terminal/src/cascadia/TerminalCore/Terminal.cpp

1301 lines
51 KiB
C++
Raw Normal View History

// Copyright (c) Microsoft Corporation.
// Licensed under the MIT license.
#include "pch.h"
#include "Terminal.hpp"
#include "../../terminal/parser/OutputStateMachineEngine.hpp"
#include "TerminalDispatch.hpp"
#include "../../inc/unicode.hpp"
#include "../../inc/argb.h"
#include "../../types/inc/utils.hpp"
#include "../../types/inc/colorTable.hpp"
Rename `Microsoft.Terminal.TerminalControl` to `.Control`; Split into dll & lib (#9472) **BE NOT AFRAID**. I know that there's 107 files in this PR, but almost all of it is just find/replacing `TerminalControl` with `Control`. This is the start of the work to move TermControl into multiple pieces, for #5000. The PR starts this work by: * Splits `TerminalControl` into separate lib and dll projects. We'll want control tests in the future, and for that, we'll need a lib. * Moves `ICoreSettings` back into the `Microsoft.Terminal.Core` namespace. We'll have other types in there soon too. * I could not tell you why this works suddenly. New VS versions? New cppwinrt version? Maybe we're just better at dealing with mdmerge bugs these days. * RENAMES `Microsoft.Terminal.TerminalControl` to `Microsoft.Terminal.Control`. This touches pretty much every file in the sln. Sorry about that (not sorry). An upcoming PR will move much of the logic in TermControl into a new `ControlCore` class that we'll add in `Microsoft.Terminal.Core`. `ControlCore` will then be unittest-able in the `UnitTests_TerminalCore`, which will help prevent regressions like #9455 ## Detailed Description of the Pull Request / Additional comments You're really gonna want to clean the sln first, then merge this into your branch, then rebuild. It's very likely that old winmds will get left behind. If you see something like ``` Error MDM2007 Cannot create type Microsoft.Terminal.TerminalControl.KeyModifiers in read-only metadata file Microsoft.Terminal.TerminalControl. ``` then that's what happened to you.
2021-03-17 21:47:24 +01:00
#include <winrt/Microsoft.Terminal.Core.h>
Rename `Microsoft.Terminal.TerminalControl` to `.Control`; Split into dll & lib (#9472) **BE NOT AFRAID**. I know that there's 107 files in this PR, but almost all of it is just find/replacing `TerminalControl` with `Control`. This is the start of the work to move TermControl into multiple pieces, for #5000. The PR starts this work by: * Splits `TerminalControl` into separate lib and dll projects. We'll want control tests in the future, and for that, we'll need a lib. * Moves `ICoreSettings` back into the `Microsoft.Terminal.Core` namespace. We'll have other types in there soon too. * I could not tell you why this works suddenly. New VS versions? New cppwinrt version? Maybe we're just better at dealing with mdmerge bugs these days. * RENAMES `Microsoft.Terminal.TerminalControl` to `Microsoft.Terminal.Control`. This touches pretty much every file in the sln. Sorry about that (not sorry). An upcoming PR will move much of the logic in TermControl into a new `ControlCore` class that we'll add in `Microsoft.Terminal.Core`. `ControlCore` will then be unittest-able in the `UnitTests_TerminalCore`, which will help prevent regressions like #9455 ## Detailed Description of the Pull Request / Additional comments You're really gonna want to clean the sln first, then merge this into your branch, then rebuild. It's very likely that old winmds will get left behind. If you see something like ``` Error MDM2007 Cannot create type Microsoft.Terminal.TerminalControl.KeyModifiers in read-only metadata file Microsoft.Terminal.TerminalControl. ``` then that's what happened to you.
2021-03-17 21:47:24 +01:00
using namespace winrt::Microsoft::Terminal::Core;
using namespace Microsoft::Terminal::Core;
using namespace Microsoft::Console;
using namespace Microsoft::Console::Render;
using namespace Microsoft::Console::Types;
using namespace Microsoft::Console::VirtualTerminal;
using PointTree = interval_tree::IntervalTree<til::point, size_t>;
static std::wstring _KeyEventsToText(std::deque<std::unique_ptr<IInputEvent>>& inEventsToWrite)
{
std::wstring wstr = L"";
for (const auto& ev : inEventsToWrite)
{
if (ev->EventType() == InputEventType::KeyEvent)
{
const auto& k = static_cast<KeyEvent&>(*ev);
const auto wch = k.GetCharData();
wstr += wch;
}
}
return wstr;
}
#pragma warning(suppress : 26455) // default constructor is throwing, too much effort to rearrange at this time.
Terminal::Terminal() :
_mutableViewport{ Viewport::Empty() },
_title{},
_colorTable{},
_defaultFg{ RGB(255, 255, 255) },
_defaultBg{ ARGB(0, 0, 0, 0) },
Add support for DECSCNM in Windows Terminal (#6809) ## Summary of the Pull Request This PR adds full support for the `DECSCNM` reverse screen mode in the Windows Terminal to align with the implementation in conhost. ## References * The conhost implementation of `DECSCNM` was in PR #3817. * WT originally inherited that functionality via the colors being passed through, but that behaviour was lost in PR #6506. ## PR Checklist * [x] Closes #6622 * [x] CLA signed. * [ ] Tests added/passed * [ ] Documentation updated. If checked, please file a pull request on [our docs repo](https://github.com/MicrosoftDocs/terminal) and link it here: #xxx * [ ] Schema updated. * [x] 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: #6622 ## Detailed Description of the Pull Request / Additional comments The `AdaptDispatch::SetScreenMode` now checks if it's in conpty mode and simply returns false to force a pass-through of the mode change. And the `TerminalDispatch` now has its own `SetScreenMode` implementation that tracks any changes to the reversed state, and triggers a redraw in the renderer. To make the renderer work, we just needed to update the `GetForegroundColor` and `GetBackgroundColor` methods of the terminal's `IRenderData` implementation to check the reversed state, and switch the colors being calculated, the same way the `LookupForegroundColor` and `LookupBackgroundColor` methods work in the conhost `Settings` class. ## Validation Steps Performed I've manually tested the `DECSCNM` functionality for Windows Terminal in Vttest, and also with some of my own test scripts.
2020-07-09 13:25:30 +02:00
_screenReversed{ false },
_pfnWriteInput{ nullptr },
_scrollOffset{ 0 },
_snapOnInput{ true },
_altGrAliasing{ true },
Move rect expansion to textbuffer; refactor selection code (#4560) - When performing chunk selection, the expansion now occurs at the time of the selection, not the rendering of the selection - `GetSelectionRects()` was moved to the `TextBuffer` and is now shared between ConHost and Windows Terminal - Some of the selection variables were renamed for clarity - Selection COORDs are now in the Text Buffer coordinate space - Fixes an issue with Shift+Click after performing a Multi-Click Selection ## References This also contributes to... - #4509: UIA Box Selection - #2447: UIA Signaling for Selection - #1354: UIA support for Wide Glyphs Now that the expansion occurs at before render-time, the selection anchors are an accurate representation of what is selected. We just need to move `GetText` to the `TextBuffer`. Then we can have those three issues just rely on code from the text buffer. This also means ConHost gets some of this stuff for free 😀 ### TextBuffer - `GetTextRects` is the abstracted form of `GetSelectionRects` - `_ExpandTextRow` is still needed to handle wide glyphs properly ### Terminal - Rename... - `_boxSelection` --> `_blockSelection` for consistency with ConHost - `_selectionAnchor` --> `_selectionStart` for consistency with UIA - `_endSelectionPosition` --> `_selectionEnd` for consistency with UIA - Selection anchors are in Text Buffer coordinates now - Really rely on `SetSelectionEnd` to accomplish appropriate chunk selection and shift+click actions ## Validation Steps Performed - Shift+Click - Multi-Click --> Shift+Click - Chunk Selection at... - top of buffer - bottom of buffer - random region in scrollback Closes #4465 Closes #4547
2020-02-28 01:42:26 +01:00
_blockSelection{ false },
_selection{ std::nullopt },
_taskbarState{ 0 },
_taskbarProgress{ 0 },
Add an ENUM setting for disabling rendering "intense" text as bold (#10759) ## Summary of the Pull Request This adds a new setting `intenseTextStyle`. It's a per-appearance, control setting, defaulting to `"all"`. * When set to `"all"` or `["bold", "bright"]`, then we'll render text as both **bold** and bright (1.10 behavior) * When set to `"bold"`, `["bold"]`, we'll render text formatted with `^[[1m` as **bold**, but not bright * When set to `"bright"`, `["bright"]`, we'll render text formatted with `^[[1m` as bright, but not bold. This is the pre 1.10 behavior * When set to `"none"`, we won't do anything special for it at all. ## references * I last did this in #10648. This time it's an enum, so we can add bright in the future. It's got positive wording this time. * ~We will want to add `"bright"` as a value in the future, to disable the auto intense->bright conversion.~ I just did that now. * #5682 is related ## PR Checklist * [x] Closes #10576 * [x] I seriously don't think we have an issue for "disable intense is bright", but I'm not crazy, people wanted that, right? https://github.com/microsoft/terminal/issues/2916#issuecomment-544880423 was the closest * [x] I work here * [x] Tests added/passed * [x] https://github.com/MicrosoftDocs/terminal/pull/381 ## Validation Steps Performed <!-- ![image](https://user-images.githubusercontent.com/18356694/125480327-07f6b711-6bca-4c1b-9a76-75fc978c702d.png) --> ![image](https://user-images.githubusercontent.com/18356694/128929228-504933ee-cf50-43a2-9982-55110ba39191.png) Yea that works. Printed some bold text, toggled it on, the text was no longer bold. hooray. ### EDIT, 10 Aug ```json "intenseTextStyle": "none", "intenseTextStyle": "bold", "intenseTextStyle": "bright", "intenseTextStyle": "all", "intenseTextStyle": ["bold", "bright"], ``` all work now. Repro script: ```sh printf "\e[1m[bold]\e[m[normal]\e[34m[blue]\e[1m[bold blue]\e[m\n" ```
2021-08-16 15:45:56 +02:00
_trimBlockSelection{ false },
_intenseIsBright{ true },
_adjustIndistinguishableColors{ true }
{
auto dispatch = std::make_unique<TerminalDispatch>(*this);
auto engine = std::make_unique<OutputStateMachineEngine>(std::move(dispatch));
_stateMachine = std::make_unique<StateMachine>(std::move(engine));
Disable the acceptance of C1 control codes by default (#11690) There are some code pages with "unmapped" code points in the C1 range, which results in them being translated into Unicode C1 control codes, even though that is not their intended use. To avoid having these characters triggering unintentional escape sequences, this PR now disables C1 controls by default. Switching to ISO-2022 encoding will re-enable them, though, since that is the most likely scenario in which they would be required. They can also be explicitly enabled, even in UTF-8 mode, with the `DECAC1` escape sequence. What I've done is add a new mode to the `StateMachine` class that controls whether C1 code points are interpreted as control characters or not. When disabled, these code points are simply dropped from the output, similar to the way a `NUL` is interpreted. This isn't exactly the way they were handled in the v1 console (which I think replaces them with the font _notdef_ glyph), but it matches the XTerm behavior, which seems more appropriate considering this is in VT mode. And it's worth noting that Windows Explorer seems to work the same way. As mentioned above, the mode can be enabled by designating the ISO-2022 coding system with a `DOCS` sequence, and it will be disabled again when UTF-8 is designated. You can also enable it explicitly with a `DECAC1` sequence (originally this was actually a DEC printer sequence, but it doesn't seem unreasonable to use it in a terminal). I've also extended the operations that save and restore "cursor state" (e.g. `DECSC` and `DECRC`) to include the state of the C1 parser mode, since it's closely tied to the code page and character sets which are also saved there. Similarly, when a `DECSTR` sequence resets the code page and character sets, I've now made it reset the C1 mode as well. I should note that the new `StateMachine` mode is controlled via a generic `SetParserMode` method (with a matching API in the `ConGetSet` interface) to allow for easier addition of other modes in the future. And I've reimplemented the existing ANSI/VT52 mode in terms of these generic methods instead of it having to have its own separate APIs. ## Validation Steps Performed Some of the unit tests for OSC sequences were using a C1 `0x9C` for the string terminator, which doesn't work by default anymore. Since that's not a good practice anyway, I thought it best to change those to a standard 7-bit terminator. However, in tests that were explicitly validating the C1 controls, I've just enabled the C1 parser mode at the start of the tests in order to get them working again. There were also some ANSI mode adapter tests that had to be updated to account for the fact that it has now been reimplemented in terms of the `SetParserMode` API. I've added a new state machine test to validate the changes in behavior when the C1 parser mode is enabled or disabled. And I've added an adapter test to verify that the `DesignateCodingSystems` and `AcceptC1Controls` methods toggle the C1 parser mode as expected. I've manually verified the test cases in #10069 and #10310 to confirm that they're no longer triggering control sequences by default. Although, as I explained above, the C1 code points are completely dropped from the output rather than displayed as _notdef_ glyphs. I think this is a reasonable compromise though. Closes #10069 Closes #10310
2021-11-18 00:40:31 +01:00
// Until we have a true pass-through mode (GH#1173), the decision as to
// whether C1 controls are interpreted or not is made at the conhost level.
// If they are being filtered out, then we will simply never receive them.
// But if they are being accepted by conhost, there's a chance they may get
// passed through in some situations, so it's important that our state
// machine is always prepared to accept them.
_stateMachine->SetParserMode(StateMachine::Mode::AcceptC1, true);
auto passAlongInput = [&](std::deque<std::unique_ptr<IInputEvent>>& inEventsToWrite) {
if (!_pfnWriteInput)
{
return;
}
std::wstring wstr = _KeyEventsToText(inEventsToWrite);
_pfnWriteInput(wstr);
};
_terminalInput = std::make_unique<TerminalInput>(passAlongInput);
_InitializeColorTable();
}
void Terminal::Create(COORD viewportSize, SHORT scrollbackLines, IRenderTarget& renderTarget)
{
_mutableViewport = Viewport::FromDimensions({ 0, 0 }, viewportSize);
_scrollbackLines = scrollbackLines;
const COORD bufferSize{ viewportSize.X,
Utils::ClampToShortMax(viewportSize.Y + scrollbackLines, 1) };
const TextAttribute attr{};
const UINT cursorSize = 12;
_buffer = std::make_unique<TextBuffer>(bufferSize, attr, cursorSize, renderTarget);
}
// Method Description:
2019-05-28 18:53:03 +02:00
// - Initializes the Terminal from the given set of settings.
// Arguments:
// - settings: the set of CoreSettings we need to use to initialize the terminal
// - renderTarget: A render target the terminal can use for paint invalidation.
void Terminal::CreateFromSettings(ICoreSettings settings,
IRenderTarget& renderTarget)
{
const COORD viewportSize{ Utils::ClampToShortMax(settings.InitialCols(), 1),
Utils::ClampToShortMax(settings.InitialRows(), 1) };
// TODO:MSFT:20642297 - Support infinite scrollback here, if HistorySize is -1
Create(viewportSize, Utils::ClampToShortMax(settings.HistorySize(), 0), renderTarget);
UpdateSettings(settings);
}
// Method Description:
// - Update our internal properties to match the new values in the provided
// CoreSettings object.
// Arguments:
// - settings: an ICoreSettings with new settings values for us to use.
void Terminal::UpdateSettings(ICoreSettings settings)
{
UpdateAppearance(settings);
_snapOnInput = settings.SnapOnInput();
_altGrAliasing = settings.AltGrAliasing();
_wordDelimiters = settings.WordDelimiters();
_suppressApplicationTitle = settings.SuppressApplicationTitle();
_startingTitle = settings.StartingTitle();
_trimBlockSelection = settings.TrimBlockSelection();
_terminalInput->ForceDisableWin32InputMode(settings.ForceVTInput());
if (settings.TabColor() == nullptr)
{
_tabColor = std::nullopt;
}
else
{
_tabColor = til::color{ settings.TabColor().Value() }.with_alpha(0xff);
}
if (!_startingTabColor && settings.StartingTabColor())
{
_startingTabColor = til::color{ settings.StartingTabColor().Value() }.with_alpha(0xff);
}
if (_pfnTabColorChanged)
{
_pfnTabColorChanged(GetTabColor());
}
// TODO:MSFT:21327402 - if HistorySize has changed, resize the buffer so we
// have a smaller scrollback. We should do this carefully - if the new buffer
// size is smaller than where the mutable viewport currently is, we'll want
// to make sure to rotate the buffer contents upwards, so the mutable viewport
// remains at the bottom of the buffer.
if (_buffer)
{
// Clear the patterns first
_buffer->ClearPatternRecognizers();
if (settings.DetectURLs())
{
// Add regex pattern recognizers to the buffer
// For now, we only add the URI regex pattern
_hyperlinkPatternId = _buffer->AddPatternRecognizer(linkPattern);
UpdatePatternsUnderLock();
}
else
{
ClearPatternTree();
}
}
}
// Method Description:
// - Update our internal properties to match the new values in the provided
// CoreAppearance object.
