+author: Dustin Howett @DHowett-MSFT
+created on: 2019-07-19
+last updated: 2019-11-05
+issue id: "#2563"
+---
+
+# Improvements to CloseOnExit
+
+## Abstract
+
+This specification describes an improvement to the `closeOnExit` profile feature and the `ITerminalConnection` interface that will offer greater flexibility and allow us to provide saner defaults in the face of unreliable software.
+
+### Conventions and Terminology
+
+The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED",  "MAY", and "OPTIONAL" in this document are to be interpreted as described in [RFC 2119](https://tools.ietf.org/html/rfc2119).
+
+## Inspiration
+
+Other terminal emulators like ConEmu have a similar feature.
+
+## Solution Design
+
+### `ITerminalConnection` Changes
+
+* The `TerminalConnection` interface will be augmented with an enumerator and a set of events regarding connection state transitions.
+    * enum `TerminalConnection::ConnectionState`
+        * This enum attempts to encompass all potential connection states, even those which do not make sense for a local terminal.
+        * The wide variety of values will be useful to indicate state changes in a user interface.
+        * `NotConnected`: All new connections will start out in this state
+        * `Connecting`: The connection has been initiated, but has not yet completed connecting.
+        * `Connected`: The connection is active.
+        * `Closing`: The connection is being closed (usually by request).
+        * `Closed`: The connection has been closed, either by request or from the remote end terminating successfully.
+        * `Failed`: The connection was unexpectedly terminated.
+    * event `StateChanged(ITerminalConnection, IInspectable)`
+        * (the `IInspectable` argument is recommended and required for a typed event handler, but it will bear no payload.)
+    * event `TerminalDisconnected` will be removed, as it is replaced by `StateChanged`
+    * **NOTE**: A conforming implementation MUST treat states as a directed acyclic graph. States MUST NOT be transitioned in reverse.
+* A helper class may be provided for managing state transitions.
+
+### `TerminalControl` Changes
+
+* As the decision as to whether to close a terminal control hosting a connection that has transitioned into a terminal state will be made by the application, the unexpressive `Close` event will be removed and replaced with a `ConnectionStateChanged` event.
+* `event ConnectionStateChanged(TerminalControl, IInspectable)` event will project its connection's `StateChanged` event.
+* TerminalControl's new `ConnectionState` will project its connection's `State`.
+    * (this is indicated for an eventual data binding; see Future Considerations.)
+
+### Application and Settings
+
+1. The existing `closeOnExit` profile key will be replaced with an enumerated string key supporting the following values (behaviors):
+    * `always` - a tab or pane hosting this profile will always be closed when the launched connection reaches a terminal state.
+    * `graceful` - a tab or pane hosting this profile will be closed if and only if the launched connection reaches the `Closed` terminal state.
+    * `never` - a tab or pane hosting this profile will not automatically close.
+    * See the Compatibility section for information on the legacy settings transition. 
+    * **The new default value for `closeOnExit` will be `graceful`.**
+2. `Pane` will remain responsible for making the final determination as to whether it is closed based on the settings of the profile it is hosting.
+
+## UI/UX Design
+
+* The existing `ITerminalConnection` implementations will be augmented to print out interesting and useful status information when they transition into a `Closed` or `Failed` state.
+    * Example (ConPTY connection)
+        * The pseudoconsole cannot be opened, or the process fails to launch.<br>`[failed to spawn 'thing': 0x80070002]`, transition to `Failed`.
+        * The process exited unexpectedly.<br>`[process exited with code 300]`, transition to `Failed`.
+        * The process exited normally.<br>`[process exited with code 0]`, transition to `Closed`.
+* _The final message will always be printed_ regardless of user configuration.
+* If the user's settings specify `closeOnExit: never/false`, the terminal hosting the connection will never be automatically closed. The message will remain on-screen.
+* If the user's settings specify `closeOnExit: graceful/true`, the terminal hosting the connection _will_ automatically be closed if the connection's state is `Closed`. A connection in the `Failed` state will not be closed, and the message will remain on-screen.
+* If the user's settings specify `closeOnExit: always`, the terminal hosting the connection will be closed. The message will not be seen.
+
+## Capabilities
+
+### Accessibility
+
+This will give users of all technologies a way to know when their shell has failed to launch or has exited with an unexpected status code.
+
+### Security
+
+There will be no impact to security.
+
+### Reliability
+
+Windows Terminal will no longer immediately terminate on startup if the user's shell doesn't exist.
+
+### Compatibility
+
+There is an existing `closeOnExit` _boolean_ key that a user may have configured in profiles.json. The boolean values should map as follows:
+
+* `true` -> `graceful`
+* `false` -> `never`
+
+This will make for a clean transition to Windows Terminal's sane new defaults.
+
+### Performance, Power, and Efficiency
+
+## Potential Issues
+
+There will be no impact to Performance, Power or Efficiency.
