starred/terminal
1
0
Fork 0
mirror of https://github.com/microsoft/terminal.git synced 2026-02-03 21:25:34 +00:00
Code Issues 19.0k Packages Projects Releases 20 Wiki Activity

ConPTY could optimize some newline movement #17822

New Issue
Closed
opened 2026-01-31 05:55:22 +00:00 by claunia · 6 comments
claunia commented 2026-01-31 05:55:22 +00:00
Owner
Copy Link

Originally created by @Tancen on GitHub (Jun 29, 2022).

Windows Terminal version

1.13.11432.0

Windows build number

10.0.22000.739

Other Software

No response

Steps to reproduce

My program(test5.exe) loads the file 'tmp2.txt' and prints each character, At the end of the file, The terminal processing characters '\1b[34;1H' incorrect, It doesn't jump to line 34, Line 33 jumped to. And some characters of line 33 were overwritten.

image

image

Codes of the test:

#include <Windows.h>
#include <stdio.h>
#include <assert.h>

int main(int argc, char* argv[])
{
    DWORD  settings;
    HANDLE hConsole;
    bool success;

    //output
    hConsole = GetStdHandle(STD_OUTPUT_HANDLE);
    assert(hConsole != INVALID_HANDLE_VALUE);
    success = GetConsoleMode(hConsole, &settings);
    assert(success);
    settings |= ENABLE_VIRTUAL_TERMINAL_PROCESSING;
    success = SetConsoleMode(hConsole, settings);
    assert(success);

    FILE* pf;
    int err = fopen_s(&pf, "e:/tmp2.txt", "rb");
    assert(!err);

    do
    {
        unsigned ch;
        int ret = fread(&ch, 1, 1, pf);
        if (ret <= 0)
            break;
        printf("%c", ch);
    } while (true);
    fclose(pf);
    return 0;
}

Expected Behavior

Output characters to line 34.

Actual Behavior

Output characters to line 33.

Originally created by @Tancen on GitHub (Jun 29, 2022). ### Windows Terminal version 1.13.11432.0 ### Windows build number 10.0.22000.739 ### Other Software _No response_ ### Steps to reproduce My program(test5.exe) loads the file 'tmp2.txt' and prints each character, At the end of the file, The terminal processing characters '\1b[34;1H' incorrect, It doesn't jump to line 34, Line 33 jumped to. And some characters of line 33 were overwritten. ![image](https://user-images.githubusercontent.com/3649637/176457686-d2fb7bd0-c6c3-4b32-b70d-fced3a717f6d.png) ![image](https://user-images.githubusercontent.com/3649637/176457869-d5514a12-fe8f-4d54-b4a2-0a1894ce9c8e.png) Codes of the test: ``` #include <Windows.h> #include <stdio.h> #include <assert.h> int main(int argc, char* argv[]) { DWORD settings; HANDLE hConsole; bool success; //output hConsole = GetStdHandle(STD_OUTPUT_HANDLE); assert(hConsole != INVALID_HANDLE_VALUE); success = GetConsoleMode(hConsole, &settings); assert(success); settings |= ENABLE_VIRTUAL_TERMINAL_PROCESSING; success = SetConsoleMode(hConsole, settings); assert(success); FILE* pf; int err = fopen_s(&pf, "e:/tmp2.txt", "rb"); assert(!err); do { unsigned ch; int ret = fread(&ch, 1, 1, pf); if (ret <= 0) break; printf("%c", ch); } while (true); fclose(pf); return 0; } ``` ### Expected Behavior Output characters to line 34. ### Actual Behavior Output characters to line 33.
claunia added the Area-RenderingIssue-BugNeeds-Tag-FixProduct-ConptyPriority-2 labels 2026-01-31 05:55:23 +00:00
claunia commented 2026-01-31 05:55:25 +00:00
Author
Owner
Copy Link

@WSLUser commented on GitHub (Jun 29, 2022):

So there's https://terminalnuget.blob.core.windows.net/packages/TerminalSequences.html, which lists what WT supports. However it has been awhile since that document was generated and number of changes have been made since then. It would be nice if a new document was generated to list the current support of WT. For example, none of the OSC sequences are even listed such as hyperlink support. WT also now supports DRCS but that isn't listed on the document due to it's age. Adding the ConEmu custom sequences currently supported to that list would also be helpful.