// Arguments:
// - appearance: an ICoreAppearance with new settings values for us to use.
void Terminal::UpdateAppearance(const ICoreAppearance& appearance)
{
// Set the default background as transparent to prevent the
// DX layer from overwriting the background image or acrylic effect
til::color newBackgroundColor{ appearance.DefaultBackground() };
_defaultBg = newBackgroundColor.with_alpha(0);
_defaultFg = appearance.DefaultForeground();
Add an ENUM setting for disabling rendering "intense" text as bold (#10759) ## Summary of the Pull Request This adds a new setting `intenseTextStyle`. It's a per-appearance, control setting, defaulting to `"all"`. * When set to `"all"` or `["bold", "bright"]`, then we'll render text as both **bold** and bright (1.10 behavior) * When set to `"bold"`, `["bold"]`, we'll render text formatted with `^[[1m` as **bold**, but not bright * When set to `"bright"`, `["bright"]`, we'll render text formatted with `^[[1m` as bright, but not bold. This is the pre 1.10 behavior * When set to `"none"`, we won't do anything special for it at all. ## references * I last did this in #10648. This time it's an enum, so we can add bright in the future. It's got positive wording this time. * ~We will want to add `"bright"` as a value in the future, to disable the auto intense->bright conversion.~ I just did that now. * #5682 is related ## PR Checklist * [x] Closes #10576 * [x] I seriously don't think we have an issue for "disable intense is bright", but I'm not crazy, people wanted that, right? https://github.com/microsoft/terminal/issues/2916#issuecomment-544880423 was the closest * [x] I work here * [x] Tests added/passed * [x] https://github.com/MicrosoftDocs/terminal/pull/381 ## Validation Steps Performed <!-- ![image](https://user-images.githubusercontent.com/18356694/125480327-07f6b711-6bca-4c1b-9a76-75fc978c702d.png) --> ![image](https://user-images.githubusercontent.com/18356694/128929228-504933ee-cf50-43a2-9982-55110ba39191.png) Yea that works. Printed some bold text, toggled it on, the text was no longer bold. hooray. ### EDIT, 10 Aug ```json "intenseTextStyle": "none", "intenseTextStyle": "bold", "intenseTextStyle": "bright", "intenseTextStyle": "all", "intenseTextStyle": ["bold", "bright"], ``` all work now. Repro script: ```sh printf "\e[1m[bold]\e[m[normal]\e[34m[blue]\e[1m[bold blue]\e[m\n" ```
2021-08-16 15:45:56 +02:00
_intenseIsBright = appearance.IntenseIsBright();
_adjustIndistinguishableColors = appearance.AdjustIndistinguishableColors();
for (int i = 0; i < 16; i++)
{
_colorTable.at(i) = til::color{ appearance.GetColorTableEntry(i) };
}
if (_adjustIndistinguishableColors)
{
_MakeAdjustedColorArray();
}
CursorType cursorShape = CursorType::VerticalBar;
switch (appearance.CursorShape())
{
case CursorStyle::Underscore:
cursorShape = CursorType::Underscore;
break;
case CursorStyle::FilledBox:
cursorShape = CursorType::FullBox;
break;
case CursorStyle::EmptyBox:
cursorShape = CursorType::EmptyBox;
break;
case CursorStyle::Vintage:
cursorShape = CursorType::Legacy;
break;
case CursorStyle::DoubleUnderscore:
cursorShape = CursorType::DoubleUnderscore;
break;
default:
case CursorStyle::Bar:
cursorShape = CursorType::VerticalBar;
break;
}
if (_buffer)
{
_buffer->GetCursor().SetStyle(appearance.CursorHeight(),
til::color{ appearance.CursorColor() },
cursorShape);
}
_defaultCursorShape = cursorShape;
}
// Method Description:
// - Resize the terminal as the result of some user interaction.
// Arguments:
// - viewportSize: the new size of the viewport, in chars
// Return Value:
// - S_OK if we successfully resized the terminal, S_FALSE if there was
// nothing to do (the viewportSize is the same as our current size), or an
// appropriate HRESULT for failing to resize.
[[nodiscard]] HRESULT Terminal::UserResize(const COORD viewportSize) noexcept
{
const auto oldDimensions = _mutableViewport.Dimensions();
if (viewportSize == oldDimensions)
{
return S_FALSE;
}
const auto dx = ::base::ClampSub(viewportSize.X, oldDimensions.X);
const short newBufferHeight = ::base::ClampAdd(viewportSize.Y, _scrollbackLines);
COORD bufferSize{ viewportSize.X, newBufferHeight };
Add support for "reflow"ing the Terminal buffer (#4741) This PR adds support for "Resize with Reflow" to the Terminal. In conhost, `ResizeWithReflow` is the function that's responsible for reflowing wrapped lines of text as the buffer gets resized. Now that #4415 has merged, we can also implement this in the Terminal. Now, when the Terminal is resized, it will reflow the lines of it's buffer in the same way that conhost does. This means, the terminal will no longer chop off the ends of lines as the buffer is too small to represent them. As a happy side effect of this PR, it also fixed #3490. This was a bug that plagued me during the investigation into this functionality. The original #3490 PR, #4354, tried to fix this bug with some heavy conpty changes. Turns out, that only made things worse, and far more complicated. When I really got to thinking about it, I realized "conhost can handle this right, why can't the Terminal?". Turns out, by adding resize with reflow, I was also able to fix this at the same time. Conhost does a little bit of math after reflowing to attempt to keep the viewport in the same relative place after a reflow. By re-using that logic in the Terminal, I was able to fix #3490. I also included that big ole test from #3490, because everyone likes adding 60 test cases in a PR. ## References * #4200 - this scenario * #405/#4415 - conpty emits wrapped lines, which was needed for this PR * #4403 - delayed EOL wrapping via conpty, which was also needed for this * #4354 - we don't speak of this PR anymore ## PR Checklist * [x] Closes #1465 * [x] Closes #3490 * [x] Closes #4771 * [x] Tests added/passed ## EDIT: Changes to this PR on 5 March 2020 I learned more since my original version of this PR. I wrote that in January, and despite my notes that say it was totally working, it _really_ wasn't. Part of the hard problem, as mentioned in #3490, is that the Terminal might request a resize to (W, H-1), and while conpty is preparing that frame, or before the terminal has received that frame, the Terminal resizes to (W, H-2). Now, there aren't enough lines in the terminal buffer to catch all the lines that conpty is about to emit. When that happens, lines get duplicated in the buffer. From a UX perspective, this certainly looks a lot worse than a couple lost lines. It looks like utter chaos. So I've introduced a new mode to conpty to try and counteract this behavior. This behavior I'm calling "quirky resize". The **TL;DR** of quirky resize mode is that conpty won't emit the entire buffer on a resize, and will trust that the terminal is prepared to reflow it's buffer on it's own. This will enable the quirky resize behavior for applications that are prepared for it. The "quirky resize" is "don't `InvalidateAll` when the terminal resizes". This is added as a quirk as to not regress other terminal applications that aren't prepared for this behavior (gnome-terminal, conhost in particular). For those kinds of terminals, when the buffer is resized, it's just going to lose lines. That's what currently happens for them. When the quirk is enabled, conpty won't repaint the entire buffer. This gets around the "duplicated lines" issue that requesting multiple resizes in a row can cause. However, for these terminals that are unprepared, the conpty cursor might end up in the wrong position after a quirky resize. The case in point is maximizing the terminal. For maximizing (height->50) from a buffer that's 30 lines tall, with the cursor on y=30, this is what happens: * With the quirk disabled, conpty reprints the entire buffer. This is 60 lines that get printed. This ends up blowing away about 20 lines of scrollback history, as the terminal app would have tried to keep the text pinned to the bottom of the window. The term. app moved the viewport up 20 lines, and then the 50 lines of conpty output (30 lines of text, and 20 blank lines at the bottom) overwrote the lines from the scrollback. This is bad, but not immediately obvious, and is **what currently happens**. * With the quirk enabled, conpty doesn't emit any lines, but the actual content of the window is still only in the top 30 lines. However, the terminal app has still moved 20 lines down from the scrollback back into the viewport. So the terminal's cursor is at y=50 now, but conpty's is at 30. This means that the terminal and conpty are out of sync, and there's not a good way of re-syncing these. It's very possible (trivial in `powershell`) that the new output will jump up to y=30 override the existing output in the terminal buffer. The Windows Terminal is already prepared for this quirky behavior, so it doesn't keep the output at the bottom of the window. It shifts it's viewport down to match what conpty things the buffer looks like. What happens when we have passthrough mode and WT is like "I would like quirky resize"? I guess things will just work fine, cause there won't be a buffer behind the passthrough app that the terminal cares about. Sure, in the passthrough case the Terminal could _not_ quirky resize, but the quirky resize won't be wrong.
2020-03-13 01:43:37 +01:00
// This will be used to determine where the viewport should be in the new buffer.
const short oldViewportTop = _mutableViewport.Top();
short newViewportTop = oldViewportTop;
short newVisibleTop = ::base::saturated_cast<short>(_VisibleStartIndex());
// If the original buffer had _no_ scroll offset, then we should be at the
// bottom in the new buffer as well. Track that case now.
const bool originalOffsetWasZero = _scrollOffset == 0;
Add support for "reflow"ing the Terminal buffer (#4741) This PR adds support for "Resize with Reflow" to the Terminal. In conhost, `ResizeWithReflow` is the function that's responsible for reflowing wrapped lines of text as the buffer gets resized. Now that #4415 has merged, we can also implement this in the Terminal. Now, when the Terminal is resized, it will reflow the lines of it's buffer in the same way that conhost does. This means, the terminal will no longer chop off the ends of lines as the buffer is too small to represent them. As a happy side effect of this PR, it also fixed #3490. This was a bug that plagued me during the investigation into this functionality. The original #3490 PR, #4354, tried to fix this bug with some heavy conpty changes. Turns out, that only made things worse, and far more complicated. When I really got to thinking about it, I realized "conhost can handle this right, why can't the Terminal?". Turns out, by adding resize with reflow, I was also able to fix this at the same time. Conhost does a little bit of math after reflowing to attempt to keep the viewport in the same relative place after a reflow. By re-using that logic in the Terminal, I was able to fix #3490. I also included that big ole test from #3490, because everyone likes adding 60 test cases in a PR. ## References * #4200 - this scenario * #405/#4415 - conpty emits wrapped lines, which was needed for this PR * #4403 - delayed EOL wrapping via conpty, which was also needed for this * #4354 - we don't speak of this PR anymore ## PR Checklist * [x] Closes #1465 * [x] Closes #3490 * [x] Closes #4771 * [x] Tests added/passed ## EDIT: Changes to this PR on 5 March 2020 I learned more since my original version of this PR. I wrote that in January, and despite my notes that say it was totally working, it _really_ wasn't. Part of the hard problem, as mentioned in #3490, is that the Terminal might request a resize to (W, H-1), and while conpty is preparing that frame, or before the terminal has received that frame, the Terminal resizes to (W, H-2). Now, there aren't enough lines in the terminal buffer to catch all the lines that conpty is about to emit. When that happens, lines get duplicated in the buffer. From a UX perspective, this certainly looks a lot worse than a couple lost lines. It looks like utter chaos. So I've introduced a new mode to conpty to try and counteract this behavior. This behavior I'm calling "quirky resize". The **TL;DR** of quirky resize mode is that conpty won't emit the entire buffer on a resize, and will trust that the terminal is prepared to reflow it's buffer on it's own. This will enable the quirky resize behavior for applications that are prepared for it. The "quirky resize" is "don't `InvalidateAll` when the terminal resizes". This is added as a quirk as to not regress other terminal applications that aren't prepared for this behavior (gnome-terminal, conhost in particular). For those kinds of terminals, when the buffer is resized, it's just going to lose lines. That's what currently happens for them. When the quirk is enabled, conpty won't repaint the entire buffer. This gets around the "duplicated lines" issue that requesting multiple resizes in a row can cause. However, for these terminals that are unprepared, the conpty cursor might end up in the wrong position after a quirky resize. The case in point is maximizing the terminal. For maximizing (height->50) from a buffer that's 30 lines tall, with the cursor on y=30, this is what happens: * With the quirk disabled, conpty reprints the entire buffer. This is 60 lines that get printed. This ends up blowing away about 20 lines of scrollback history, as the terminal app would have tried to keep the text pinned to the bottom of the window. The term. app moved the viewport up 20 lines, and then the 50 lines of conpty output (30 lines of text, and 20 blank lines at the bottom) overwrote the lines from the scrollback. This is bad, but not immediately obvious, and is **what currently happens**. * With the quirk enabled, conpty doesn't emit any lines, but the actual content of the window is still only in the top 30 lines. However, the terminal app has still moved 20 lines down from the scrollback back into the viewport. So the terminal's cursor is at y=50 now, but conpty's is at 30. This means that the terminal and conpty are out of sync, and there's not a good way of re-syncing these. It's very possible (trivial in `powershell`) that the new output will jump up to y=30 override the existing output in the terminal buffer. The Windows Terminal is already prepared for this quirky behavior, so it doesn't keep the output at the bottom of the window. It shifts it's viewport down to match what conpty things the buffer looks like. What happens when we have passthrough mode and WT is like "I would like quirky resize"? I guess things will just work fine, cause there won't be a buffer behind the passthrough app that the terminal cares about. Sure, in the passthrough case the Terminal could _not_ quirky resize, but the quirky resize won't be wrong.
2020-03-13 01:43:37 +01:00
Show a double width cursor for double width characters (#5319) # Summary of the Pull Request This PR will allow the cursor to be double width when on top of a double width character. This required changing `IsCursorDoubleWidth` to check whether the glyph the cursor's on top of is double width. This code is exactly the same as the original PR that addressed this issue in #2932. That one got reverted at some point due to the crashes related to it, but due to a combination of Terminal having come further since that PR and other changes to address use-after-frees, some of the crashes may/may not be relevant now. The ones that seemed to be relevant/repro-able, I attempt to address in this PR. The `IsCursorDoubleWidth` check would fail during the `TextBuffer::Reflow` call inside of `Terminal::UserResize` occasionally, particularly when `newCursor.EndDeferDrawing()` is called. This is because when we tell the newCursor to `EndDefer`, the renderer will attempt to redraw the cursor. As part of this redraw, it'll ask if `IsCursorDoubleWidth`, and if the renderer managed to ask this before `UserResize` swapped out the old buffer with the new one from `Reflow`, the renderer will be asking the old buffer if its out-of-bounds cursor is double width. This was pretty easily repro'd using `cmatrix -u0` and resizing the window like a madman. As a solution, I've moved the Start/End DeferDrawing calls out of `Reflow` and into `UserResize`. This way, I can "clamp" the portion of the code where the newBuffer is getting created and reflowed and swapped into the Terminal buffer, and only allow the renderer to draw once the swap is done. This also means that ConHost's `ResizeWithReflow` needed to change slightly. In addition, I've added a WriteLock to `SetCursorOn`. It was mentioned as a fix for a crash in #2965 (although I can't repro), and I also figured it would be good to try to emulate where ConHost locks with regards to Cursor operations, and this seemed to be one that we were missing. ## PR Checklist * [x] Closes #2713 * [x] CLA signed * [x] Tests added/passed ## Validation Steps Performed Manual validation that the cursor is indeed chonky, added a test case to check that we are correctly saying that the cursor is double width (not too sure if I put it in the right place). Also open to other test case ideas and thoughts on what else I should be careful for since I am quite nervous about what other crashes might occur.
2020-04-15 21:23:06 +02:00
// skip any drawing updates that might occur until we swap _buffer with the new buffer or if we exit early.
_buffer->GetCursor().StartDeferDrawing();
// we're capturing _buffer by reference here because when we exit, we want to EndDefer on the current active buffer.
auto endDefer = wil::scope_exit([&]() noexcept { _buffer->GetCursor().EndDeferDrawing(); });
Add support for "reflow"ing the Terminal buffer (#4741) This PR adds support for "Resize with Reflow" to the Terminal. In conhost, `ResizeWithReflow` is the function that's responsible for reflowing wrapped lines of text as the buffer gets resized. Now that #4415 has merged, we can also implement this in the Terminal. Now, when the Terminal is resized, it will reflow the lines of it's buffer in the same way that conhost does. This means, the terminal will no longer chop off the ends of lines as the buffer is too small to represent them. As a happy side effect of this PR, it also fixed #3490. This was a bug that plagued me during the investigation into this functionality. The original #3490 PR, #4354, tried to fix this bug with some heavy conpty changes. Turns out, that only made things worse, and far more complicated. When I really got to thinking about it, I realized "conhost can handle this right, why can't the Terminal?". Turns out, by adding resize with reflow, I was also able to fix this at the same time. Conhost does a little bit of math after reflowing to attempt to keep the viewport in the same relative place after a reflow. By re-using that logic in the Terminal, I was able to fix #3490. I also included that big ole test from #3490, because everyone likes adding 60 test cases in a PR. ## References * #4200 - this scenario * #405/#4415 - conpty emits wrapped lines, which was needed for this PR * #4403 - delayed EOL wrapping via conpty, which was also needed for this * #4354 - we don't speak of this PR anymore ## PR Checklist * [x] Closes #1465 * [x] Closes #3490 * [x] Closes #4771 * [x] Tests added/passed ## EDIT: Changes to this PR on 5 March 2020 I learned more since my original version of this PR. I wrote that in January, and despite my notes that say it was totally working, it _really_ wasn't. Part of the hard problem, as mentioned in #3490, is that the Terminal might request a resize to (W, H-1), and while conpty is preparing that frame, or before the terminal has received that frame, the Terminal resizes to (W, H-2). Now, there aren't enough lines in the terminal buffer to catch all the lines that conpty is about to emit. When that happens, lines get duplicated in the buffer. From a UX perspective, this certainly looks a lot worse than a couple lost lines. It looks like utter chaos. So I've introduced a new mode to conpty to try and counteract this behavior. This behavior I'm calling "quirky resize". The **TL;DR** of quirky resize mode is that conpty won't emit the entire buffer on a resize, and will trust that the terminal is prepared to reflow it's buffer on it's own. This will enable the quirky resize behavior for applications that are prepared for it. The "quirky resize" is "don't `InvalidateAll` when the terminal resizes". This is added as a quirk as to not regress other terminal applications that aren't prepared for this behavior (gnome-terminal, conhost in particular). For those kinds of terminals, when the buffer is resized, it's just going to lose lines. That's what currently happens for them. When the quirk is enabled, conpty won't repaint the entire buffer. This gets around the "duplicated lines" issue that requesting multiple resizes in a row can cause. However, for these terminals that are unprepared, the conpty cursor might end up in the wrong position after a quirky resize. The case in point is maximizing the terminal. For maximizing (height->50) from a buffer that's 30 lines tall, with the cursor on y=30, this is what happens: * With the quirk disabled, conpty reprints the entire buffer. This is 60 lines that get printed. This ends up blowing away about 20 lines of scrollback history, as the terminal app would have tried to keep the text pinned to the bottom of the window. The term. app moved the viewport up 20 lines, and then the 50 lines of conpty output (30 lines of text, and 20 blank lines at the bottom) overwrote the lines from the scrollback. This is bad, but not immediately obvious, and is **what currently happens**. * With the quirk enabled, conpty doesn't emit any lines, but the actual content of the window is still only in the top 30 lines. However, the terminal app has still moved 20 lines down from the scrollback back into the viewport. So the terminal's cursor is at y=50 now, but conpty's is at 30. This means that the terminal and conpty are out of sync, and there's not a good way of re-syncing these. It's very possible (trivial in `powershell`) that the new output will jump up to y=30 override the existing output in the terminal buffer. The Windows Terminal is already prepared for this quirky behavior, so it doesn't keep the output at the bottom of the window. It shifts it's viewport down to match what conpty things the buffer looks like. What happens when we have passthrough mode and WT is like "I would like quirky resize"? I guess things will just work fine, cause there won't be a buffer behind the passthrough app that the terminal cares about. Sure, in the passthrough case the Terminal could _not_ quirky resize, but the quirky resize won't be wrong.