+
+## Future considerations
+
+* Eventually, we may want to implement a feature like "only close on graceful exit if the shell was running for more than X seconds". This puts us in a better position to do that, as we can detect graceful and clumsy exits more readily.
+    * (potential suggestion: `{ "closeOnExit": "10s" }`
+* The enumerator values for transitioning connection states will be useful for connections that require internet access.
+* Since the connection states are exposed through `TerminalControl`, they should be able to be data-bound to other Xaml elements. This can be used to provide discrete UI states for terminal controls, panes or tabs _hosting_ terminal controls.
+    * Example: a tab hosting a terminal control whose connection has been broken MAY display a red border.
+    * Example: an inactive tab that reaches the `Connected` state MAY flash to indicate that it is ready.

doc/specs/#532 - Panes and Split Windows.md

@@ -0,0 +1,234 @@
+---
+author: "Mike Griese @zadjii-msft"
+created on: 2019-05-16
+last updated: 2019-07-07
+issue id: 523
+---
+
+# Panes in the Windows Terminal
+
+## Abstract
+
+Panes are an abstraction by which the terminal can display multiple terminal
+instances simultaneously in a single terminal window. While tabs allow for a
+single terminal window to have many terminal sessions running simultaneously
+within a single window, only one tab can be visible at a time. Panes, on the
+other hand, allow a user to have many different terminal sessions visible to the
+user within the context of a single window at the same time. This can enable
+greater productivity from the user, as they can see the output of one terminal
+window while working in another.
+
+This spec will help outline the design of the implementation of panes in the
+Windows Terminal.
+
+## Inspirations
+
+Panes within the context of a single terminal window are not a new idea. The
+design of the panes for the Windows Terminal was heavily inspired by the
+application `tmux`, which is a commandline application which acts as a "terminal
+multiplexer", allowing for the easy management of many terminal sessions from a
+single application.
+
+Other applications that include pane-like functionality include (but are not
+limited to):
+
+* screen
+* terminator
+* emacs & vim
+* Iterm2
+
+## Design
+
+The architecture of the Windows Terminal can be broken into two main pieces:
+Tabs and Panes. The Windows Terminal supports _top-level_ tabs, with nested
+panes inside the tabs. This means that there's a single strip of tabs along the
+application, and each tab has a set of panes that are visible within the context
+of that tab.
+
+Panes are implemented as a binary tree of panes. A Pane can either be a leaf
+pane, with it's own terminal control that it displays, or it could be a parent
+pane, where it has two children, each with their own terminal control.
+
+When a pane is a parent, its two children are either split vertically or
+horizontally. Parent nodes don't have a terminal of their own, they merely
+display the terminals of their children.
+
+ * If a Pane is split vertically, the two panes are separated by a vertical
+   split, as to appear side-by-side. Think `[|]`
+ * If a Pane is split horizontally, the two panes are split by a horizontal
+   separator, and appear above/below one another. Think `[-]`.
+
+As additional panes are created, panes will continue to subdivide the space of
+their parent. It's up to the parent pane to control the sizing and display of
+it's children.
+
+### Example
+
+We'll start by taking the terminal and creating a single vertical split. There
+are now two panes in the terminal, side by side. The original terminal is `A`,
+and the newly created one is `B`. The terminal now looks like this:
+
+```
+  +---------------+
+  |       |       |    1: parent [|]
+  |       |       |    ├── 2: A
+  |       |       |    └── 3: B
+  |   A   |   B   |
+  |       |       |
+  |       |       |
+  |       |       |
+  +---------------+
+```
+
+Here, there are actually 3 nodes: 1 is the parent of both 2 and 3. 2 is the node
+containing the `A` terminal, and 3 is the node with the `B` terminal.
+
+
+We could now split `B` in two horizontally, creating a third terminal pane `C`.
+
+```
+  +---------------+
+  |       |       |    1: parent [|]
+  |       |   B   |    ├── 2: A
+  |       |       |    └── 3: parent [-]
+  |   A   +-------+        ├── 4: B
+  |       |       |        └── 5: C
+  |       |   C   |
+  |       |       |
+  +---------------+
+```
+
+Node 3 is now a parent node, and the terminal `B` has moved into a new node as a
+sibling of the new terminal `C`.
+
+We could also split `A` in horizontally, creating a fourth terminal pane `D`.