@WSLUser commented on GitHub (Jun 29, 2022): So there's https://terminalnuget.blob.core.windows.net/packages/TerminalSequences.html, which lists what WT supports. However it has been awhile since that document was generated and number of changes have been made since then. It would be nice if a new document was generated to list the current support of WT. For example, none of the OSC sequences are even listed such as hyperlink support. WT also now supports DRCS but that isn't listed on the document due to it's age. Adding the ConEmu custom sequences currently supported to that list would also be helpful.
claunia commented 2026-01-31 05:55:25 +00:00
Author
Owner
Copy Link

@DHowett commented on GitHub (Jun 29, 2022):

Can you share the test file as well? I have a couple questions that it could answer 😄

@DHowett commented on GitHub (Jun 29, 2022): Can you share the test file as well? I have a couple questions that it could answer 😄
claunia commented 2026-01-31 05:55:26 +00:00
Author
Owner
Copy Link

@Tancen commented on GitHub (Jun 30, 2022):

Can you share the test file as well? I have a couple questions that it could answer 😄

tmp2.txt

@Tancen commented on GitHub (Jun 30, 2022): > Can you share the test file as well? I have a couple questions that it could answer 😄 [tmp2.txt](https://github.com/Tancen/tmp/blob/8bb54befcb7ec8671f2d1946e4c28365afb4933b/tmp2.txt)
claunia commented 2026-01-31 05:55:26 +00:00
Author
Owner
Copy Link

@j4james commented on GitHub (Jul 2, 2022):

I'm assuming your code page is 936, but it would help if you could confirm that, because that might be a factor (just type chcp at the cmd prompt to see the active code page).

But I suspect the problem is that your screen height was set to 32 lines or less when you ran the test, because that's the only way I can reproduce the output you're seeing.

The \1b[34;1H escape sequence is trying to move to line 34, but if there aren't 34 lines on the screen it's just going to move down as far as it can go. And at that point in time, the last line is 5 涓洰褰?67,369,209,856 鍙敤瀛楄妭, so that's why that line is overwritten.

@j4james commented on GitHub (Jul 2, 2022): I'm assuming your code page is 936, but it would help if you could confirm that, because that might be a factor (just type `chcp` at the cmd prompt to see the active code page). But I suspect the problem is that your screen height was set to 32 lines or less when you ran the test, because that's the only way I can reproduce the output you're seeing. The `\1b[34;1H` escape sequence is trying to move to line 34, but if there aren't 34 lines on the screen it's just going to move down as far as it can go. And at that point in time, the last line is `5 涓洰褰?67,369,209,856 鍙敤瀛楄妭`, so that's why that line is overwritten.
claunia commented 2026-01-31 05:55:26 +00:00
Author
Owner
Copy Link

@Tancen commented on GitHub (Jul 4, 2022):

I'm assuming your code page is 936, but it would help if you could confirm that, because that might be a factor (just type chcp at the cmd prompt to see the active code page).

But I suspect the problem is that your screen height was set to 32 lines or less when you ran the test, because that's the only way I can reproduce the output you're seeing.

The \1b[34;1H escape sequence is trying to move to line 34, but if there aren't 34 lines on the screen it's just going to move down as far as it can go. And at that point in time, the last line is 5 涓洰褰?67,369,209,856 鍙敤瀛楄妭, so that's why that line is overwritten.