2020-03-13 01:43:37 +01:00
// First allocate a new text buffer to take the place of the current one.
std::unique_ptr<TextBuffer> newTextBuffer;
try
{
// GH#3848 - Stash away the current attributes the old text buffer is
// using. We'll initialize the new buffer with the default attributes,
// but after the resize, we'll want to make sure that the new buffer's
// current attributes (the ones used for printing new text) match the
// old buffer's.
const auto oldBufferAttributes = _buffer->GetCurrentAttributes();
Add support for "reflow"ing the Terminal buffer (#4741) This PR adds support for "Resize with Reflow" to the Terminal. In conhost, `ResizeWithReflow` is the function that's responsible for reflowing wrapped lines of text as the buffer gets resized. Now that #4415 has merged, we can also implement this in the Terminal. Now, when the Terminal is resized, it will reflow the lines of it's buffer in the same way that conhost does. This means, the terminal will no longer chop off the ends of lines as the buffer is too small to represent them. As a happy side effect of this PR, it also fixed #3490. This was a bug that plagued me during the investigation into this functionality. The original #3490 PR, #4354, tried to fix this bug with some heavy conpty changes. Turns out, that only made things worse, and far more complicated. When I really got to thinking about it, I realized "conhost can handle this right, why can't the Terminal?". Turns out, by adding resize with reflow, I was also able to fix this at the same time. Conhost does a little bit of math after reflowing to attempt to keep the viewport in the same relative place after a reflow. By re-using that logic in the Terminal, I was able to fix #3490. I also included that big ole test from #3490, because everyone likes adding 60 test cases in a PR. ## References * #4200 - this scenario * #405/#4415 - conpty emits wrapped lines, which was needed for this PR * #4403 - delayed EOL wrapping via conpty, which was also needed for this * #4354 - we don't speak of this PR anymore ## PR Checklist * [x] Closes #1465 * [x] Closes #3490 * [x] Closes #4771 * [x] Tests added/passed ## EDIT: Changes to this PR on 5 March 2020 I learned more since my original version of this PR. I wrote that in January, and despite my notes that say it was totally working, it _really_ wasn't. Part of the hard problem, as mentioned in #3490, is that the Terminal might request a resize to (W, H-1), and while conpty is preparing that frame, or before the terminal has received that frame, the Terminal resizes to (W, H-2). Now, there aren't enough lines in the terminal buffer to catch all the lines that conpty is about to emit. When that happens, lines get duplicated in the buffer. From a UX perspective, this certainly looks a lot worse than a couple lost lines. It looks like utter chaos. So I've introduced a new mode to conpty to try and counteract this behavior. This behavior I'm calling "quirky resize". The **TL;DR** of quirky resize mode is that conpty won't emit the entire buffer on a resize, and will trust that the terminal is prepared to reflow it's buffer on it's own. This will enable the quirky resize behavior for applications that are prepared for it. The "quirky resize" is "don't `InvalidateAll` when the terminal resizes". This is added as a quirk as to not regress other terminal applications that aren't prepared for this behavior (gnome-terminal, conhost in particular). For those kinds of terminals, when the buffer is resized, it's just going to lose lines. That's what currently happens for them. When the quirk is enabled, conpty won't repaint the entire buffer. This gets around the "duplicated lines" issue that requesting multiple resizes in a row can cause. However, for these terminals that are unprepared, the conpty cursor might end up in the wrong position after a quirky resize. The case in point is maximizing the terminal. For maximizing (height->50) from a buffer that's 30 lines tall, with the cursor on y=30, this is what happens: * With the quirk disabled, conpty reprints the entire buffer. This is 60 lines that get printed. This ends up blowing away about 20 lines of scrollback history, as the terminal app would have tried to keep the text pinned to the bottom of the window. The term. app moved the viewport up 20 lines, and then the 50 lines of conpty output (30 lines of text, and 20 blank lines at the bottom) overwrote the lines from the scrollback. This is bad, but not immediately obvious, and is **what currently happens**. * With the quirk enabled, conpty doesn't emit any lines, but the actual content of the window is still only in the top 30 lines. However, the terminal app has still moved 20 lines down from the scrollback back into the viewport. So the terminal's cursor is at y=50 now, but conpty's is at 30. This means that the terminal and conpty are out of sync, and there's not a good way of re-syncing these. It's very possible (trivial in `powershell`) that the new output will jump up to y=30 override the existing output in the terminal buffer. The Windows Terminal is already prepared for this quirky behavior, so it doesn't keep the output at the bottom of the window. It shifts it's viewport down to match what conpty things the buffer looks like. What happens when we have passthrough mode and WT is like "I would like quirky resize"? I guess things will just work fine, cause there won't be a buffer behind the passthrough app that the terminal cares about. Sure, in the passthrough case the Terminal could _not_ quirky resize, but the quirky resize won't be wrong.
2020-03-13 01:43:37 +01:00
newTextBuffer = std::make_unique<TextBuffer>(bufferSize,
TextAttribute{},
Add support for "reflow"ing the Terminal buffer (#4741) This PR adds support for "Resize with Reflow" to the Terminal. In conhost, `ResizeWithReflow` is the function that's responsible for reflowing wrapped lines of text as the buffer gets resized. Now that #4415 has merged, we can also implement this in the Terminal. Now, when the Terminal is resized, it will reflow the lines of it's buffer in the same way that conhost does. This means, the terminal will no longer chop off the ends of lines as the buffer is too small to represent them. As a happy side effect of this PR, it also fixed #3490. This was a bug that plagued me during the investigation into this functionality. The original #3490 PR, #4354, tried to fix this bug with some heavy conpty changes. Turns out, that only made things worse, and far more complicated. When I really got to thinking about it, I realized "conhost can handle this right, why can't the Terminal?". Turns out, by adding resize with reflow, I was also able to fix this at the same time. Conhost does a little bit of math after reflowing to attempt to keep the viewport in the same relative place after a reflow. By re-using that logic in the Terminal, I was able to fix #3490. I also included that big ole test from #3490, because everyone likes adding 60 test cases in a PR. ## References * #4200 - this scenario * #405/#4415 - conpty emits wrapped lines, which was needed for this PR * #4403 - delayed EOL wrapping via conpty, which was also needed for this * #4354 - we don't speak of this PR anymore ## PR Checklist * [x] Closes #1465 * [x] Closes #3490 * [x] Closes #4771 * [x] Tests added/passed ## EDIT: Changes to this PR on 5 March 2020 I learned more since my original version of this PR. I wrote that in January, and despite my notes that say it was totally working, it _really_ wasn't. Part of the hard problem, as mentioned in #3490, is that the Terminal might request a resize to (W, H-1), and while conpty is preparing that frame, or before the terminal has received that frame, the Terminal resizes to (W, H-2). Now, there aren't enough lines in the terminal buffer to catch all the lines that conpty is about to emit. When that happens, lines get duplicated in the buffer. From a UX perspective, this certainly looks a lot worse than a couple lost lines. It looks like utter chaos. So I've introduced a new mode to conpty to try and counteract this behavior. This behavior I'm calling "quirky resize". The **TL;DR** of quirky resize mode is that conpty won't emit the entire buffer on a resize, and will trust that the terminal is prepared to reflow it's buffer on it's own. This will enable the quirky resize behavior for applications that are prepared for it. The "quirky resize" is "don't `InvalidateAll` when the terminal resizes". This is added as a quirk as to not regress other terminal applications that aren't prepared for this behavior (gnome-terminal, conhost in particular). For those kinds of terminals, when the buffer is resized, it's just going to lose lines. That's what currently happens for them. When the quirk is enabled, conpty won't repaint the entire buffer. This gets around the "duplicated lines" issue that requesting multiple resizes in a row can cause. However, for these terminals that are unprepared, the conpty cursor might end up in the wrong position after a quirky resize. The case in point is maximizing the terminal. For maximizing (height->50) from a buffer that's 30 lines tall, with the cursor on y=30, this is what happens: * With the quirk disabled, conpty reprints the entire buffer. This is 60 lines that get printed. This ends up blowing away about 20 lines of scrollback history, as the terminal app would have tried to keep the text pinned to the bottom of the window. The term. app moved the viewport up 20 lines, and then the 50 lines of conpty output (30 lines of text, and 20 blank lines at the bottom) overwrote the lines from the scrollback. This is bad, but not immediately obvious, and is **what currently happens**. * With the quirk enabled, conpty doesn't emit any lines, but the actual content of the window is still only in the top 30 lines. However, the terminal app has still moved 20 lines down from the scrollback back into the viewport. So the terminal's cursor is at y=50 now, but conpty's is at 30. This means that the terminal and conpty are out of sync, and there's not a good way of re-syncing these. It's very possible (trivial in `powershell`) that the new output will jump up to y=30 override the existing output in the terminal buffer. The Windows Terminal is already prepared for this quirky behavior, so it doesn't keep the output at the bottom of the window. It shifts it's viewport down to match what conpty things the buffer looks like. What happens when we have passthrough mode and WT is like "I would like quirky resize"? I guess things will just work fine, cause there won't be a buffer behind the passthrough app that the terminal cares about. Sure, in the passthrough case the Terminal could _not_ quirky resize, but the quirky resize won't be wrong.
2020-03-13 01:43:37 +01:00
0, // temporarily set size to 0 so it won't render.
_buffer->GetRenderTarget());
Show a double width cursor for double width characters (#5319) # Summary of the Pull Request This PR will allow the cursor to be double width when on top of a double width character. This required changing `IsCursorDoubleWidth` to check whether the glyph the cursor's on top of is double width. This code is exactly the same as the original PR that addressed this issue in #2932. That one got reverted at some point due to the crashes related to it, but due to a combination of Terminal having come further since that PR and other changes to address use-after-frees, some of the crashes may/may not be relevant now. The ones that seemed to be relevant/repro-able, I attempt to address in this PR. The `IsCursorDoubleWidth` check would fail during the `TextBuffer::Reflow` call inside of `Terminal::UserResize` occasionally, particularly when `newCursor.EndDeferDrawing()` is called. This is because when we tell the newCursor to `EndDefer`, the renderer will attempt to redraw the cursor. As part of this redraw, it'll ask if `IsCursorDoubleWidth`, and if the renderer managed to ask this before `UserResize` swapped out the old buffer with the new one from `Reflow`, the renderer will be asking the old buffer if its out-of-bounds cursor is double width. This was pretty easily repro'd using `cmatrix -u0` and resizing the window like a madman. As a solution, I've moved the Start/End DeferDrawing calls out of `Reflow` and into `UserResize`. This way, I can "clamp" the portion of the code where the newBuffer is getting created and reflowed and swapped into the Terminal buffer, and only allow the renderer to draw once the swap is done. This also means that ConHost's `ResizeWithReflow` needed to change slightly. In addition, I've added a WriteLock to `SetCursorOn`. It was mentioned as a fix for a crash in #2965 (although I can't repro), and I also figured it would be good to try to emulate where ConHost locks with regards to Cursor operations, and this seemed to be one that we were missing. ## PR Checklist * [x] Closes #2713 * [x] CLA signed * [x] Tests added/passed ## Validation Steps Performed Manual validation that the cursor is indeed chonky, added a test case to check that we are correctly saying that the cursor is double width (not too sure if I put it in the right place). Also open to other test case ideas and thoughts on what else I should be careful for since I am quite nervous about what other crashes might occur.
2020-04-15 21:23:06 +02:00
newTextBuffer->GetCursor().StartDeferDrawing();
// Build a PositionInformation to track the position of both the top of
// the mutable viewport and the top of the visible viewport in the new
// buffer.
// * the new value of mutableViewportTop will be used to figure out
// where we should place the mutable viewport in the new buffer. This
// requires a bit of trickiness to remain consistent with conpty's
// buffer (as seen below).
// * the new value of visibleViewportTop will be used to calculate the
// new scrollOffset in the new buffer, so that the visible lines on
// the screen remain roughly the same.
TextBuffer::PositionInformation oldRows{ 0 };
oldRows.mutableViewportTop = oldViewportTop;
oldRows.visibleViewportTop = newVisibleTop;
const std::optional<short> oldViewStart{ oldViewportTop };
Add support for "reflow"ing the Terminal buffer (#4741) This PR adds support for "Resize with Reflow" to the Terminal. In conhost, `ResizeWithReflow` is the function that's responsible for reflowing wrapped lines of text as the buffer gets resized. Now that #4415 has merged, we can also implement this in the Terminal. Now, when the Terminal is resized, it will reflow the lines of it's buffer in the same way that conhost does. This means, the terminal will no longer chop off the ends of lines as the buffer is too small to represent them. As a happy side effect of this PR, it also fixed #3490. This was a bug that plagued me during the investigation into this functionality. The original #3490 PR, #4354, tried to fix this bug with some heavy conpty changes. Turns out, that only made things worse, and far more complicated. When I really got to thinking about it, I realized "conhost can handle this right, why can't the Terminal?". Turns out, by adding resize with reflow, I was also able to fix this at the same time. Conhost does a little bit of math after reflowing to attempt to keep the viewport in the same relative place after a reflow. By re-using that logic in the Terminal, I was able to fix #3490. I also included that big ole test from #3490, because everyone likes adding 60 test cases in a PR. ## References * #4200 - this scenario * #405/#4415 - conpty emits wrapped lines, which was needed for this PR * #4403 - delayed EOL wrapping via conpty, which was also needed for this * #4354 - we don't speak of this PR anymore ## PR Checklist * [x] Closes #1465 * [x] Closes #3490 * [x] Closes #4771 * [x] Tests added/passed ## EDIT: Changes to this PR on 5 March 2020 I learned more since my original version of this PR. I wrote that in January, and despite my notes that say it was totally working, it _really_ wasn't. Part of the hard problem, as mentioned in #3490, is that the Terminal might request a resize to (W, H-1), and while conpty is preparing that frame, or before the terminal has received that frame, the Terminal resizes to (W, H-2). Now, there aren't enough lines in the terminal buffer to catch all the lines that conpty is about to emit. When that happens, lines get duplicated in the buffer. From a UX perspective, this certainly looks a lot worse than a couple lost lines. It looks like utter chaos. So I've introduced a new mode to conpty to try and counteract this behavior. This behavior I'm calling "quirky resize". The **TL;DR** of quirky resize mode is that conpty won't emit the entire buffer on a resize, and will trust that the terminal is prepared to reflow it's buffer on it's own. This will enable the quirky resize behavior for applications that are prepared for it. The "quirky resize" is "don't `InvalidateAll` when the terminal resizes". This is added as a quirk as to not regress other terminal applications that aren't prepared for this behavior (gnome-terminal, conhost in particular). For those kinds of terminals, when the buffer is resized, it's just going to lose lines. That's what currently happens for them. When the quirk is enabled, conpty won't repaint the entire buffer. This gets around the "duplicated lines" issue that requesting multiple resizes in a row can cause. However, for these terminals that are unprepared, the conpty cursor might end up in the wrong position after a quirky resize. The case in point is maximizing the terminal. For maximizing (height->50) from a buffer that's 30 lines tall, with the cursor on y=30, this is what happens: * With the quirk disabled, conpty reprints the entire buffer. This is 60 lines that get printed. This ends up blowing away about 20 lines of scrollback history, as the terminal app would have tried to keep the text pinned to the bottom of the window. The term. app moved the viewport up 20 lines, and then the 50 lines of conpty output (30 lines of text, and 20 blank lines at the bottom) overwrote the lines from the scrollback. This is bad, but not immediately obvious, and is **what currently happens**. * With the quirk enabled, conpty doesn't emit any lines, but the actual content of the window is still only in the top 30 lines. However, the terminal app has still moved 20 lines down from the scrollback back into the viewport. So the terminal's cursor is at y=50 now, but conpty's is at 30. This means that the terminal and conpty are out of sync, and there's not a good way of re-syncing these. It's very possible (trivial in `powershell`) that the new output will jump up to y=30 override the existing output in the terminal buffer. The Windows Terminal is already prepared for this quirky behavior, so it doesn't keep the output at the bottom of the window. It shifts it's viewport down to match what conpty things the buffer looks like. What happens when we have passthrough mode and WT is like "I would like quirky resize"? I guess things will just work fine, cause there won't be a buffer behind the passthrough app that the terminal cares about. Sure, in the passthrough case the Terminal could _not_ quirky resize, but the quirky resize won't be wrong.