+
+```
+  +---------------+
+  |       |       |    1: parent [|]
+  |   A   |   B   |    ├── 2: parent [-]
+  |       |       |    |   ├── 4: A
+  +-------+-------+    |   └── 5: D
+  |       |       |    └── 3: parent [-]
+  |   D   |   C   |        ├── 4: B
+  |       |       |        └── 5: C
+  +---------------+
+```
+
+While it may appear that there's a single horizontal separator and a single
+vertical separator here, that's not actually the case. Due to the tree-like
+structure of the pane splitting, the horizontal splits exist only between the
+two panes they're splitting. So, the user could move each of the horizontal
+splits independently, without affecting the other set of panes. As an example:
+
+```
+  +---------------+
+  |       |       |
+  |   A   |       |
+  +-------+   B   |
+  |       |       |
+  |   D   |       |
+  |       +-------+
+  |       |   C   |
+  +---------------+
+```
+
+### Creating a pane
+
+In the basic use case, the user will decide to split the currently focused pane.
+The currently focused pane is always a leaf, because as parent's can't be
+focused (they don't have their own terminal). When a user decides to add a new
+pane, the child will:
+
+ 1. Convert into a parent
+ 2. Move its terminal into its first child
+ 3. Split its UI in half, and display each child in one half.
+
+It's up to the app hosting the panes to tell the pane what kind of terminal in
+wants created in the new pane. By default, the new pane will be created with the
+default settings profile.
+
+### While panes are open
+
+When a tab has multiple panes open, only one is the "active" pane. This is the
+pane that was last focused in the tab. If the tab is the currently open tab,
+then this is the pane with the currently focused terminal control. When the user
+brings the tab into focus, the last focused pane is the pane that should become
+focused again.
+
+The tab's state will be updated to reflect the state of it's focused pane. The
+title text and icon of the tab will reflect that of the focused pane. Should the
+focus switch from one pane to another, the tab's text and icon should update to
+reflect the newly focused control. Any additional state that the tab would
+display for a single pane should also be reflected in the tab for a tab with
+multiple panes.
+
+While panes are open, the user should be able to move any split between panes.
+In moving the split, the sizes of the terminal controls should be resized to
+match.
+
+### Closing a pane
+
+A pane can either be closed by the user manually, or when the terminal it's
+attached to raises its ConnectionClosed event. When this happens, we should
+remove this pane from the tree. The parent of the closing pane will have to
+remove the pane as one of it's children. If the sibling of the closing pane is a
+leaf, then the parent should just take all of the state from the remaining pane.
+This will cause the remaining pane's content to expand to take the entire
+boundaries of the parent's pane. If the remaining child was a parent itself,
+then the parent will take both the children of the remaining pane, and make them
+the parent's children, as if the parent node was taken from the tree and
+replaced by the remaining child.
+
+## Future considerations
+
+The Pane implementation isn't complete in it's current form. There are many
+additional things that could be done to improve the user experience. This is by
+no means a comprehensive list.
+
+* [ ] Panes should be resizable with the mouse. The user should be able to drag
+  the separator for a pair of panes, and have the content between them resize as
+  the separator moves.
+* [ ] There's no keyboard shortcut for "ClosePane"
+* [ ] The user should be able to configure what profile is used for splitting a
+  pane. Currently, the default profile is used, but it's possible a user might
+  want to create a new pane with the parent pane's profile.
+* [ ] There should be some sort of UI to indicate that a particular pane is
+  focused, more than just the blinking cursor. `tmux` accomplishes this by
+  colorizing the separators adjacent to the active pane. Another idea is
+  displaying a small outline around the focused pane (like when tabbing through
+  controls on a webpage).
+* [ ] The user should be able to navigate the focus of panes with the keyboard,
+  instead of requiring the mouse.
+* [ ] The user should be able to zoom a pane, to make the pane take the entire
+  size of the terminal window temporarily.
+* [ ] A pane doesn't necessarily need to host a terminal. It could potentially
+  host another UIElement. One could imagine enabling a user to quickly open up a
+  Browser pane to search for a particular string without needing to leave the
+  terminal.
+
+## Footnotes
+
+### Why not top-level panes, and nested tabs?
+
+If each pane were to have it's own set of tabs, then each pane would need to
+reserve screen real estate for a row of tabs. As a user continued to split the
+window, more and more of the screen would be dedicated to just displaying a row
+of tabs, which isn't really the important part of the application, the terminal
+is.
+
+Additionally, if there were top-level panes, once the root was split, it would
+not be possible to move a single pane to be the full size of the window. The
+user would need to somehow close the other panes, to be able to make the split
+the size of the dull window.
+
+One con of this design is that if a control is hosted in a pane, the current
+design makes it hard to move out of a pane into it's own tab, or into another
+pane. This could be solved a number of ways. There could be keyboard shortcuts
+for swapping the positions of tabs, or a shortcut for both "zooming" a tab
+(temporarily making it the full size) or even popping a pane out to it's own
+tab. Additionally, a right-click menu option could be added to do the
+aforementioned actions. Discoverability of these two actions is not as high as
+just dragging a tab from one pane to another; however, it's believed that panes
+are more of a power-user scenario, and power users will not necessarily be
+turned off by the feature's discoverability.