Yes, You are right. I ran the test codes(a little changes for the codes) on Ubuntu 20.04, , It's not pretty also.

image

In fact, 'tmp2.txt' is the output of ConPTY typing some commands after executing 'cmd.exe'.
The output looks likely not good for the test codes:

#include <Windows.h>
#include <stdio.h>
#include <assert.h>
#include <string>
#include <memory>

int main(int argc, char* argv[])
{

    HPCON hPC = 0;
    PROCESS_INFORMATION pi = {0};
    HANDLE outPipeOurSide = nullptr, inPipeOurSide = nullptr;
    HANDLE outPipePseudoConsoleSide = nullptr, inPipePseudoConsoleSide = nullptr;
    int err;
    bool success;

    // create pty
    CreatePipe(&inPipePseudoConsoleSide, &inPipeOurSide, NULL, 0);
    CreatePipe(&outPipeOurSide, &outPipePseudoConsoleSide, NULL, 0);

    err = CreatePseudoConsole(COORD{ 800, 600 }, inPipePseudoConsoleSide, outPipePseudoConsoleSide, 0, &hPC);
    assert(err == S_OK);
    
    STARTUPINFOEXW  si;
    memset(&si, 0, sizeof(si));
    si.StartupInfo.cb = sizeof(STARTUPINFOEXW);

    size_t size;
    InitializeProcThreadAttributeList(NULL, 1, 0, &size);
    si.lpAttributeList = reinterpret_cast<PPROC_THREAD_ATTRIBUTE_LIST>(new BYTE[size]);
    success = InitializeProcThreadAttributeList(si.lpAttributeList, 1, 0, (PSIZE_T)&size);
    assert(success);

    success = UpdateProcThreadAttribute(
        si.lpAttributeList,
        0,
        PROC_THREAD_ATTRIBUTE_PSEUDOCONSOLE,
        hPC,
        sizeof(hPC),
        NULL,
        NULL);
    assert(success);

    std::wstring commandline = L"cmd.exe";
    std::unique_ptr<wchar_t> bufCommandline(new wchar_t[commandline.length() + 1]);
    memcpy(bufCommandline.get(), commandline.c_str(), (commandline.length() + 1) * sizeof(wchar_t));
    success = CreateProcessW(
        nullptr,
        bufCommandline.get(),
        nullptr,
        nullptr,
        TRUE,
        EXTENDED_STARTUPINFO_PRESENT,
        nullptr,
        nullptr,
        &si.StartupInfo,
        &pi);

    assert(success);
    DeleteProcThreadAttributeList(si.lpAttributeList);


    //type command
    std::string cmdListDir = "dir\r\n";
    DWORD l;
    success = WriteFile(inPipeOurSide, cmdListDir.c_str(), cmdListDir.length(), &l, nullptr);
    assert(success && l == cmdListDir.length());

    success = WriteFile(inPipeOurSide, cmdListDir.c_str(), cmdListDir.length(), &l, nullptr);
    assert(success && l == cmdListDir.length());


    DWORD  settings;
    HANDLE hConsole;

    //change code page
    system("chcp  65001 > nul");    // utf-8

    //print output
    hConsole = GetStdHandle(STD_OUTPUT_HANDLE);
    assert(hConsole != INVALID_HANDLE_VALUE);
    success = GetConsoleMode(hConsole, &settings);
    assert(success);
    settings |= ENABLE_VIRTUAL_TERMINAL_PROCESSING;
    success = SetConsoleMode(hConsole, settings);
    assert(success);

    const int BUF_LEN = 4096;
    std::unique_ptr<char> buf(new char[BUF_LEN]);
    do
    {
        DWORD n = 0;
        success = PeekNamedPipe(outPipeOurSide, nullptr, 0, nullptr, &n, nullptr);
        assert(success);

        if (n)
        {
            if (n > BUF_LEN)
                n = BUF_LEN;

            DWORD l;
            success = ReadFile(outPipeOurSide, buf.get(), n, &l, nullptr);
            assert(success);

            for (int i = 0; i < l; i ++)
                printf("%c", buf.get()[i]);
        }
    } while (true);


    CloseHandle(pi.hProcess);

    CloseHandle(outPipeOurSide);
    CloseHandle(inPipeOurSide);
    CloseHandle(outPipePseudoConsoleSide);
    CloseHandle(inPipePseudoConsoleSide);

    ClosePseudoConsole(hPC);
    return 0;
}

image

I tried to simulate the above test and ran the PTY on Linux, print the output on Windows, It's looks every good.
image

I don't know what I missed.