2020-03-13 01:43:37 +01:00
RETURN_IF_FAILED(TextBuffer::Reflow(*_buffer.get(),
*newTextBuffer.get(),
_mutableViewport,
{ oldRows }));
newViewportTop = oldRows.mutableViewportTop;
newVisibleTop = oldRows.visibleViewportTop;
// Restore the active text attributes
newTextBuffer->SetCurrentAttributes(oldBufferAttributes);
Add support for "reflow"ing the Terminal buffer (#4741) This PR adds support for "Resize with Reflow" to the Terminal. In conhost, `ResizeWithReflow` is the function that's responsible for reflowing wrapped lines of text as the buffer gets resized. Now that #4415 has merged, we can also implement this in the Terminal. Now, when the Terminal is resized, it will reflow the lines of it's buffer in the same way that conhost does. This means, the terminal will no longer chop off the ends of lines as the buffer is too small to represent them. As a happy side effect of this PR, it also fixed #3490. This was a bug that plagued me during the investigation into this functionality. The original #3490 PR, #4354, tried to fix this bug with some heavy conpty changes. Turns out, that only made things worse, and far more complicated. When I really got to thinking about it, I realized "conhost can handle this right, why can't the Terminal?". Turns out, by adding resize with reflow, I was also able to fix this at the same time. Conhost does a little bit of math after reflowing to attempt to keep the viewport in the same relative place after a reflow. By re-using that logic in the Terminal, I was able to fix #3490. I also included that big ole test from #3490, because everyone likes adding 60 test cases in a PR. ## References * #4200 - this scenario * #405/#4415 - conpty emits wrapped lines, which was needed for this PR * #4403 - delayed EOL wrapping via conpty, which was also needed for this * #4354 - we don't speak of this PR anymore ## PR Checklist * [x] Closes #1465 * [x] Closes #3490 * [x] Closes #4771 * [x] Tests added/passed ## EDIT: Changes to this PR on 5 March 2020 I learned more since my original version of this PR. I wrote that in January, and despite my notes that say it was totally working, it _really_ wasn't. Part of the hard problem, as mentioned in #3490, is that the Terminal might request a resize to (W, H-1), and while conpty is preparing that frame, or before the terminal has received that frame, the Terminal resizes to (W, H-2). Now, there aren't enough lines in the terminal buffer to catch all the lines that conpty is about to emit. When that happens, lines get duplicated in the buffer. From a UX perspective, this certainly looks a lot worse than a couple lost lines. It looks like utter chaos. So I've introduced a new mode to conpty to try and counteract this behavior. This behavior I'm calling "quirky resize". The **TL;DR** of quirky resize mode is that conpty won't emit the entire buffer on a resize, and will trust that the terminal is prepared to reflow it's buffer on it's own. This will enable the quirky resize behavior for applications that are prepared for it. The "quirky resize" is "don't `InvalidateAll` when the terminal resizes". This is added as a quirk as to not regress other terminal applications that aren't prepared for this behavior (gnome-terminal, conhost in particular). For those kinds of terminals, when the buffer is resized, it's just going to lose lines. That's what currently happens for them. When the quirk is enabled, conpty won't repaint the entire buffer. This gets around the "duplicated lines" issue that requesting multiple resizes in a row can cause. However, for these terminals that are unprepared, the conpty cursor might end up in the wrong position after a quirky resize. The case in point is maximizing the terminal. For maximizing (height->50) from a buffer that's 30 lines tall, with the cursor on y=30, this is what happens: * With the quirk disabled, conpty reprints the entire buffer. This is 60 lines that get printed. This ends up blowing away about 20 lines of scrollback history, as the terminal app would have tried to keep the text pinned to the bottom of the window. The term. app moved the viewport up 20 lines, and then the 50 lines of conpty output (30 lines of text, and 20 blank lines at the bottom) overwrote the lines from the scrollback. This is bad, but not immediately obvious, and is **what currently happens**. * With the quirk enabled, conpty doesn't emit any lines, but the actual content of the window is still only in the top 30 lines. However, the terminal app has still moved 20 lines down from the scrollback back into the viewport. So the terminal's cursor is at y=50 now, but conpty's is at 30. This means that the terminal and conpty are out of sync, and there's not a good way of re-syncing these. It's very possible (trivial in `powershell`) that the new output will jump up to y=30 override the existing output in the terminal buffer. The Windows Terminal is already prepared for this quirky behavior, so it doesn't keep the output at the bottom of the window. It shifts it's viewport down to match what conpty things the buffer looks like. What happens when we have passthrough mode and WT is like "I would like quirky resize"? I guess things will just work fine, cause there won't be a buffer behind the passthrough app that the terminal cares about. Sure, in the passthrough case the Terminal could _not_ quirky resize, but the quirky resize won't be wrong.
2020-03-13 01:43:37 +01:00
}
CATCH_RETURN();
// Conpty resizes a little oddly - if the height decreased, and there were
// blank lines at the bottom, those lines will get trimmed. If there's not
// blank lines, then the top will get "shifted down", moving the top line
// into scrollback. See GH#3490 for more details.
//
// If the final position in the buffer is on the bottom row of the new
// viewport, then we're going to need to move the top down. Otherwise, move
// the bottom up.
//
// There are also important things to consider with line wrapping.
// * If a line in scrollback wrapped that didn't previously, we'll need to
// make sure to have the new viewport down another line. This will cause
// our top to move down.
// * If a line _in the viewport_ wrapped that didn't previously, then the
// conpty buffer will also have that wrapped line, and will move the
// cursor & text down a line in response. This causes our bottom to move
// down.
//
// We're going to use a combo of both these things to calculate where the
// new viewport should be. To keep in sync with conpty, we'll need to make
// sure that any lines that entered the scrollback _stay in scrollback_. We
// do that by taking the max of
// * Where the old top line in the viewport exists in the new buffer (as
// calculated by TextBuffer::Reflow)
// * Where the bottom of the text in the new buffer is (and using that to
// calculate another proposed top location).
const COORD newCursorPos = newTextBuffer->GetCursor().GetPosition();
#pragma warning(push)
#pragma warning(disable : 26496) // cpp core checks wants this const, but it's assigned immediately below...
COORD newLastChar = newCursorPos;
try
{
newLastChar = newTextBuffer->GetLastNonSpaceCharacter();
}
CATCH_LOG();
#pragma warning(pop)
const auto maxRow = std::max(newLastChar.Y, newCursorPos.Y);
const short proposedTopFromLastLine = ::base::ClampAdd(::base::ClampSub(maxRow, viewportSize.Y), 1);
Add support for "reflow"ing the Terminal buffer (#4741) This PR adds support for "Resize with Reflow" to the Terminal. In conhost, `ResizeWithReflow` is the function that's responsible for reflowing wrapped lines of text as the buffer gets resized. Now that #4415 has merged, we can also implement this in the Terminal. Now, when the Terminal is resized, it will reflow the lines of it's buffer in the same way that conhost does. This means, the terminal will no longer chop off the ends of lines as the buffer is too small to represent them. As a happy side effect of this PR, it also fixed #3490. This was a bug that plagued me during the investigation into this functionality. The original #3490 PR, #4354, tried to fix this bug with some heavy conpty changes. Turns out, that only made things worse, and far more complicated. When I really got to thinking about it, I realized "conhost can handle this right, why can't the Terminal?". Turns out, by adding resize with reflow, I was also able to fix this at the same time. Conhost does a little bit of math after reflowing to attempt to keep the viewport in the same relative place after a reflow. By re-using that logic in the Terminal, I was able to fix #3490. I also included that big ole test from #3490, because everyone likes adding 60 test cases in a PR. ## References * #4200 - this scenario * #405/#4415 - conpty emits wrapped lines, which was needed for this PR * #4403 - delayed EOL wrapping via conpty, which was also needed for this * #4354 - we don't speak of this PR anymore ## PR Checklist * [x] Closes #1465 * [x] Closes #3490 * [x] Closes #4771 * [x] Tests added/passed ## EDIT: Changes to this PR on 5 March 2020 I learned more since my original version of this PR. I wrote that in January, and despite my notes that say it was totally working, it _really_ wasn't. Part of the hard problem, as mentioned in #3490, is that the Terminal might request a resize to (W, H-1), and while conpty is preparing that frame, or before the terminal has received that frame, the Terminal resizes to (W, H-2). Now, there aren't enough lines in the terminal buffer to catch all the lines that conpty is about to emit. When that happens, lines get duplicated in the buffer. From a UX perspective, this certainly looks a lot worse than a couple lost lines. It looks like utter chaos. So I've introduced a new mode to conpty to try and counteract this behavior. This behavior I'm calling "quirky resize". The **TL;DR** of quirky resize mode is that conpty won't emit the entire buffer on a resize, and will trust that the terminal is prepared to reflow it's buffer on it's own. This will enable the quirky resize behavior for applications that are prepared for it. The "quirky resize" is "don't `InvalidateAll` when the terminal resizes". This is added as a quirk as to not regress other terminal applications that aren't prepared for this behavior (gnome-terminal, conhost in particular). For those kinds of terminals, when the buffer is resized, it's just going to lose lines. That's what currently happens for them. When the quirk is enabled, conpty won't repaint the entire buffer. This gets around the "duplicated lines" issue that requesting multiple resizes in a row can cause. However, for these terminals that are unprepared, the conpty cursor might end up in the wrong position after a quirky resize. The case in point is maximizing the terminal. For maximizing (height->50) from a buffer that's 30 lines tall, with the cursor on y=30, this is what happens: * With the quirk disabled, conpty reprints the entire buffer. This is 60 lines that get printed. This ends up blowing away about 20 lines of scrollback history, as the terminal app would have tried to keep the text pinned to the bottom of the window. The term. app moved the viewport up 20 lines, and then the 50 lines of conpty output (30 lines of text, and 20 blank lines at the bottom) overwrote the lines from the scrollback. This is bad, but not immediately obvious, and is **what currently happens**. * With the quirk enabled, conpty doesn't emit any lines, but the actual content of the window is still only in the top 30 lines. However, the terminal app has still moved 20 lines down from the scrollback back into the viewport. So the terminal's cursor is at y=50 now, but conpty's is at 30. This means that the terminal and conpty are out of sync, and there's not a good way of re-syncing these. It's very possible (trivial in `powershell`) that the new output will jump up to y=30 override the existing output in the terminal buffer. The Windows Terminal is already prepared for this quirky behavior, so it doesn't keep the output at the bottom of the window. It shifts it's viewport down to match what conpty things the buffer looks like. What happens when we have passthrough mode and WT is like "I would like quirky resize"? I guess things will just work fine, cause there won't be a buffer behind the passthrough app that the terminal cares about. Sure, in the passthrough case the Terminal could _not_ quirky resize, but the quirky resize won't be wrong.
2020-03-13 01:43:37 +01:00
const short proposedTopFromScrollback = newViewportTop;
short proposedTop = std::max(proposedTopFromLastLine,
proposedTopFromScrollback);
// If we're using the new location of the old top line to place the
// viewport, we might need to make an adjustment to it.
//
// We're using the last cell of the line to calculate where the top line is
// in the new buffer. If that line wrapped, then all the lines below it
// shifted down in the buffer. If there's space for all those lines in the
// conpty buffer, then the originally unwrapped top line will _still_ be in
// the buffer. In that case, don't stick to the _end_ of the old top line,
// instead stick to the _start_, which is one line up.
//
// We can know if there's space in the conpty buffer by checking if the
// maxRow (the highest row we've written text to) is above the viewport from
// this proposed top position.
if (proposedTop == proposedTopFromScrollback)
{
const auto proposedViewFromTop = Viewport::FromDimensions({ 0, proposedTopFromScrollback }, viewportSize);
if (maxRow < proposedViewFromTop.BottomInclusive())
{
if (dx < 0 && proposedTop > 0)
{
try
{
Greatly reduce allocations in the conhost/OpenConsole startup path (#8489) I was looking at conhost/OpenConsole and noticed it was being pretty inefficient with allocations due to some usages of std::deque and std::vector that didn't need to be done quite that way. So this uses std::vector for the TextBuffer's storage of ROW objects, which allows one allocation to contiguously reserve space for all the ROWs - on Desktop this is 9001 ROW objects which means it saves 9000 allocations that the std::deque would have done. Plus it has the benefit of increasing locality of the ROW objects since deque is going to chase pointers more often with its data structure. Then, within each ROW there are CharRow and ATTR_ROW objects that use std::vector today. This changes them to use Boost's small_vector, which is a variation of vector that allows for the so-called "small string optimization." Since we know the typical size of these vectors, we can pre-reserve the right number of elements directly in the CharRow/ATTR_ROW instances, avoiding any heap allocations at all for constructing these objects. There are a ton of variations on this "small_vector" concept out there in the world - this one in Boost, LLVM has one called SmallVector, Electronic Arts' STL has a small_vector, Facebook's folly library has one...there are a silly number of these out there. But Boost seems like it's by far the easiest to consume in terms of integration into this repo, the CI/CD pipeline, licensing, and stuff like that, so I went with the boost version. In terms of numbers, I measured the startup path of OpenConsole.exe on my dev box for Release x64 configuration. My box is an i7-6700k @ 4 Ghz, with 32 GB RAM, not that I think machine config matters much here: | | Allocation count | Allocated bytes | CPU usage (ms) | | ------ | ------------------- | ------------------ | -------------- | | Before | 29,461 | 4,984,640 | 103 | | After | 2,459 (-91%) | 4,853,931 (-2.6%) | 96 (-7%) | Along the way, I also fixed a dynamic initializer I happened to spot in the registry code, and updated some docs. ## Validation Steps Performed - Ran "runut", "runft" and "runuia" locally and confirmed results are the same as the main branch - Profiled the before/after numbers in the Visual Studio profiler, for the numbers shown in the table Co-authored-by: Austin Lamb <austinl@microsoft.com>
2020-12-16 19:40:30 +01:00
auto& row = newTextBuffer->GetRowByOffset(::base::ClampSub(proposedTop, 1));
if (row.WasWrapForced())
Add support for "reflow"ing the Terminal buffer (#4741) This PR adds support for "Resize with Reflow" to the Terminal. In conhost, `ResizeWithReflow` is the function that's responsible for reflowing wrapped lines of text as the buffer gets resized. Now that #4415 has merged, we can also implement this in the Terminal. Now, when the Terminal is resized, it will reflow the lines of it's buffer in the same way that conhost does. This means, the terminal will no longer chop off the ends of lines as the buffer is too small to represent them. As a happy side effect of this PR, it also fixed #3490. This was a bug that plagued me during the investigation into this functionality. The original #3490 PR, #4354, tried to fix this bug with some heavy conpty changes. Turns out, that only made things worse, and far more complicated. When I really got to thinking about it, I realized "conhost can handle this right, why can't the Terminal?". Turns out, by adding resize with reflow, I was also able to fix this at the same time. Conhost does a little bit of math after reflowing to attempt to keep the viewport in the same relative place after a reflow. By re-using that logic in the Terminal, I was able to fix #3490. I also included that big ole test from #3490, because everyone likes adding 60 test cases in a PR. ## References * #4200 - this scenario * #405/#4415 - conpty emits wrapped lines, which was needed for this PR * #4403 - delayed EOL wrapping via conpty, which was also needed for this * #4354 - we don't speak of this PR anymore ## PR Checklist * [x] Closes #1465 * [x] Closes #3490 * [x] Closes #4771 * [x] Tests added/passed ## EDIT: Changes to this PR on 5 March 2020 I learned more since my original version of this PR. I wrote that in January, and despite my notes that say it was totally working, it _really_ wasn't. Part of the hard problem, as mentioned in #3490, is that the Terminal might request a resize to (W, H-1), and while conpty is preparing that frame, or before the terminal has received that frame, the Terminal resizes to (W, H-2). Now, there aren't enough lines in the terminal buffer to catch all the lines that conpty is about to emit. When that happens, lines get duplicated in the buffer. From a UX perspective, this certainly looks a lot worse than a couple lost lines. It looks like utter chaos. So I've introduced a new mode to conpty to try and counteract this behavior. This behavior I'm calling "quirky resize". The **TL;DR** of quirky resize mode is that conpty won't emit the entire buffer on a resize, and will trust that the terminal is prepared to reflow it's buffer on it's own. This will enable the quirky resize behavior for applications that are prepared for it. The "quirky resize" is "don't `InvalidateAll` when the terminal resizes". This is added as a quirk as to not regress other terminal applications that aren't prepared for this behavior (gnome-terminal, conhost in particular). For those kinds of terminals, when the buffer is resized, it's just going to lose lines. That's what currently happens for them. When the quirk is enabled, conpty won't repaint the entire buffer. This gets around the "duplicated lines" issue that requesting multiple resizes in a row can cause. However, for these terminals that are unprepared, the conpty cursor might end up in the wrong position after a quirky resize. The case in point is maximizing the terminal. For maximizing (height->50) from a buffer that's 30 lines tall, with the cursor on y=30, this is what happens: * With the quirk disabled, conpty reprints the entire buffer. This is 60 lines that get printed. This ends up blowing away about 20 lines of scrollback history, as the terminal app would have tried to keep the text pinned to the bottom of the window. The term. app moved the viewport up 20 lines, and then the 50 lines of conpty output (30 lines of text, and 20 blank lines at the bottom) overwrote the lines from the scrollback. This is bad, but not immediately obvious, and is **what currently happens**. * With the quirk enabled, conpty doesn't emit any lines, but the actual content of the window is still only in the top 30 lines. However, the terminal app has still moved 20 lines down from the scrollback back into the viewport. So the terminal's cursor is at y=50 now, but conpty's is at 30. This means that the terminal and conpty are out of sync, and there's not a good way of re-syncing these. It's very possible (trivial in `powershell`) that the new output will jump up to y=30 override the existing output in the terminal buffer. The Windows Terminal is already prepared for this quirky behavior, so it doesn't keep the output at the bottom of the window. It shifts it's viewport down to match what conpty things the buffer looks like. What happens when we have passthrough mode and WT is like "I would like quirky resize"? I guess things will just work fine, cause there won't be a buffer behind the passthrough app that the terminal cares about. Sure, in the passthrough case the Terminal could _not_ quirky resize, but the quirky resize won't be wrong.
2020-03-13 01:43:37 +01:00
{
proposedTop--;
}
}
CATCH_LOG();
}
}
}
// If the new bottom would be higher than the last row of text, then we
// definitely want to use the last row of text to determine where the
// viewport should be.
const auto proposedViewFromTop = Viewport::FromDimensions({ 0, proposedTopFromScrollback }, viewportSize);
if (maxRow > proposedViewFromTop.BottomInclusive())
{
proposedTop = proposedTopFromLastLine;
}
// Make sure the proposed viewport is within the bounds of the buffer.
// First make sure the top is >=0
proposedTop = std::max(static_cast<short>(0), proposedTop);
// If the new bottom would be below the bottom of the buffer, then slide the
// top up so that we'll still fit within the buffer.
