starred/terminal
1
0
Fork 0
mirror of https://github.com/microsoft/terminal.git synced 2026-09-24 07:54:35 +00:00
Code Issues 19.0k Packages Projects Releases 20 Wiki Activity

Reading input byte by byte causes lines with even numbers of characters #21144

New Issue
Closed
opened 2026-01-31 07:34:28 +00:00 by claunia · 2 comments
claunia commented 2026-01-31 07:34:28 +00:00
Owner
Copy Link

Originally created by @paperclover on GitHub (Jan 26, 2024).

Windows Terminal version

1.18.3181.0

Windows build number

10.0.22621.0

Other Software

This bug report is a minimized reproduction of https://github.com/oven-sh/bun/issues/8504.

Steps to reproduce

Demo application "input.c"

#include <stdio.h>
#include <windows.h>

int main()
{
    HANDLE hStdin;
    DWORD len;
    char buf[1];

    hStdin = GetStdHandle(STD_INPUT_HANDLE);
    if (hStdin == INVALID_HANDLE_VALUE)
        printf("GetStdHandle failed with %lu\n", GetLastError());

    while (1)
    {
        if (!ReadFile(hStdin, buf, 1, &len, NULL)) // number of records read
            printf("ReadConsoleInput failed with %lu\n", GetLastError());

        printf("byte: %d\n", buf[0]);
    }
    return 0;
}

(This program is mimicking the behavior of the Zig standard library function std.io.Reader().readUntilDelimiterAlloc, which does a byte by byte read until it hits a specified character)

clang input.c (I used llvm 16.0.4, but I do not think this matters)

Run in Windows terminal. .\a.exe

If you type 0 or an even number of characters, the \n byte 10 is not read. only the \r byte is.

If you type an odd number of characters, everything is as normal.

Expected Behavior

I expect all lines to end with \r\n, to match the behavior of conhost.exe, ConEmu, Visual Studio Code's Integrated terminal, and also various terminals on Linux/MacOS over an ssh connection.

Actual Behavior

my inputs are: 1<return>12<return>123<return><return>, and then an S pressed by mistake when trying to take a screenshot.

image

Originally created by @paperclover on GitHub (Jan 26, 2024). ### Windows Terminal version 1.18.3181.0 ### Windows build number 10.0.22621.0 ### Other Software This bug report is a minimized reproduction of https://github.com/oven-sh/bun/issues/8504. ### Steps to reproduce Demo application "`input.c`" ```c #include <stdio.h> #include <windows.h> int main() { HANDLE hStdin; DWORD len; char buf[1]; hStdin = GetStdHandle(STD_INPUT_HANDLE); if (hStdin == INVALID_HANDLE_VALUE) printf("GetStdHandle failed with %lu\n", GetLastError()); while (1) { if (!ReadFile(hStdin, buf, 1, &len, NULL)) // number of records read printf("ReadConsoleInput failed with %lu\n", GetLastError()); printf("byte: %d\n", buf[0]); } return 0; } ``` (This program is mimicking the behavior of the Zig standard library function `std.io.Reader().readUntilDelimiterAlloc`, which does a byte by byte read until it hits a specified character) `clang input.c` (I used llvm 16.0.4, but I do not think this matters) Run in Windows terminal. `.\a.exe` If you type 0 or an even number of characters, the `\n` byte 10 is not read. only the `\r` byte is. If you type an odd number of characters, everything is as normal. ### Expected Behavior I expect all lines to end with `\r\n`, to match the behavior of conhost.exe, ConEmu, Visual Studio Code's Integrated terminal, and also various terminals on Linux/MacOS over an ssh connection. ### Actual Behavior my inputs are: `1<return>12<return>123<return><return>`, and then an S pressed by mistake when trying to take a screenshot. ![image](https://github.com/microsoft/terminal/assets/24465214/b4c689f4-740b-43ac-b502-5d499f2eba97)
claunia commented 2026-01-31 07:34:30 +00:00
Author
Owner
Copy Link

@lhecker commented on GitHub (Jan 29, 2024):

I've found that this PR fixed the issue: #16313
The issue was caused by this PR: #14745

It's planned to be shipped in the next service release for both 1.18 (Stable) and 1.19 (Preview). I'm sorry that it's been taking some time to ship this! But since the bug only started appearing in 1.18, I personally believe it's fine to not hotfix Zig just yet. (Although I do sort of think that using a buffered reader is a great idea, if it's an option. The ReadFile syscalls are very expensive.)