@Tancen commented on GitHub (Jul 4, 2022): > I'm assuming your code page is 936, but it would help if you could confirm that, because that might be a factor (just type `chcp` at the cmd prompt to see the active code page). > > But I suspect the problem is that your screen height was set to 32 lines or less when you ran the test, because that's the only way I can reproduce the output you're seeing. > > The `\1b[34;1H` escape sequence is trying to move to line 34, but if there aren't 34 lines on the screen it's just going to move down as far as it can go. And at that point in time, the last line is `5 涓洰褰?67,369,209,856 鍙敤瀛楄妭`, so that's why that line is overwritten. Yes, You are right. I ran the test codes(a little changes for the codes) on Ubuntu 20.04, , It's not pretty also. ![image](https://user-images.githubusercontent.com/3649637/177158041-94d47df2-2d98-4933-bded-23b2cc1b24a0.png) In fact, 'tmp2.txt' is the output of ConPTY typing some commands after executing 'cmd.exe'. The output looks likely not good for the test codes: ``` #include <Windows.h> #include <stdio.h> #include <assert.h> #include <string> #include <memory> int main(int argc, char* argv[]) { HPCON hPC = 0; PROCESS_INFORMATION pi = {0}; HANDLE outPipeOurSide = nullptr, inPipeOurSide = nullptr; HANDLE outPipePseudoConsoleSide = nullptr, inPipePseudoConsoleSide = nullptr; int err; bool success; // create pty CreatePipe(&inPipePseudoConsoleSide, &inPipeOurSide, NULL, 0); CreatePipe(&outPipeOurSide, &outPipePseudoConsoleSide, NULL, 0); err = CreatePseudoConsole(COORD{ 800, 600 }, inPipePseudoConsoleSide, outPipePseudoConsoleSide, 0, &hPC); assert(err == S_OK); STARTUPINFOEXW si; memset(&si, 0, sizeof(si)); si.StartupInfo.cb = sizeof(STARTUPINFOEXW); size_t size; InitializeProcThreadAttributeList(NULL, 1, 0, &size); si.lpAttributeList = reinterpret_cast<PPROC_THREAD_ATTRIBUTE_LIST>(new BYTE[size]); success = InitializeProcThreadAttributeList(si.lpAttributeList, 1, 0, (PSIZE_T)&size); assert(success); success = UpdateProcThreadAttribute( si.lpAttributeList, 0, PROC_THREAD_ATTRIBUTE_PSEUDOCONSOLE, hPC, sizeof(hPC), NULL, NULL); assert(success); std::wstring commandline = L"cmd.exe"; std::unique_ptr<wchar_t> bufCommandline(new wchar_t[commandline.length() + 1]); memcpy(bufCommandline.get(), commandline.c_str(), (commandline.length() + 1) * sizeof(wchar_t)); success = CreateProcessW( nullptr, bufCommandline.get(), nullptr, nullptr, TRUE, EXTENDED_STARTUPINFO_PRESENT, nullptr, nullptr, &si.StartupInfo, &pi); assert(success); DeleteProcThreadAttributeList(si.lpAttributeList); //type command std::string cmdListDir = "dir\r\n"; DWORD l; success = WriteFile(inPipeOurSide, cmdListDir.c_str(), cmdListDir.length(), &l, nullptr); assert(success && l == cmdListDir.length()); success = WriteFile(inPipeOurSide, cmdListDir.c_str(), cmdListDir.length(), &l, nullptr); assert(success && l == cmdListDir.length()); DWORD settings; HANDLE hConsole; //change code page system("chcp 65001 > nul"); // utf-8 //print output hConsole = GetStdHandle(STD_OUTPUT_HANDLE); assert(hConsole != INVALID_HANDLE_VALUE); success = GetConsoleMode(hConsole, &settings); assert(success); settings |= ENABLE_VIRTUAL_TERMINAL_PROCESSING; success = SetConsoleMode(hConsole, settings); assert(success); const int BUF_LEN = 4096; std::unique_ptr<char> buf(new char[BUF_LEN]); do { DWORD n = 0; success = PeekNamedPipe(outPipeOurSide, nullptr, 0, nullptr, &n, nullptr); assert(success); if (n) { if (n > BUF_LEN) n = BUF_LEN; DWORD l; success = ReadFile(outPipeOurSide, buf.get(), n, &l, nullptr); assert(success); for (int i = 0; i < l; i ++) printf("%c", buf.get()[i]); } } while (true); CloseHandle(pi.hProcess); CloseHandle(outPipeOurSide); CloseHandle(inPipeOurSide); CloseHandle(outPipePseudoConsoleSide); CloseHandle(inPipePseudoConsoleSide); ClosePseudoConsole(hPC); return 0; } ``` ![image](https://user-images.githubusercontent.com/3649637/177163696-d069be84-ac47-48f1-9554-1f788fd19202.png) I tried to simulate the above test and ran the PTY on Linux, print the output on Windows, It's looks every good. ![image](https://user-images.githubusercontent.com/3649637/177166662-828555b4-9951-48a8-a728-e414a43354de.png) I don't know what I missed.
claunia commented 2026-01-31 05:55:26 +00:00
Author
Owner
Copy Link