Add support for "reflow"ing the Terminal buffer (#4741) This PR adds support for "Resize with Reflow" to the Terminal. In conhost, `ResizeWithReflow` is the function that's responsible for reflowing wrapped lines of text as the buffer gets resized. Now that #4415 has merged, we can also implement this in the Terminal. Now, when the Terminal is resized, it will reflow the lines of it's buffer in the same way that conhost does. This means, the terminal will no longer chop off the ends of lines as the buffer is too small to represent them. As a happy side effect of this PR, it also fixed #3490. This was a bug that plagued me during the investigation into this functionality. The original #3490 PR, #4354, tried to fix this bug with some heavy conpty changes. Turns out, that only made things worse, and far more complicated. When I really got to thinking about it, I realized "conhost can handle this right, why can't the Terminal?". Turns out, by adding resize with reflow, I was also able to fix this at the same time. Conhost does a little bit of math after reflowing to attempt to keep the viewport in the same relative place after a reflow. By re-using that logic in the Terminal, I was able to fix #3490. I also included that big ole test from #3490, because everyone likes adding 60 test cases in a PR. ## References * #4200 - this scenario * #405/#4415 - conpty emits wrapped lines, which was needed for this PR * #4403 - delayed EOL wrapping via conpty, which was also needed for this * #4354 - we don't speak of this PR anymore ## PR Checklist * [x] Closes #1465 * [x] Closes #3490 * [x] Closes #4771 * [x] Tests added/passed ## EDIT: Changes to this PR on 5 March 2020 I learned more since my original version of this PR. I wrote that in January, and despite my notes that say it was totally working, it _really_ wasn't. Part of the hard problem, as mentioned in #3490, is that the Terminal might request a resize to (W, H-1), and while conpty is preparing that frame, or before the terminal has received that frame, the Terminal resizes to (W, H-2). Now, there aren't enough lines in the terminal buffer to catch all the lines that conpty is about to emit. When that happens, lines get duplicated in the buffer. From a UX perspective, this certainly looks a lot worse than a couple lost lines. It looks like utter chaos. So I've introduced a new mode to conpty to try and counteract this behavior. This behavior I'm calling "quirky resize". The **TL;DR** of quirky resize mode is that conpty won't emit the entire buffer on a resize, and will trust that the terminal is prepared to reflow it's buffer on it's own. This will enable the quirky resize behavior for applications that are prepared for it. The "quirky resize" is "don't `InvalidateAll` when the terminal resizes". This is added as a quirk as to not regress other terminal applications that aren't prepared for this behavior (gnome-terminal, conhost in particular). For those kinds of terminals, when the buffer is resized, it's just going to lose lines. That's what currently happens for them. When the quirk is enabled, conpty won't repaint the entire buffer. This gets around the "duplicated lines" issue that requesting multiple resizes in a row can cause. However, for these terminals that are unprepared, the conpty cursor might end up in the wrong position after a quirky resize. The case in point is maximizing the terminal. For maximizing (height->50) from a buffer that's 30 lines tall, with the cursor on y=30, this is what happens: * With the quirk disabled, conpty reprints the entire buffer. This is 60 lines that get printed. This ends up blowing away about 20 lines of scrollback history, as the terminal app would have tried to keep the text pinned to the bottom of the window. The term. app moved the viewport up 20 lines, and then the 50 lines of conpty output (30 lines of text, and 20 blank lines at the bottom) overwrote the lines from the scrollback. This is bad, but not immediately obvious, and is **what currently happens**. * With the quirk enabled, conpty doesn't emit any lines, but the actual content of the window is still only in the top 30 lines. However, the terminal app has still moved 20 lines down from the scrollback back into the viewport. So the terminal's cursor is at y=50 now, but conpty's is at 30. This means that the terminal and conpty are out of sync, and there's not a good way of re-syncing these. It's very possible (trivial in `powershell`) that the new output will jump up to y=30 override the existing output in the terminal buffer. The Windows Terminal is already prepared for this quirky behavior, so it doesn't keep the output at the bottom of the window. It shifts it's viewport down to match what conpty things the buffer looks like. What happens when we have passthrough mode and WT is like "I would like quirky resize"? I guess things will just work fine, cause there won't be a buffer behind the passthrough app that the terminal cares about. Sure, in the passthrough case the Terminal could _not_ quirky resize, but the quirky resize won't be wrong.
2020-03-13 01:43:37 +01:00
const auto newView = Viewport::FromDimensions({ 0, proposedTop }, viewportSize);
const auto proposedBottom = newView.BottomExclusive();
if (proposedBottom > bufferSize.Y)
{
proposedTop = ::base::ClampSub(proposedTop, ::base::ClampSub(proposedBottom, bufferSize.Y));
}
_mutableViewport = Viewport::FromDimensions({ 0, proposedTop }, viewportSize);
Add support for "reflow"ing the Terminal buffer (#4741) This PR adds support for "Resize with Reflow" to the Terminal. In conhost, `ResizeWithReflow` is the function that's responsible for reflowing wrapped lines of text as the buffer gets resized. Now that #4415 has merged, we can also implement this in the Terminal. Now, when the Terminal is resized, it will reflow the lines of it's buffer in the same way that conhost does. This means, the terminal will no longer chop off the ends of lines as the buffer is too small to represent them. As a happy side effect of this PR, it also fixed #3490. This was a bug that plagued me during the investigation into this functionality. The original #3490 PR, #4354, tried to fix this bug with some heavy conpty changes. Turns out, that only made things worse, and far more complicated. When I really got to thinking about it, I realized "conhost can handle this right, why can't the Terminal?". Turns out, by adding resize with reflow, I was also able to fix this at the same time. Conhost does a little bit of math after reflowing to attempt to keep the viewport in the same relative place after a reflow. By re-using that logic in the Terminal, I was able to fix #3490. I also included that big ole test from #3490, because everyone likes adding 60 test cases in a PR. ## References * #4200 - this scenario * #405/#4415 - conpty emits wrapped lines, which was needed for this PR * #4403 - delayed EOL wrapping via conpty, which was also needed for this * #4354 - we don't speak of this PR anymore ## PR Checklist * [x] Closes #1465 * [x] Closes #3490 * [x] Closes #4771 * [x] Tests added/passed ## EDIT: Changes to this PR on 5 March 2020 I learned more since my original version of this PR. I wrote that in January, and despite my notes that say it was totally working, it _really_ wasn't. Part of the hard problem, as mentioned in #3490, is that the Terminal might request a resize to (W, H-1), and while conpty is preparing that frame, or before the terminal has received that frame, the Terminal resizes to (W, H-2). Now, there aren't enough lines in the terminal buffer to catch all the lines that conpty is about to emit. When that happens, lines get duplicated in the buffer. From a UX perspective, this certainly looks a lot worse than a couple lost lines. It looks like utter chaos. So I've introduced a new mode to conpty to try and counteract this behavior. This behavior I'm calling "quirky resize". The **TL;DR** of quirky resize mode is that conpty won't emit the entire buffer on a resize, and will trust that the terminal is prepared to reflow it's buffer on it's own. This will enable the quirky resize behavior for applications that are prepared for it. The "quirky resize" is "don't `InvalidateAll` when the terminal resizes". This is added as a quirk as to not regress other terminal applications that aren't prepared for this behavior (gnome-terminal, conhost in particular). For those kinds of terminals, when the buffer is resized, it's just going to lose lines. That's what currently happens for them. When the quirk is enabled, conpty won't repaint the entire buffer. This gets around the "duplicated lines" issue that requesting multiple resizes in a row can cause. However, for these terminals that are unprepared, the conpty cursor might end up in the wrong position after a quirky resize. The case in point is maximizing the terminal. For maximizing (height->50) from a buffer that's 30 lines tall, with the cursor on y=30, this is what happens: * With the quirk disabled, conpty reprints the entire buffer. This is 60 lines that get printed. This ends up blowing away about 20 lines of scrollback history, as the terminal app would have tried to keep the text pinned to the bottom of the window. The term. app moved the viewport up 20 lines, and then the 50 lines of conpty output (30 lines of text, and 20 blank lines at the bottom) overwrote the lines from the scrollback. This is bad, but not immediately obvious, and is **what currently happens**. * With the quirk enabled, conpty doesn't emit any lines, but the actual content of the window is still only in the top 30 lines. However, the terminal app has still moved 20 lines down from the scrollback back into the viewport. So the terminal's cursor is at y=50 now, but conpty's is at 30. This means that the terminal and conpty are out of sync, and there's not a good way of re-syncing these. It's very possible (trivial in `powershell`) that the new output will jump up to y=30 override the existing output in the terminal buffer. The Windows Terminal is already prepared for this quirky behavior, so it doesn't keep the output at the bottom of the window. It shifts it's viewport down to match what conpty things the buffer looks like. What happens when we have passthrough mode and WT is like "I would like quirky resize"? I guess things will just work fine, cause there won't be a buffer behind the passthrough app that the terminal cares about. Sure, in the passthrough case the Terminal could _not_ quirky resize, but the quirky resize won't be wrong.
2020-03-13 01:43:37 +01:00
_buffer.swap(newTextBuffer);
// GH#3494: Maintain scrollbar position during resize
// Make sure that we don't scroll past the mutableViewport at the bottom of the buffer
newVisibleTop = std::min(newVisibleTop, _mutableViewport.Top());
// Make sure we don't scroll past the top of the scrollback
newVisibleTop = std::max<short>(newVisibleTop, 0);
// If the old scrolloffset was 0, then we weren't scrolled back at all
// before, and shouldn't be now either.
_scrollOffset = originalOffsetWasZero ? 0 : static_cast<int>(::base::ClampSub(_mutableViewport.Top(), newVisibleTop));
// GH#5029 - make sure to InvalidateAll here, so that we'll paint the entire visible viewport.
try
{
_buffer->GetRenderTarget().TriggerRedrawAll();
}
CATCH_LOG();
_NotifyScrollEvent();
return S_OK;
}
void Terminal::Write(std::wstring_view stringView)
{
auto lock = LockForWriting();
_stateMachine->ProcessString(stringView);
}
void Terminal::WritePastedText(std::wstring_view stringView)
{
auto option = ::Microsoft::Console::Utils::FilterOption::CarriageReturnNewline |
::Microsoft::Console::Utils::FilterOption::ControlCodes;
std::wstring filtered = ::Microsoft::Console::Utils::FilterStringForPaste(stringView, option);
if (IsXtermBracketedPasteModeEnabled())
{
filtered.insert(0, L"\x1b[200~");
filtered.append(L"\x1b[201~");
}
if (_pfnWriteInput)
{
_pfnWriteInput(filtered);
}
}
2019-11-13 03:12:43 +01:00
// Method Description:
// - Attempts to snap to the bottom of the buffer, if SnapOnInput is true. Does
// nothing if SnapOnInput is set to false, or we're already at the bottom of
// the buffer.
// Arguments:
// - <none>
// Return Value:
// - <none>
void Terminal::TrySnapOnInput()
{
if (_snapOnInput && _scrollOffset != 0)
{
auto lock = LockForWriting();
_scrollOffset = 0;
_NotifyScrollEvent();
}
}
// Routine Description:
// - Relays if we are tracking mouse input
// Parameters:
// - <none>
// Return value:
// - true, if we are tracking mouse input. False, otherwise
bool Terminal::IsTrackingMouseInput() const noexcept
{
return _terminalInput->IsTrackingMouseInput();
}
OSC 8 support for conhost and terminal (#7251) <!-- 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 Conhost can now support OSC8 sequences (as specified [here](https://gist.github.com/egmontkob/eb114294efbcd5adb1944c9f3cb5feda)). Terminal also supports those sequences and additionally hyperlinks can be opened by Ctrl+LeftClicking on them. <!-- Other than the issue solved, is this relevant to any other issues/existing PRs? --> ## References #204 <!-- Please review the items on the PR checklist before submitting--> ## PR Checklist * [X] Closes #204 * [ ] CLA signed. If not, go over [here](https://cla.opensource.microsoft.com/microsoft/Terminal) and sign the CLA * [ ] Tests added/passed * [ ] Documentation updated. If checked, please file a pull request on [our docs repo](https://github.com/MicrosoftDocs/terminal) and link it here: #xxx * [ ] Schema 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 Added support to: - parse OSC8 sequences and extract URIs from them (conhost and terminal) - add hyperlink uri data to textbuffer/screeninformation, associated with a hyperlink id (conhost and terminal) - attach hyperlink ids to text to allow for uri extraction from the textbuffer/screeninformation (conhost and terminal) - process ctrl+leftclick to open a hyperlink in the clicked region if present <!-- Describe how you validated the behavior. Add automated tests wherever possible, but list manual validation steps taken as well --> ## Validation Steps Performed Open up a PowerShell tab and type ```PowerShell ${ESC}=[char]27 Write-Host "${ESC}]8;;https://github.com/microsoft/terminal${ESC}\This is a link!${ESC}]8;;${ESC}\" ``` Ctrl+LeftClick on the link correctly brings you to the terminal page on github ![hyperlink](https://user-images.githubusercontent.com/26824113/89953536-45a6f580-dbfd-11ea-8e0d-8a3cd25c634a.gif)
2020-09-03 19:52:39 +02:00
// Method Description:
// - Given a coord, get the URI at that location
OSC 8 support for conhost and terminal (#7251) <!-- 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 Conhost can now support OSC8 sequences (as specified [here](https://gist.github.com/egmontkob/eb114294efbcd5adb1944c9f3cb5feda)). Terminal also supports those sequences and additionally hyperlinks can be opened by Ctrl+LeftClicking on them. <!-- Other than the issue solved, is this relevant to any other issues/existing PRs? --> ## References #204 <!-- Please review the items on the PR checklist before submitting--> ## PR Checklist * [X] Closes #204 * [ ] CLA signed. If not, go over [here](https://cla.opensource.microsoft.com/microsoft/Terminal) and sign the CLA * [ ] Tests added/passed * [ ] Documentation updated. If checked, please file a pull request on [our docs repo](https://github.com/MicrosoftDocs/terminal) and link it here: #xxx * [ ] Schema 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 Added support to: - parse OSC8 sequences and extract URIs from them (conhost and terminal) - add hyperlink uri data to textbuffer/screeninformation, associated with a hyperlink id (conhost and terminal) - attach hyperlink ids to text to allow for uri extraction from the textbuffer/screeninformation (conhost and terminal) - process ctrl+leftclick to open a hyperlink in the clicked region if present <!-- Describe how you validated the behavior. Add automated tests wherever possible, but list manual validation steps taken as well --> ## Validation Steps Performed Open up a PowerShell tab and type ```PowerShell ${ESC}=[char]27 Write-Host "${ESC}]8;;https://github.com/microsoft/terminal${ESC}\This is a link!${ESC}]8;;${ESC}\" ``` Ctrl+LeftClick on the link correctly brings you to the terminal page on github ![hyperlink](https://user-images.githubusercontent.com/26824113/89953536-45a6f580-dbfd-11ea-8e0d-8a3cd25c634a.gif)
2020-09-03 19:52:39 +02:00
// Arguments:
// - The position
OSC 8 support for conhost and terminal (#7251) <!-- 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 Conhost can now support OSC8 sequences (as specified [here](https://gist.github.com/egmontkob/eb114294efbcd5adb1944c9f3cb5feda)). Terminal also supports those sequences and additionally hyperlinks can be opened by Ctrl+LeftClicking on them. <!-- Other than the issue solved, is this relevant to any other issues/existing PRs? --> ## References #204 <!-- Please review the items on the PR checklist before submitting--> ## PR Checklist * [X] Closes #204 * [ ] CLA signed. If not, go over [here](https://cla.opensource.microsoft.com/microsoft/Terminal) and sign the CLA * [ ] Tests added/passed * [ ] Documentation updated. If checked, please file a pull request on [our docs repo](https://github.com/MicrosoftDocs/terminal) and link it here: #xxx * [ ] Schema 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 Added support to: - parse OSC8 sequences and extract URIs from them (conhost and terminal) - add hyperlink uri data to textbuffer/screeninformation, associated with a hyperlink id (conhost and terminal) - attach hyperlink ids to text to allow for uri extraction from the textbuffer/screeninformation (conhost and terminal) - process ctrl+leftclick to open a hyperlink in the clicked region if present <!-- Describe how you validated the behavior. Add automated tests wherever possible, but list manual validation steps taken as well --> ## Validation Steps Performed Open up a PowerShell tab and type ```PowerShell ${ESC}=[char]27 Write-Host "${ESC}]8;;https://github.com/microsoft/terminal${ESC}\This is a link!${ESC}]8;;${ESC}\" ``` Ctrl+LeftClick on the link correctly brings you to the terminal page on github ![hyperlink](https://user-images.githubusercontent.com/26824113/89953536-45a6f580-dbfd-11ea-8e0d-8a3cd25c634a.gif)
2020-09-03 19:52:39 +02:00
std::wstring Terminal::GetHyperlinkAtPosition(const COORD position)
{
auto attr = _buffer->GetCellDataAt(_ConvertToBufferCell(position))->TextAttr();
if (attr.IsHyperlink())
{
auto uri = _buffer->GetHyperlinkUriFromId(attr.GetHyperlinkId());
return uri;
}
// also look through our known pattern locations in our pattern interval tree
const auto result = GetHyperlinkIntervalFromPosition(position);
if (result.has_value() && result->value == _hyperlinkPatternId)
{
const auto start = result->start;
const auto end = result->stop;
std::wstring uri;
const auto startIter = _buffer->GetCellDataAt(_ConvertToBufferCell(start));
const auto endIter = _buffer->GetCellDataAt(_ConvertToBufferCell(end));
for (auto iter = startIter; iter != endIter; ++iter)
{
uri += iter->Chars();
}
return uri;
}
OSC 8 support for conhost and terminal (#7251) <!-- 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 Conhost can now support OSC8 sequences (as specified [here](https://gist.github.com/egmontkob/eb114294efbcd5adb1944c9f3cb5feda)). Terminal also supports those sequences and additionally hyperlinks can be opened by Ctrl+LeftClicking on them. <!-- Other than the issue solved, is this relevant to any other issues/existing PRs? --> ## References #204 <!-- Please review the items on the PR checklist before submitting--> ## PR Checklist * [X] Closes #204 * [ ] CLA signed. If not, go over [here](https://cla.opensource.microsoft.com/microsoft/Terminal) and sign the CLA * [ ] Tests added/passed * [ ] Documentation updated. If checked, please file a pull request on [our docs repo](https://github.com/MicrosoftDocs/terminal) and link it here: #xxx * [ ] Schema 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 Added support to: - parse OSC8 sequences and extract URIs from them (conhost and terminal) - add hyperlink uri data to textbuffer/screeninformation, associated with a hyperlink id (conhost and terminal) - attach hyperlink ids to text to allow for uri extraction from the textbuffer/screeninformation (conhost and terminal) - process ctrl+leftclick to open a hyperlink in the clicked region if present <!-- Describe how you validated the behavior. Add automated tests wherever possible, but list manual validation steps taken as well --> ## Validation Steps Performed Open up a PowerShell tab and type ```PowerShell ${ESC}=[char]27 Write-Host "${ESC}]8;;https://github.com/microsoft/terminal${ESC}\This is a link!${ESC}]8;;${ESC}\" ``` Ctrl+LeftClick on the link correctly brings you to the terminal page on github ![hyperlink](https://user-images.githubusercontent.com/26824113/89953536-45a6f580-dbfd-11ea-8e0d-8a3cd25c634a.gif)
2020-09-03 19:52:39 +02:00
return {};
}
// Method Description:
// - Gets the hyperlink ID of the text at the given terminal position
// Arguments:
// - The position of the text
// Return value:
// - The hyperlink ID
uint16_t Terminal::GetHyperlinkIdAtPosition(const COORD position)
{
return _buffer->GetCellDataAt(_ConvertToBufferCell(position))->TextAttr().GetHyperlinkId();
}
// Method description:
// - Given a position in a URI pattern, gets the start and end coordinates of the URI
// Arguments:
// - The position
// Return value:
// - The interval representing the start and end coordinates
std::optional<PointTree::interval> Terminal::GetHyperlinkIntervalFromPosition(const COORD position)
{
const auto results = _patternIntervalTree.findOverlapping(COORD{ position.X + 1, position.Y }, position);
if (results.size() > 0)
{
for (const auto& result : results)
{
if (result.value == _hyperlinkPatternId)
{
return result;
}
}
}
return std::nullopt;
}
// Method Description:
Delegate all character input to the character event handler (#4192) My basic idea was that `WM_CHAR` is just the better `WM_KEYDOWN`. The latter fails to properly support common dead key sequences like in #3516. As such I added some logic to `Terminal::SendKeyEvent` to make it return false if the pressed key represents a printable character. This causes us to receive a character event with a (hopefully) correctly composed code unit, which then gets sent to `Terminal::SendCharEvent`. `Terminal::SendCharEvent` in turn had to be modified to support potentially pressed modifier keys, since `Terminal::SendKeyEvent` isn't doing that for us anymore. Lastly `TerminalInput` had to be modified heavily to support character events with modifier key states. In order to do so I merged its `HandleKey` and `HandleChar` methods into a single one, that now handles both cases. Since key events will now contain character data and character events key codes the decision logic in `TerminalInput::HandleKey` had to be rewritten. ## PR Checklist * [x] CLA signed * [x] Tests added/passed * [x] I've discussed this with core contributors already. ## Validation Steps Performed * See #3516. * I don't have any keyboard that generates surrogate characters. Due to this I modified `TermControl::_SendPastedTextToConnection` to send the data to `_terminal->SendCharEvent()` instead. I then pasted the test string ""𐐌𐐜𐐬" and ensured that the new `TerminalInput::_SendChar` method still correctly assembles surrogate pairs. Closes #3516 Closes #3554 (obsoleted by this PR) Potentially impacts #391, which sounds like a duplicate of #3516
2020-04-07 21:09:28 +02:00
// - Send this particular (non-character) key event to the terminal.