If you want to test the fix in the meantime, you can find the nightly Canary build here: https://aka.ms/terminal-canary-installer

@lhecker commented on GitHub (Jan 29, 2024): I've found that this PR fixed the issue: #16313 The issue was caused by this PR: #14745 It's planned to be shipped in the next service release for both 1.18 (Stable) and 1.19 (Preview). I'm sorry that it's been taking some time to ship this! But since the bug only started appearing in 1.18, I personally believe it's fine to not hotfix Zig just yet. (Although I do sort of think that using a buffered reader is a great idea, if it's an option. The `ReadFile` syscalls are very expensive.) If you want to test the fix in the meantime, you can find the nightly Canary build here: https://aka.ms/terminal-canary-installer
claunia commented 2026-01-31 07:34:30 +00:00
Author
Owner
Copy Link

@paperclover commented on GitHub (Jan 29, 2024):

Thank you.

In Bun i have changed our code to use a buffered reader for all platforms. This makes the readByte call less expensive since it will buffer an amount already.

It probably makes sense for us to leave that commit in regardless because it improves performance slightly. These are only used for terminal input, where the ReadFile is going to always be blocked on user input (bun init and the prompt() JS api).

Maybe I'll revert my change if there are issues consuming stdin from a user program while using the prompt api, but I suspect this case is very unlikely, and by the time this becomes an issue for anyone, Windows Terminal should be propagated to everyone.

@paperclover commented on GitHub (Jan 29, 2024): Thank you. In Bun i have changed our code to use a buffered reader for all platforms. This makes the `readByte` call less expensive since it will buffer an amount already. It probably makes sense for us to leave that commit in regardless because it improves performance slightly. These are only used for terminal input, where the ReadFile is going to always be blocked on user input (`bun init` and the `prompt()` JS api). Maybe I'll revert my change if there are issues consuming stdin from a user program while using the `prompt` api, but I suspect this case is very unlikely, and by the time this becomes an issue for anyone, Windows Terminal should be propagated to everyone.
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
⛺ Reserved
A11yCO
A11yMAS
A11ySev1
A11ySev2
A11ySev3
A11yTTValidated
A11yUsable
A11yVoiceAccess
A11yWCAG
Area-Accessibility
Area-AtlasEngine
Area-AzureShell
Area-Build
Area-Build
Area-Chat
Area-CmdPal
Area-CodeHealth
Area-Commandline
Area-CookedRead
Area-DefApp
Area-Extensibility
Area-Fonts
Area-GroupPolicy
Area-i18n
Area-Input
Area-Interaction
Area-Interop
Area-Localization
Area-Output
Area-Performance
Area-Portable
Area-Quality
Area-Remoting
Area-Rendering
Area-Schema
Area-Server
Area-Settings
Area-SettingsUI
Area-ShellExtension
Area-ShellExtension
Area-ShellExtension
Area-Suggestions
Area-Suggestions
Area-TerminalConnection
Area-TerminalControl
Area-Theming
Area-UserInterface
Area-VT
Area-Windowing
Area-WPFControl
AutoMerge
Blocking-Ingestion
Culprit-Centennial
Culprit-WinUI
Disability-All
Disability-Blind
Disability-LowVision
Disability-Mobility
External-Blocked-WinUI3
Fixed
Gathering-Data
good first issue
HCL-E+D
HCL-WindowsTerminal
Help Wanted
Impact-Compatibility
Impact-Compliance
Impact-Correctness
Impact-Visual
In-PR
InclusionBacklog
InclusionBacklog-Windows TerminalWin32
InclusionCommitted-202206
Issue-Bug
Issue-Docs
Issue-Feature
Issue-Feature
Issue-Question
Issue-Samples
Issue-Scenario
Issue-Task
Needs-Attention
Needs-Author-Feedback
Needs-Bisect
Needs-Discussion
Needs-Repro
Needs-Tag-Fix
Needs-Tag-Fix
Needs-Triage
No-Recent-Activity
Priority-0
Priority-1
Priority-2
Priority-3
Product-Cmd.exe
Product-Colortool
Product-Colortool
Product-Colortool
Product-Conhost
Product-Conpty
Product-Meta
Product-Powershell
Product-Terminal
Product-WSL
pull-request
Mirrored from GitHub Pull Request
 Resolution-Answered
Resolution-By-Design
Resolution-Duplicate
Resolution-External
Resolution-Fix-Available
Resolution-Fix-Committed
Resolution-No-Repro
Resolution-Won't-Fix
Severity-Blocking
Severity-Crash
Severity-DataLoss
spam
this-will-be-a-breaking-change
Tracking-External
WindowsTerminal_Win32
Work-Item
zAskModeBug
zInbox-Bug
No Label Area-Input Issue-Bug Priority-1 Product-Conhost Resolution-Fix-Committed
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
claunia
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: starred/terminal#21144
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?