@zadjii-msft commented on GitHub (Aug 1, 2022):

Yea I mean, this definitely just looks like this was ConPTY output that was recorded at one buffer size, and then replayed at another. Theoretically, conpty should be a bit more resistant to something like that, it could maybe output something a bit more...

Hmm. Y'know, I'm pretty confident ConPTY actually should emit \r\n's at the end of lines, rather that positioning the cursor manually. At least in the case here we probably should... Hmm.

I wonder if this is just because conpty is moving two lines down. Because it's not just "on to the next line", we instead choose to optimize, and send the  instead of \r\n\r\n. You can see that other places too, like  near the top.

I suppose it actually always makes sense to optimize out "move two lines down" into nl/cr pairs. Heck, even 3 lines is equally optimal, and more literal. I wonder if there's anything horrifying that would regress if we did that.

For OP's sake: You'll probably have better results making sure that any verbatim conpty output is replayed in a terminal window of the exact same dimensions. ConPTY isn't really meant to be replayed. It's more designed to be used interactively.
@zadjii-msft commented on GitHub (Aug 1, 2022): Yea I mean, this definitely just looks like this was ConPTY output that was recorded at one buffer size, and then replayed at another. Theoretically, conpty should be a bit more resistant to something like that, it could maybe output something a bit more... Hmm. Y'know, I'm pretty confident ConPTY actually should emit `\r\n`'s at the end of lines, rather that positioning the cursor manually. At least in the case here we probably should... Hmm. I wonder if this is just because conpty is moving _two_ lines down. Because it's not just "on to the next line", we instead choose to optimize, and send the `` instead of `\r\n\r\n`. You can see that other places too, like `` near the top. I suppose it actually always makes sense to optimize out "move two lines down" into nl/cr pairs. Heck, even 3 lines is equally optimal, and more literal. **I wonder if there's anything horrifying that would regress if we did that**. <hr> For OP's sake: You'll probably have better results making sure that any verbatim conpty output is replayed in a terminal window of the exact same dimensions. ConPTY isn't really meant to be replayed. It's more designed to be used interactively.
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-Rendering Issue-Bug Needs-Tag-Fix Priority-2 Product-Conpty
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#17822
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?