// - The terminal will translate the key and the modifiers pressed into the
// appropriate VT sequence for that key chord. If we do translate the key,
// we'll return true. In that case, the event should NOT be processed any further.
// - Character events (e.g. WM_CHAR) are generally the best way to properly receive
// keyboard input on Windows though, as the OS is suited best at handling the
// translation of the current keyboard layout, dead keys, etc.
// As a result of this false is returned for all key events that contain characters.
// SendCharEvent may then be called with the data obtained from a character event.
// - As a special case we'll always handle VK_TAB key events.
// This must be done due to TermControl::_KeyDownHandler (one of the callers)
// always marking tab key events as handled, causing no character event to be raised.
// Arguments:
Delegate all character input to the character event handler (#4192) My basic idea was that `WM_CHAR` is just the better `WM_KEYDOWN`. The latter fails to properly support common dead key sequences like in #3516. As such I added some logic to `Terminal::SendKeyEvent` to make it return false if the pressed key represents a printable character. This causes us to receive a character event with a (hopefully) correctly composed code unit, which then gets sent to `Terminal::SendCharEvent`. `Terminal::SendCharEvent` in turn had to be modified to support potentially pressed modifier keys, since `Terminal::SendKeyEvent` isn't doing that for us anymore. Lastly `TerminalInput` had to be modified heavily to support character events with modifier key states. In order to do so I merged its `HandleKey` and `HandleChar` methods into a single one, that now handles both cases. Since key events will now contain character data and character events key codes the decision logic in `TerminalInput::HandleKey` had to be rewritten. ## PR Checklist * [x] CLA signed * [x] Tests added/passed * [x] I've discussed this with core contributors already. ## Validation Steps Performed * See #3516. * I don't have any keyboard that generates surrogate characters. Due to this I modified `TermControl::_SendPastedTextToConnection` to send the data to `_terminal->SendCharEvent()` instead. I then pasted the test string ""𐐌𐐜𐐬" and ensured that the new `TerminalInput::_SendChar` method still correctly assembles surrogate pairs. Closes #3516 Closes #3554 (obsoleted by this PR) Potentially impacts #391, which sounds like a duplicate of #3516
2020-04-07 21:09:28 +02:00
// - vkey: The vkey of the last pressed key.
// - scanCode: The scan code of the last pressed key.
// - states: The Microsoft::Terminal::Core::ControlKeyStates representing the modifier key states.
// - keyDown: If true, the key was pressed, otherwise the key was released.
// Return Value:
// - true if we translated the key event, and it should not be processed any further.
// - false if we did not translate the key, and it should be processed into a character.
bool Terminal::SendKeyEvent(const WORD vkey,
const WORD scanCode,
const ControlKeyStates states,
const bool keyDown)
{
// GH#6423 - don't snap on this key if the key that was pressed was a
// modifier key. We'll wait for a real keystroke to snap to the bottom.
// GH#6481 - Additionally, make sure the key was actually pressed. This
// check will make sure we behave the same as before GH#6309
if (!KeyEvent::IsModifierKey(vkey) && keyDown)
{
TrySnapOnInput();
}
Delegate all character input to the character event handler (#4192) My basic idea was that `WM_CHAR` is just the better `WM_KEYDOWN`. The latter fails to properly support common dead key sequences like in #3516. As such I added some logic to `Terminal::SendKeyEvent` to make it return false if the pressed key represents a printable character. This causes us to receive a character event with a (hopefully) correctly composed code unit, which then gets sent to `Terminal::SendCharEvent`. `Terminal::SendCharEvent` in turn had to be modified to support potentially pressed modifier keys, since `Terminal::SendKeyEvent` isn't doing that for us anymore. Lastly `TerminalInput` had to be modified heavily to support character events with modifier key states. In order to do so I merged its `HandleKey` and `HandleChar` methods into a single one, that now handles both cases. Since key events will now contain character data and character events key codes the decision logic in `TerminalInput::HandleKey` had to be rewritten. ## PR Checklist * [x] CLA signed * [x] Tests added/passed * [x] I've discussed this with core contributors already. ## Validation Steps Performed * See #3516. * I don't have any keyboard that generates surrogate characters. Due to this I modified `TermControl::_SendPastedTextToConnection` to send the data to `_terminal->SendCharEvent()` instead. I then pasted the test string ""𐐌𐐜𐐬" and ensured that the new `TerminalInput::_SendChar` method still correctly assembles surrogate pairs. Closes #3516 Closes #3554 (obsoleted by this PR) Potentially impacts #391, which sounds like a duplicate of #3516
2020-04-07 21:09:28 +02:00
_StoreKeyEvent(vkey, scanCode);
// Certain applications like AutoHotKey and its keyboard remapping feature,
// send us key events using SendInput() whose values are outside of the valid range.
// GH#7064
if (vkey == 0 || vkey >= 0xff)
{
return false;
}
// While not explicitly permitted, a wide range of software, including Windows' own touch keyboard,
// sets the wScan member of the KEYBDINPUT structure to 0, resulting in scanCode being 0 as well.
// --> Alternatively get the scanCode from the vkey if possible.
// GH#7495
const auto sc = scanCode ? scanCode : _ScanCodeFromVirtualKey(vkey);
if (sc == 0)
{
return false;
}
Delegate all character input to the character event handler (#4192) My basic idea was that `WM_CHAR` is just the better `WM_KEYDOWN`. The latter fails to properly support common dead key sequences like in #3516. As such I added some logic to `Terminal::SendKeyEvent` to make it return false if the pressed key represents a printable character. This causes us to receive a character event with a (hopefully) correctly composed code unit, which then gets sent to `Terminal::SendCharEvent`. `Terminal::SendCharEvent` in turn had to be modified to support potentially pressed modifier keys, since `Terminal::SendKeyEvent` isn't doing that for us anymore. Lastly `TerminalInput` had to be modified heavily to support character events with modifier key states. In order to do so I merged its `HandleKey` and `HandleChar` methods into a single one, that now handles both cases. Since key events will now contain character data and character events key codes the decision logic in `TerminalInput::HandleKey` had to be rewritten. ## PR Checklist * [x] CLA signed * [x] Tests added/passed * [x] I've discussed this with core contributors already. ## Validation Steps Performed * See #3516. * I don't have any keyboard that generates surrogate characters. Due to this I modified `TermControl::_SendPastedTextToConnection` to send the data to `_terminal->SendCharEvent()` instead. I then pasted the test string ""𐐌𐐜𐐬" and ensured that the new `TerminalInput::_SendChar` method still correctly assembles surrogate pairs. Closes #3516 Closes #3554 (obsoleted by this PR) Potentially impacts #391, which sounds like a duplicate of #3516
2020-04-07 21:09:28 +02:00
const auto isAltOnlyPressed = states.IsAltPressed() && !states.IsCtrlPressed();
// By default Windows treats Ctrl+Alt as an alias for AltGr.
// When the altGrAliasing setting is set to false, this behaviour should be disabled.
//
// Whenever possible _CharacterFromKeyEvent() will return a valid character.
// For instance both Ctrl+Alt+Q as well as AltGr+Q return @ on a German keyboard.
//
// We can achieve the altGrAliasing functionality by skipping the call to _CharacterFromKeyEvent,
// as TerminalInput::HandleKey will then fall back to using the vkey which
// is the underlying ASCII character (e.g. A-Z) on the keyboard in our case.
// See GH#5525/GH#6211 for more details
const auto isSuppressedAltGrAlias = !_altGrAliasing && states.IsAltPressed() && states.IsCtrlPressed() && !states.IsAltGrPressed();
const auto ch = isSuppressedAltGrAlias ? UNICODE_NULL : _CharacterFromKeyEvent(vkey, sc, states);
Delegate all character input to the character event handler (#4192) My basic idea was that `WM_CHAR` is just the better `WM_KEYDOWN`. The latter fails to properly support common dead key sequences like in #3516. As such I added some logic to `Terminal::SendKeyEvent` to make it return false if the pressed key represents a printable character. This causes us to receive a character event with a (hopefully) correctly composed code unit, which then gets sent to `Terminal::SendCharEvent`. `Terminal::SendCharEvent` in turn had to be modified to support potentially pressed modifier keys, since `Terminal::SendKeyEvent` isn't doing that for us anymore. Lastly `TerminalInput` had to be modified heavily to support character events with modifier key states. In order to do so I merged its `HandleKey` and `HandleChar` methods into a single one, that now handles both cases. Since key events will now contain character data and character events key codes the decision logic in `TerminalInput::HandleKey` had to be rewritten. ## PR Checklist * [x] CLA signed * [x] Tests added/passed * [x] I've discussed this with core contributors already. ## Validation Steps Performed * See #3516. * I don't have any keyboard that generates surrogate characters. Due to this I modified `TermControl::_SendPastedTextToConnection` to send the data to `_terminal->SendCharEvent()` instead. I then pasted the test string ""𐐌𐐜𐐬" and ensured that the new `TerminalInput::_SendChar` method still correctly assembles surrogate pairs. Closes #3516 Closes #3554 (obsoleted by this PR) Potentially impacts #391, which sounds like a duplicate of #3516
2020-04-07 21:09:28 +02:00
// Delegate it to the character event handler if this key event can be
// mapped to one (see method description above). For Alt+key combinations
// we'll not receive another character event for some reason though.
// -> Don't delegate the event if this is a Alt+key combination.
//
// As a special case we'll furthermore always handle VK_TAB
// key events here instead of in Terminal::SendCharEvent.
// See the method description for more information.
if (!isAltOnlyPressed && vkey != VK_TAB && ch != UNICODE_NULL)
{
Delegate all character input to the character event handler (#4192) My basic idea was that `WM_CHAR` is just the better `WM_KEYDOWN`. The latter fails to properly support common dead key sequences like in #3516. As such I added some logic to `Terminal::SendKeyEvent` to make it return false if the pressed key represents a printable character. This causes us to receive a character event with a (hopefully) correctly composed code unit, which then gets sent to `Terminal::SendCharEvent`. `Terminal::SendCharEvent` in turn had to be modified to support potentially pressed modifier keys, since `Terminal::SendKeyEvent` isn't doing that for us anymore. Lastly `TerminalInput` had to be modified heavily to support character events with modifier key states. In order to do so I merged its `HandleKey` and `HandleChar` methods into a single one, that now handles both cases. Since key events will now contain character data and character events key codes the decision logic in `TerminalInput::HandleKey` had to be rewritten. ## PR Checklist * [x] CLA signed * [x] Tests added/passed * [x] I've discussed this with core contributors already. ## Validation Steps Performed * See #3516. * I don't have any keyboard that generates surrogate characters. Due to this I modified `TermControl::_SendPastedTextToConnection` to send the data to `_terminal->SendCharEvent()` instead. I then pasted the test string ""𐐌𐐜𐐬" and ensured that the new `TerminalInput::_SendChar` method still correctly assembles surrogate pairs. Closes #3516 Closes #3554 (obsoleted by this PR) Potentially impacts #391, which sounds like a duplicate of #3516
2020-04-07 21:09:28 +02:00
return false;
}
KeyEvent keyEv{ keyDown, 1, vkey, sc, ch, states.Value() };
Delegate all character input to the character event handler (#4192) My basic idea was that `WM_CHAR` is just the better `WM_KEYDOWN`. The latter fails to properly support common dead key sequences like in #3516. As such I added some logic to `Terminal::SendKeyEvent` to make it return false if the pressed key represents a printable character. This causes us to receive a character event with a (hopefully) correctly composed code unit, which then gets sent to `Terminal::SendCharEvent`. `Terminal::SendCharEvent` in turn had to be modified to support potentially pressed modifier keys, since `Terminal::SendKeyEvent` isn't doing that for us anymore. Lastly `TerminalInput` had to be modified heavily to support character events with modifier key states. In order to do so I merged its `HandleKey` and `HandleChar` methods into a single one, that now handles both cases. Since key events will now contain character data and character events key codes the decision logic in `TerminalInput::HandleKey` had to be rewritten. ## PR Checklist * [x] CLA signed * [x] Tests added/passed * [x] I've discussed this with core contributors already. ## Validation Steps Performed * See #3516. * I don't have any keyboard that generates surrogate characters. Due to this I modified `TermControl::_SendPastedTextToConnection` to send the data to `_terminal->SendCharEvent()` instead. I then pasted the test string ""𐐌𐐜𐐬" and ensured that the new `TerminalInput::_SendChar` method still correctly assembles surrogate pairs. Closes #3516 Closes #3554 (obsoleted by this PR) Potentially impacts #391, which sounds like a duplicate of #3516
2020-04-07 21:09:28 +02:00
return _terminalInput->HandleKey(&keyEv);
}
// Method Description:
// - Send this particular mouse event to the terminal. The terminal will translate
// the button and the modifiers pressed into the appropriate VT sequence for that
// mouse event. If we do translate the key, we'll return true. In that case, the
// event should NOT be processed any further. If we return false, the event
// was NOT translated, and we should instead use the event normally
// Arguments:
// - viewportPos: the position of the mouse event relative to the viewport origin.
// - uiButton: the WM mouse button event code
// - states: The Microsoft::Terminal::Core::ControlKeyStates representing the modifier key states.
// - wheelDelta: the amount that the scroll wheel changed (should be 0 unless button is a WM_MOUSE*WHEEL)
// Return Value:
// - true if we translated the key event, and it should not be processed any further.
// - false if we did not translate the key, and it should be processed into a character.
bool Terminal::SendMouseEvent(const COORD viewportPos, const unsigned int uiButton, const ControlKeyStates states, const short wheelDelta, const TerminalInput::MouseButtonState state)
{
// GH#6401: VT applications should be able to receive mouse events from outside the
// terminal buffer. This is likely to happen when the user drags the cursor offscreen.
// We shouldn't throw away perfectly good events when they're offscreen, so we just
// clamp them to be within the range [(0, 0), (W, H)].
#pragma warning(suppress : 26496) // analysis can't tell we're assigning through a reference below
auto clampedPos{ viewportPos };
_mutableViewport.ToOrigin().Clamp(clampedPos);
return _terminalInput->HandleMouse(clampedPos, uiButton, GET_KEYSTATE_WPARAM(states.Value()), wheelDelta, state);
}
Delegate all character input to the character event handler (#4192) My basic idea was that `WM_CHAR` is just the better `WM_KEYDOWN`. The latter fails to properly support common dead key sequences like in #3516. As such I added some logic to `Terminal::SendKeyEvent` to make it return false if the pressed key represents a printable character. This causes us to receive a character event with a (hopefully) correctly composed code unit, which then gets sent to `Terminal::SendCharEvent`. `Terminal::SendCharEvent` in turn had to be modified to support potentially pressed modifier keys, since `Terminal::SendKeyEvent` isn't doing that for us anymore. Lastly `TerminalInput` had to be modified heavily to support character events with modifier key states. In order to do so I merged its `HandleKey` and `HandleChar` methods into a single one, that now handles both cases. Since key events will now contain character data and character events key codes the decision logic in `TerminalInput::HandleKey` had to be rewritten. ## PR Checklist * [x] CLA signed * [x] Tests added/passed * [x] I've discussed this with core contributors already. ## Validation Steps Performed * See #3516. * I don't have any keyboard that generates surrogate characters. Due to this I modified `TermControl::_SendPastedTextToConnection` to send the data to `_terminal->SendCharEvent()` instead. I then pasted the test string ""𐐌𐐜𐐬" and ensured that the new `TerminalInput::_SendChar` method still correctly assembles surrogate pairs. Closes #3516 Closes #3554 (obsoleted by this PR) Potentially impacts #391, which sounds like a duplicate of #3516
2020-04-07 21:09:28 +02:00
// Method Description:
// - Send this particular character to the terminal.
// - This method is the counterpart to SendKeyEvent and behaves almost identical.
// The difference is the focus on sending characters to the terminal,
// whereas SendKeyEvent handles the sending of keys like the arrow keys.
// Arguments:
// - ch: The UTF-16 code unit to be sent.
// - scanCode: The scan code of the last pressed key. Can be left 0.
// - states: The Microsoft::Terminal::Core::ControlKeyStates representing the modifier key states.
// Return Value:
// - true if we translated the character event, and it should not be processed any further.
// - false otherwise.
bool Terminal::SendCharEvent(const wchar_t ch, const WORD scanCode, const ControlKeyStates states)
{
Delegate all character input to the character event handler (#4192) My basic idea was that `WM_CHAR` is just the better `WM_KEYDOWN`. The latter fails to properly support common dead key sequences like in #3516. As such I added some logic to `Terminal::SendKeyEvent` to make it return false if the pressed key represents a printable character. This causes us to receive a character event with a (hopefully) correctly composed code unit, which then gets sent to `Terminal::SendCharEvent`. `Terminal::SendCharEvent` in turn had to be modified to support potentially pressed modifier keys, since `Terminal::SendKeyEvent` isn't doing that for us anymore. Lastly `TerminalInput` had to be modified heavily to support character events with modifier key states. In order to do so I merged its `HandleKey` and `HandleChar` methods into a single one, that now handles both cases. Since key events will now contain character data and character events key codes the decision logic in `TerminalInput::HandleKey` had to be rewritten. ## PR Checklist * [x] CLA signed * [x] Tests added/passed * [x] I've discussed this with core contributors already. ## Validation Steps Performed * See #3516. * I don't have any keyboard that generates surrogate characters. Due to this I modified `TermControl::_SendPastedTextToConnection` to send the data to `_terminal->SendCharEvent()` instead. I then pasted the test string ""𐐌𐐜𐐬" and ensured that the new `TerminalInput::_SendChar` method still correctly assembles surrogate pairs. Closes #3516 Closes #3554 (obsoleted by this PR) Potentially impacts #391, which sounds like a duplicate of #3516
2020-04-07 21:09:28 +02:00
auto vkey = _TakeVirtualKeyFromLastKeyEvent(scanCode);
if (vkey == 0 && scanCode != 0)
{
vkey = _VirtualKeyFromScanCode(scanCode);
}
if (vkey == 0)
{
vkey = _VirtualKeyFromCharacter(ch);
}
// Unfortunately, the UI doesn't give us both a character down and a
// character up event, only a character received event. So fake sending both
// to the terminal input translator. Unless it's in win32-input-mode, it'll
// ignore the keyup.
KeyEvent keyDown{ true, 1, vkey, scanCode, ch, states.Value() };
KeyEvent keyUp{ false, 1, vkey, scanCode, ch, states.Value() };
const auto handledDown = _terminalInput->HandleKey(&keyDown);
const auto handledUp = _terminalInput->HandleKey(&keyUp);
return handledDown || handledUp;
}
// Method Description:
// - Invalidates the regions described in the given pattern tree for the rendering purposes
// Arguments:
// - The interval tree containing regions that need to be invalidated
void Terminal::_InvalidatePatternTree(interval_tree::IntervalTree<til::point, size_t>& tree)
{
const auto vis = _VisibleStartIndex();
auto invalidate = [=](const PointTree::interval& interval) {
COORD startCoord{ gsl::narrow<SHORT>(interval.start.x()), gsl::narrow<SHORT>(interval.start.y() + vis) };
COORD endCoord{ gsl::narrow<SHORT>(interval.stop.x()), gsl::narrow<SHORT>(interval.stop.y() + vis) };
_InvalidateFromCoords(startCoord, endCoord);
};
tree.visit_all(invalidate);
}
// Method Description:
// - Given start and end coords, invalidates all the regions between them
// Arguments:
// - The start and end coords
void Terminal::_InvalidateFromCoords(const COORD start, const COORD end)
{
if (start.Y == end.Y)
{
SMALL_RECT region{ start.X, start.Y, end.X, end.Y };
_buffer->GetRenderTarget().TriggerRedraw(Viewport::FromInclusive(region));
}
else
{
const auto rowSize = gsl::narrow<SHORT>(_buffer->GetRowByOffset(0).size());
// invalidate the first line
SMALL_RECT region{ start.X, start.Y, rowSize - 1, start.Y };
_buffer->GetRenderTarget().TriggerRedraw(Viewport::FromInclusive(region));
if ((end.Y - start.Y) > 1)
{
// invalidate the lines in between the first and last line
region = til::rectangle(0, start.Y + 1, rowSize - 1, end.Y - 1);
_buffer->GetRenderTarget().TriggerRedraw(Viewport::FromInclusive(region));
}
// invalidate the last line
region = til::rectangle(0, end.Y, end.X, end.Y);
_buffer->GetRenderTarget().TriggerRedraw(Viewport::FromInclusive(region));
}
}
// Method Description:
// - Returns the keyboard's scan code for the given virtual key code.
// Arguments:
// - vkey: The virtual key code.
// Return Value:
// - The keyboard's scan code.
WORD Terminal::_ScanCodeFromVirtualKey(const WORD vkey) noexcept
{
return LOWORD(MapVirtualKeyW(vkey, MAPVK_VK_TO_VSC));
}
Delegate all character input to the character event handler (#4192) My basic idea was that `WM_CHAR` is just the better `WM_KEYDOWN`. The latter fails to properly support common dead key sequences like in #3516. As such I added some logic to `Terminal::SendKeyEvent` to make it return false if the pressed key represents a printable character. This causes us to receive a character event with a (hopefully) correctly composed code unit, which then gets sent to `Terminal::SendCharEvent`. `Terminal::SendCharEvent` in turn had to be modified to support potentially pressed modifier keys, since `Terminal::SendKeyEvent` isn't doing that for us anymore. Lastly `TerminalInput` had to be modified heavily to support character events with modifier key states. In order to do so I merged its `HandleKey` and `HandleChar` methods into a single one, that now handles both cases. Since key events will now contain character data and character events key codes the decision logic in `TerminalInput::HandleKey` had to be rewritten. ## PR Checklist * [x] CLA signed * [x] Tests added/passed * [x] I've discussed this with core contributors already. ## Validation Steps Performed * See #3516. * I don't have any keyboard that generates surrogate characters. Due to this I modified `TermControl::_SendPastedTextToConnection` to send the data to `_terminal->SendCharEvent()` instead. I then pasted the test string ""𐐌𐐜𐐬" and ensured that the new `TerminalInput::_SendChar` method still correctly assembles surrogate pairs. Closes #3516 Closes #3554 (obsoleted by this PR) Potentially impacts #391, which sounds like a duplicate of #3516
2020-04-07 21:09:28 +02:00
// Method Description:
// - Returns the virtual key code for the given keyboard's scan code.
// Arguments:
// - scanCode: The keyboard's scan code.
// Return Value:
// - The virtual key code. 0 if no mapping can be found.
WORD Terminal::_VirtualKeyFromScanCode(const WORD scanCode) noexcept
{
return LOWORD(MapVirtualKeyW(scanCode, MAPVK_VSC_TO_VK));
}
// Method Description:
// - Returns any virtual key code that produces the given character.
// Arguments:
// - scanCode: The keyboard's scan code.
// Return Value:
// - The virtual key code. 0 if no mapping can be found.
WORD Terminal::_VirtualKeyFromCharacter(const wchar_t ch) noexcept
{
const auto vkey = LOWORD(VkKeyScanW(ch));
return vkey == -1 ? 0 : vkey;
}
// Method Description:
// - Translates the specified virtual key code and keyboard state to the corresponding character.
// Arguments:
// - vkey: The virtual key code that initiated this keyboard event.
// - scanCode: The scan code that initiated this keyboard event.
// - states: The current keyboard state.
// Return Value:
// - The character that would result from this virtual key code and keyboard state.
wchar_t Terminal::_CharacterFromKeyEvent(const WORD vkey, const WORD scanCode, const ControlKeyStates states) noexcept
try
{
// We might want to use GetKeyboardState() instead of building our own keyState.
// The question is whether that's necessary though. For now it seems to work fine as it is.
std::array<BYTE, 256> keyState = {};
keyState.at(VK_SHIFT) = states.IsShiftPressed() ? 0x80 : 0;
keyState.at(VK_CONTROL) = states.IsCtrlPressed() ? 0x80 : 0;
keyState.at(VK_MENU) = states.IsAltPressed() ? 0x80 : 0;
2019-11-13 17:59:28 +01:00
// For the following use of ToUnicodeEx() please look here:
// https://docs.microsoft.com/en-us/windows/win32/api/winuser/nf-winuser-tounicodeex
// Technically ToUnicodeEx() can produce arbitrarily long sequences of diacritics etc.
// Since we only handle the case of a single UTF-16 code point, we can set the buffer size to 2 though.
std::array<wchar_t, 2> buffer;
2019-11-13 17:59:28 +01:00
// wFlags:
// * If bit 0 is set, a menu is active.
// If this flag is not specified ToUnicodeEx will send us character events on certain Alt+Key combinations (e.g. Alt+Arrow-Up).
// * If bit 2 is set, keyboard state is not changed (Windows 10, version 1607 and newer)
const auto result = ToUnicodeEx(vkey, scanCode, keyState.data(), buffer.data(), gsl::narrow_cast<int>(buffer.size()), 0b101, nullptr);
// TODO:GH#2853 We're only handling single UTF-16 code points right now, since that's the only thing KeyEvent supports.
return result == 1 || result == -1 ? buffer.at(0) : 0;
}
catch (...)
{
LOG_CAUGHT_EXCEPTION();
return UNICODE_INVALID;
}
Delegate all character input to the character event handler (#4192) My basic idea was that `WM_CHAR` is just the better `WM_KEYDOWN`. The latter fails to properly support common dead key sequences like in #3516. As such I added some logic to `Terminal::SendKeyEvent` to make it return false if the pressed key represents a printable character. This causes us to receive a character event with a (hopefully) correctly composed code unit, which then gets sent to `Terminal::SendCharEvent`. `Terminal::SendCharEvent` in turn had to be modified to support potentially pressed modifier keys, since `Terminal::SendKeyEvent` isn't doing that for us anymore. Lastly `TerminalInput` had to be modified heavily to support character events with modifier key states. In order to do so I merged its `HandleKey` and `HandleChar` methods into a single one, that now handles both cases. Since key events will now contain character data and character events key codes the decision logic in `TerminalInput::HandleKey` had to be rewritten. ## PR Checklist * [x] CLA signed * [x] Tests added/passed * [x] I've discussed this with core contributors already. ## Validation Steps Performed * See #3516. * I don't have any keyboard that generates surrogate characters. Due to this I modified `TermControl::_SendPastedTextToConnection` to send the data to `_terminal->SendCharEvent()` instead. I then pasted the test string ""𐐌𐐜𐐬" and ensured that the new `TerminalInput::_SendChar` method still correctly assembles surrogate pairs. Closes #3516 Closes #3554 (obsoleted by this PR) Potentially impacts #391, which sounds like a duplicate of #3516
2020-04-07 21:09:28 +02:00
// Method Description:
// - It's possible for a single scan code on a keyboard to
// produce different key codes depending on the keyboard state.
// MapVirtualKeyW(scanCode, MAPVK_VSC_TO_VK) will always chose one of the
// possibilities no matter what though and thus can't be used in SendCharEvent.
// - This method stores the key code from a key event (SendKeyEvent).
// If the key event contains character data, handling of the event will be
// denied, in order to delegate the work to the character event handler.
// - The character event handler (SendCharEvent) will now pick up
// the stored key code to restore the full key event data.
// Arguments:
// - vkey: The virtual key code.
// - scanCode: The scan code.
void Terminal::_StoreKeyEvent(const WORD vkey, const WORD scanCode)
{
_lastKeyEventCodes.emplace(KeyEventCodes{ vkey, scanCode });
}
// Method Description:
// - This method acts as a counterpart to _StoreKeyEvent and extracts a stored
// key code. As a safety measure it'll ensure that the given scan code
// matches the stored scan code from the previous key event.
// - See _StoreKeyEvent for more information.
// Arguments:
// - scanCode: The scan code.
// Return Value:
// - The key code matching the given scan code. Otherwise 0.
WORD Terminal::_TakeVirtualKeyFromLastKeyEvent(const WORD scanCode) noexcept
{
const auto codes = _lastKeyEventCodes.value_or(KeyEventCodes{});
_lastKeyEventCodes.reset();
return codes.ScanCode == scanCode ? codes.VirtualKey : 0;
}
// Method Description:
// - Acquire a read lock on the terminal.
// Return Value:
// - a shared_lock which can be used to unlock the terminal. The shared_lock
// will release this lock when it's destructed.
[[nodiscard]] std::unique_lock<til::ticket_lock> Terminal::LockForReading()
{
#ifdef NDEBUG
return std::unique_lock{ _readWriteLock };
#else
auto lock = std::unique_lock{ _readWriteLock };
_lastLocker = GetCurrentThreadId();
return lock;
#endif
}
// Method Description:
// - Acquire a write lock on the terminal.
// Return Value:
// - a unique_lock which can be used to unlock the terminal. The unique_lock
// will release this lock when it's destructed.
[[nodiscard]] std::unique_lock<til::ticket_lock> Terminal::LockForWriting()
{
#ifdef NDEBUG
return std::unique_lock{ _readWriteLock };
#else
auto lock = std::unique_lock{ _readWriteLock };
_lastLocker = GetCurrentThreadId();
return lock;
#endif
}
Viewport Terminal::_GetMutableViewport() const noexcept
{
return _mutableViewport;
}
short Terminal::GetBufferHeight() const noexcept
{
return _mutableViewport.BottomExclusive();
}
Search - add search box control and implement search experience (#3590) <!-- 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)? --> This is the PR for feature Search: #605 This PR includes the newly introduced SearchBoxControl in TermControl dir, which is the search bar for the search experience. And the codes that enable Search in Windows Terminal. <!-- Other than the issue solved, is this relevant to any other issues/existing PRs? --> The PR that migrates the Conhost search module: https://github.com/microsoft/terminal/pull/3279 Spec (still actively updating): https://github.com/microsoft/terminal/pull/3299 <!-- Please review the items on the PR checklist before submitting--> ## PR Checklist * [x] Closes #605 * [ ] 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 --> These functionalities are included in the search experience. 1. Search in Terminal text buffer. 2. Automatic wrap-around. 3. Search up or down switch by clicking different buttons. 4. Search case sensitively/insensitively by clicking a button. S. Move the search box to the top/bottom by clicking a button. 6. Close by clicking 'X'. 7. Open search by ctrl + F. When the searchbox is open, the user could still interact with the terminal by clicking the terminal input area. While I already have the search functionalities, currently there are still some known to-do works and I will keep updating my PR: 1. Optimize the search box UI, this includes: 1) Theme adaptation. The search box background and font color should change according to the theme, 2) Add background. Currently the elements in search box are all transparent. However, we need a background. 3) Move button should be highlighted once clicked. 2. Accessibility: search process should be able to performed without mouse. Once the search box is focused, the user should be able to navigate between all interactive elements on the searchbox using keyboard. <!-- Describe how you validated the behavior. Add automated tests wherever possible, but list manual validation steps taken as well --> To test: 1. checkout this branch. 2. Build the project. 3. Start Windows Terminal and press Ctrl+F 4. The search box should appear on the top right corner.
2019-12-17 16:52:37 +01:00
// ViewStartIndex is also the length of the scrollback
int Terminal::ViewStartIndex() const noexcept
{
return _mutableViewport.Top();
}
Search - add search box control and implement search experience (#3590) <!-- 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)? --> This is the PR for feature Search: #605 This PR includes the newly introduced SearchBoxControl in TermControl dir, which is the search bar for the search experience. And the codes that enable Search in Windows Terminal. <!-- Other than the issue solved, is this relevant to any other issues/existing PRs? --> The PR that migrates the Conhost search module: https://github.com/microsoft/terminal/pull/3279 Spec (still actively updating): https://github.com/microsoft/terminal/pull/3299 <!-- Please review the items on the PR checklist before submitting--> ## PR Checklist * [x] Closes #605 * [ ] 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 --> These functionalities are included in the search experience. 1. Search in Terminal text buffer. 2. Automatic wrap-around. 3. Search up or down switch by clicking different buttons. 4. Search case sensitively/insensitively by clicking a button. S. Move the search box to the top/bottom by clicking a button. 6. Close by clicking 'X'. 7. Open search by ctrl + F. When the searchbox is open, the user could still interact with the terminal by clicking the terminal input area. While I already have the search functionalities, currently there are still some known to-do works and I will keep updating my PR: 1. Optimize the search box UI, this includes: 1) Theme adaptation. The search box background and font color should change according to the theme, 2) Add background. Currently the elements in search box are all transparent. However, we need a background. 3) Move button should be highlighted once clicked. 2. Accessibility: search process should be able to performed without mouse. Once the search box is focused, the user should be able to navigate between all interactive elements on the searchbox using keyboard. <!-- Describe how you validated the behavior. Add automated tests wherever possible, but list manual validation steps taken as well --> To test: 1. checkout this branch. 2. Build the project. 3. Start Windows Terminal and press Ctrl+F 4. The search box should appear on the top right corner.
2019-12-17 16:52:37 +01:00
int Terminal::ViewEndIndex() const noexcept
{
return _mutableViewport.BottomInclusive();
}
// _VisibleStartIndex is the first visible line of the buffer
int Terminal::_VisibleStartIndex() const noexcept
{
Search - add search box control and implement search experience (#3590) <!-- 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)? --> This is the PR for feature Search: #605 This PR includes the newly introduced SearchBoxControl in TermControl dir, which is the search bar for the search experience. And the codes that enable Search in Windows Terminal. <!-- Other than the issue solved, is this relevant to any other issues/existing PRs? --> The PR that migrates the Conhost search module: https://github.com/microsoft/terminal/pull/3279 Spec (still actively updating): https://github.com/microsoft/terminal/pull/3299 <!-- Please review the items on the PR checklist before submitting--> ## PR Checklist * [x] Closes #605 * [ ] 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 --> These functionalities are included in the search experience. 1. Search in Terminal text buffer. 2. Automatic wrap-around. 3. Search up or down switch by clicking different buttons. 4. Search case sensitively/insensitively by clicking a button. S. Move the search box to the top/bottom by clicking a button. 6. Close by clicking 'X'. 7. Open search by ctrl + F. When the searchbox is open, the user could still interact with the terminal by clicking the terminal input area. While I already have the search functionalities, currently there are still some known to-do works and I will keep updating my PR: 1. Optimize the search box UI, this includes: 1) Theme adaptation. The search box background and font color should change according to the theme, 2) Add background. Currently the elements in search box are all transparent. However, we need a background. 3) Move button should be highlighted once clicked. 2. Accessibility: search process should be able to performed without mouse. Once the search box is focused, the user should be able to navigate between all interactive elements on the searchbox using keyboard. <!-- Describe how you validated the behavior. Add automated tests wherever possible, but list manual validation steps taken as well --> To test: 1. checkout this branch. 2. Build the project. 3. Start Windows Terminal and press Ctrl+F 4. The search box should appear on the top right corner.
2019-12-17 16:52:37 +01:00
return std::max(0, ViewStartIndex() - _scrollOffset);
}
int Terminal::_VisibleEndIndex() const noexcept
{
return std::max(0, ViewEndIndex() - _scrollOffset);
}
Viewport Terminal::_GetVisibleViewport() const noexcept
{
const COORD origin{ 0, gsl::narrow<short>(_VisibleStartIndex()) };
return Viewport::FromDimensions(origin,
_mutableViewport.Dimensions());
}
// Writes a string of text to the buffer, then moves the cursor (and viewport)
// in accordance with the written text.
// This method is our proverbial `WriteCharsLegacy`, and great care should be made to
// keep it minimal and orderly, lest it become WriteCharsLegacy2ElectricBoogaloo
// TODO: MSFT 21006766
// This needs to become stream logic on the buffer itself sooner rather than later
// because it's otherwise impossible to avoid the Electric Boogaloo-ness here.
// I had to make a bunch of hacks to get Japanese and emoji to work-ish.
void Terminal::_WriteBuffer(const std::wstring_view& stringView)
{
auto& cursor = _buffer->GetCursor();
// Defer the cursor drawing while we are iterating the string, for a better performance.
// We can not waste time displaying a cursor event when we know more text is coming right behind it.
cursor.StartDeferDrawing();
for (size_t i = 0; i < stringView.size(); i++)
{
const auto wch = stringView.at(i);
const COORD cursorPosBefore = cursor.GetPosition();
COORD proposedCursorPosition = cursorPosBefore;
Remove unneeded VT-specific control character handling (#4289) ## Summary of the Pull Request This PR removes all of the VT-specific functionality from the `WriteCharsLegacy` function that dealt with control characters, since those controls are now handled in the state machine when in VT mode. It also removes most of the control character handling from the `Terminal::_WriteBuffer` method for the same reason. ## References This is a followup to PR #4171 ## PR Checklist * [x] Closes #3971 * [x] 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 * [x] 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: https://github.com/microsoft/terminal/issues/780#issuecomment-570287435 ## Detailed Description of the Pull Request / Additional comments There are four changes to the `WriteCharsLegacy` implementation: 1. The `TAB` character had special case handling in VT mode which is now no longer required. This fixes a bug in the Python REPL editor (when run from a cmd shell in Windows Terminal), which would prevent you tabbing past the end of the line. It also fixes #3971. 2. Following on from point 1, the `WC_NONDESTRUCTIVE_TAB` flag could also now be removed. It only ever applied in VT mode, in which case the `TAB` character isn't handled in `WriteCharsLegacy`, so there isn't a need for a non-destructive version. 3. There used to be special case handling for a `BS` character at the beginning of the line when in VT mode, and that is also no longer required. This fixes an edge-case bug which would prevent a glyph being output for code point 8 when `ENABLE_PROCESSED_OUTPUT` was disabled. 4. There was quite a lot of special case handling for control characters in the "end-of-line wrap" implementation, which is no longer required. This fixes a bug which would prevent "low ASCII" characters from wrapping when output at the end of a line. Then in the `Terminal::_WriteBuffer` implementation, I've simply removed all control character handling, except for `LF`. The Terminal is always in VT mode, so the control characters are always handled by the state machine. The exception for the `LF` character is simply because it doesn't have a proper implementation yet, so it still passes the character through to `_WriteBuffer`. That will get cleaned up eventually, but I thought that could wait for a later PR. Finally, with the removal of the VT mode handling in `WriteCharsLegacy`, there was no longer a need for the `SCREEN_INFORMATION::InVTMode` method to be publicly accessible. That has now been made private. ## Validation Steps Performed I've only tested manually, making sure the conhost and Windows Terminal still basically work, and confirming that the above-mentioned bugs are fixed by these changes.
2020-01-29 20:18:46 +01:00
// TODO: MSFT 21006766
// This is not great but I need it demoable. Fix by making a buffer stream writer.
//
// If wch is a surrogate character we need to read 2 code units
// from the stringView to form a single code point.
const auto isSurrogate = wch >= 0xD800 && wch <= 0xDFFF;
const auto view = stringView.substr(i, isSurrogate ? 2 : 1);
const OutputCellIterator it{ view, _buffer->GetCurrentAttributes() };
const auto end = _buffer->Write(it);
const auto cellDistance = end.GetCellDistance(it);
const auto inputDistance = end.GetInputDistance(it);
if (inputDistance > 0)
{
Remove unneeded VT-specific control character handling (#4289) ## Summary of the Pull Request This PR removes all of the VT-specific functionality from the `WriteCharsLegacy` function that dealt with control characters, since those controls are now handled in the state machine when in VT mode. It also removes most of the control character handling from the `Terminal::_WriteBuffer` method for the same reason. ## References This is a followup to PR #4171 ## PR Checklist * [x] Closes #3971 * [x] 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 * [x] 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: https://github.com/microsoft/terminal/issues/780#issuecomment-570287435 ## Detailed Description of the Pull Request / Additional comments There are four changes to the `WriteCharsLegacy` implementation: 1. The `TAB` character had special case handling in VT mode which is now no longer required. This fixes a bug in the Python REPL editor (when run from a cmd shell in Windows Terminal), which would prevent you tabbing past the end of the line. It also fixes #3971. 2. Following on from point 1, the `WC_NONDESTRUCTIVE_TAB` flag could also now be removed. It only ever applied in VT mode, in which case the `TAB` character isn't handled in `WriteCharsLegacy`, so there isn't a need for a non-destructive version. 3. There used to be special case handling for a `BS` character at the beginning of the line when in VT mode, and that is also no longer required. This fixes an edge-case bug which would prevent a glyph being output for code point 8 when `ENABLE_PROCESSED_OUTPUT` was disabled. 4. There was quite a lot of special case handling for control characters in the "end-of-line wrap" implementation, which is no longer required. This fixes a bug which would prevent "low ASCII" characters from wrapping when output at the end of a line. Then in the `Terminal::_WriteBuffer` implementation, I've simply removed all control character handling, except for `LF`. The Terminal is always in VT mode, so the control characters are always handled by the state machine. The exception for the `LF` character is simply because it doesn't have a proper implementation yet, so it still passes the character through to `_WriteBuffer`. That will get cleaned up eventually, but I thought that could wait for a later PR. Finally, with the removal of the VT mode handling in `WriteCharsLegacy`, there was no longer a need for the `SCREEN_INFORMATION::InVTMode` method to be publicly accessible. That has now been made private. ## Validation Steps Performed I've only tested manually, making sure the conhost and Windows Terminal still basically work, and confirming that the above-mentioned bugs are fixed by these changes.
2020-01-29 20:18:46 +01:00
// If "wch" was a surrogate character, we just consumed 2 code units above.
// -> Increment "i" by 1 in that case and thus by 2 in total in this iteration.
proposedCursorPosition.X += gsl::narrow<SHORT>(cellDistance);
i += inputDistance - 1;
}
else
{
Remove unneeded VT-specific control character handling (#4289) ## Summary of the Pull Request This PR removes all of the VT-specific functionality from the `WriteCharsLegacy` function that dealt with control characters, since those controls are now handled in the state machine when in VT mode. It also removes most of the control character handling from the `Terminal::_WriteBuffer` method for the same reason. ## References This is a followup to PR #4171 ## PR Checklist * [x] Closes #3971 * [x] 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 * [x] 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: https://github.com/microsoft/terminal/issues/780#issuecomment-570287435 ## Detailed Description of the Pull Request / Additional comments There are four changes to the `WriteCharsLegacy` implementation: 1. The `TAB` character had special case handling in VT mode which is now no longer required. This fixes a bug in the Python REPL editor (when run from a cmd shell in Windows Terminal), which would prevent you tabbing past the end of the line. It also fixes #3971. 2. Following on from point 1, the `WC_NONDESTRUCTIVE_TAB` flag could also now be removed. It only ever applied in VT mode, in which case the `TAB` character isn't handled in `WriteCharsLegacy`, so there isn't a need for a non-destructive version. 3. There used to be special case handling for a `BS` character at the beginning of the line when in VT mode, and that is also no longer required. This fixes an edge-case bug which would prevent a glyph being output for code point 8 when `ENABLE_PROCESSED_OUTPUT` was disabled. 4. There was quite a lot of special case handling for control characters in the "end-of-line wrap" implementation, which is no longer required. This fixes a bug which would prevent "low ASCII" characters from wrapping when output at the end of a line. Then in the `Terminal::_WriteBuffer` implementation, I've simply removed all control character handling, except for `LF`. The Terminal is always in VT mode, so the control characters are always handled by the state machine. The exception for the `LF` character is simply because it doesn't have a proper implementation yet, so it still passes the character through to `_WriteBuffer`. That will get cleaned up eventually, but I thought that could wait for a later PR. Finally, with the removal of the VT mode handling in `WriteCharsLegacy`, there was no longer a need for the `SCREEN_INFORMATION::InVTMode` method to be publicly accessible. That has now been made private. ## Validation Steps Performed I've only tested manually, making sure the conhost and Windows Terminal still basically work, and confirming that the above-mentioned bugs are fixed by these changes.
2020-01-29 20:18:46 +01:00
// If _WriteBuffer() is called with a consecutive string longer than the viewport/buffer width
// the call to _buffer->Write() will refuse to write anything on the current line.
// GetInputDistance() thus returns 0, which would in turn cause i to be
// decremented by 1 below and force the outer loop to loop forever.
// This if() basically behaves as if "\r\n" had been encountered above and retries the write.
// With well behaving shells during normal operation this safeguard should normally not be encountered.
proposedCursorPosition.X = 0;
proposedCursorPosition.Y++;
Make Conpty emit wrapped lines as actually wrapped lines (#4415) ## Summary of the Pull Request Changes how conpty emits text to preserve line-wrap state, and additionally adds rudimentary support to the Windows Terminal for wrapped lines. ## References * Does _not_ fix (!) #3088, but that might be lower down in conhost. This makes wt behave like conhost, so at least there's that * Still needs a proper deferred EOL wrap implementation in #780, which is left as a todo * #4200 is the mega bucket with all this work * MSFT:16485846 was the first attempt at this task, which caused the regression MSFT:18123777 so we backed it out. * #4403 - I made sure this worked with that PR before I even sent #4403 ## PR Checklist * [x] Closes #405 * [x] Closes #3367 * [x] I work here * [x] Tests added/passed * [n/a] Requires documentation to be updated ## Detailed Description of the Pull Request / Additional comments I started with the following implementation: When conpty is about to write the last column, note that we wrapped this line here. If the next character the vt renderer is told to paint get is supposed to be at the start of the following line, then we know that the previous line had wrapped, so we _won't_ emit the usual `\r\n` here, and we'll just continue emitting text. However, this isn't _exactly_ right - if someone fills the row _exactly_ with text, the information that's available to the vt renderer isn't enough to know for sure if this line broke or not. It is possible for the client to write a full line of text, with a `\n` at the end, to manually break the line. So, I had to also add the `lineWrapped` param to the `IRenderEngine` interface, which is about half the files in this changelist. ## Validation Steps Performed * Ran tests * Checked how the Windows Terminal behaves with these changes * Made sure that conhost/inception and gnome-terminal both act as you'd expect with wrapped lines from conpty
2020-02-27 17:40:11 +01:00
// Try the character again.
i--;
Add support for "reflow"ing the Terminal buffer (#4741) This PR adds support for "Resize with Reflow" to the Terminal. In conhost, `ResizeWithReflow` is the function that's responsible for reflowing wrapped lines of text as the buffer gets resized. Now that #4415 has merged, we can also implement this in the Terminal. Now, when the Terminal is resized, it will reflow the lines of it's buffer in the same way that conhost does. This means, the terminal will no longer chop off the ends of lines as the buffer is too small to represent them. As a happy side effect of this PR, it also fixed #3490. This was a bug that plagued me during the investigation into this functionality. The original #3490 PR, #4354, tried to fix this bug with some heavy conpty changes. Turns out, that only made things worse, and far more complicated. When I really got to thinking about it, I realized "conhost can handle this right, why can't the Terminal?". Turns out, by adding resize with reflow, I was also able to fix this at the same time. Conhost does a little bit of math after reflowing to attempt to keep the viewport in the same relative place after a reflow. By re-using that logic in the Terminal, I was able to fix #3490. I also included that big ole test from #3490, because everyone likes adding 60 test cases in a PR. ## References * #4200 - this scenario * #405/#4415 - conpty emits wrapped lines, which was needed for this PR * #4403 - delayed EOL wrapping via conpty, which was also needed for this * #4354 - we don't speak of this PR anymore ## PR Checklist * [x] Closes #1465 * [x] Closes #3490 * [x] Closes #4771 * [x] Tests added/passed ## EDIT: Changes to this PR on 5 March 2020 I learned more since my original version of this PR. I wrote that in January, and despite my notes that say it was totally working, it _really_ wasn't. Part of the hard problem, as mentioned in #3490, is that the Terminal might request a resize to (W, H-1), and while conpty is preparing that frame, or before the terminal has received that frame, the Terminal resizes to (W, H-2). Now, there aren't enough lines in the terminal buffer to catch all the lines that conpty is about to emit. When that happens, lines get duplicated in the buffer. From a UX perspective, this certainly looks a lot worse than a couple lost lines. It looks like utter chaos. So I've introduced a new mode to conpty to try and counteract this behavior. This behavior I'm calling "quirky resize". The **TL;DR** of quirky resize mode is that conpty won't emit the entire buffer on a resize, and will trust that the terminal is prepared to reflow it's buffer on it's own. This will enable the quirky resize behavior for applications that are prepared for it. The "quirky resize" is "don't `InvalidateAll` when the terminal resizes". This is added as a quirk as to not regress other terminal applications that aren't prepared for this behavior (gnome-terminal, conhost in particular). For those kinds of terminals, when the buffer is resized, it's just going to lose lines. That's what currently happens for them. When the quirk is enabled, conpty won't repaint the entire buffer. This gets around the "duplicated lines" issue that requesting multiple resizes in a row can cause. However, for these terminals that are unprepared, the conpty cursor might end up in the wrong position after a quirky resize. The case in point is maximizing the terminal. For maximizing (height->50) from a buffer that's 30 lines tall, with the cursor on y=30, this is what happens: * With the quirk disabled, conpty reprints the entire buffer. This is 60 lines that get printed. This ends up blowing away about 20 lines of scrollback history, as the terminal app would have tried to keep the text pinned to the bottom of the window. The term. app moved the viewport up 20 lines, and then the 50 lines of conpty output (30 lines of text, and 20 blank lines at the bottom) overwrote the lines from the scrollback. This is bad, but not immediately obvious, and is **what currently happens**. * With the quirk enabled, conpty doesn't emit any lines, but the actual content of the window is still only in the top 30 lines. However, the terminal app has still moved 20 lines down from the scrollback back into the viewport. So the terminal's cursor is at y=50 now, but conpty's is at 30. This means that the terminal and conpty are out of sync, and there's not a good way of re-syncing these. It's very possible (trivial in `powershell`) that the new output will jump up to y=30 override the existing output in the terminal buffer. The Windows Terminal is already prepared for this quirky behavior, so it doesn't keep the output at the bottom of the window. It shifts it's viewport down to match what conpty things the buffer looks like. What happens when we have passthrough mode and WT is like "I would like quirky resize"? I guess things will just work fine, cause there won't be a buffer behind the passthrough app that the terminal cares about. Sure, in the passthrough case the Terminal could _not_ quirky resize, but the quirky resize won't be wrong.
2020-03-13 01:43:37 +01:00
// If we write the last cell of the row here, TextBuffer::Write will
// mark this line as wrapped for us. If the next character we
// process is a newline, the Terminal::CursorLineFeed will unmark
// this line as wrapped.
Make Conpty emit wrapped lines as actually wrapped lines (#4415) ## Summary of the Pull Request Changes how conpty emits text to preserve line-wrap state, and additionally adds rudimentary support to the Windows Terminal for wrapped lines. ## References * Does _not_ fix (!) #3088, but that might be lower down in conhost. This makes wt behave like conhost, so at least there's that * Still needs a proper deferred EOL wrap implementation in #780, which is left as a todo * #4200 is the mega bucket with all this work * MSFT:16485846 was the first attempt at this task, which caused the regression MSFT:18123777 so we backed it out. * #4403 - I made sure this worked with that PR before I even sent #4403 ## PR Checklist * [x] Closes #405 * [x] Closes #3367 * [x] I work here * [x] Tests added/passed * [n/a] Requires documentation to be updated ## Detailed Description of the Pull Request / Additional comments I started with the following implementation: When conpty is about to write the last column, note that we wrapped this line here. If the next character the vt renderer is told to paint get is supposed to be at the start of the following line, then we know that the previous line had wrapped, so we _won't_ emit the usual `\r\n` here, and we'll just continue emitting text. However, this isn't _exactly_ right - if someone fills the row _exactly_ with text, the information that's available to the vt renderer isn't enough to know for sure if this line broke or not. It is possible for the client to write a full line of text, with a `\n` at the end, to manually break the line. So, I had to also add the `lineWrapped` param to the `IRenderEngine` interface, which is about half the files in this changelist. ## Validation Steps Performed * Ran tests * Checked how the Windows Terminal behaves with these changes * Made sure that conhost/inception and gnome-terminal both act as you'd expect with wrapped lines from conpty
2020-02-27 17:40:11 +01:00
// TODO: GH#780 - This should really be a _deferred_ newline. If
// the next character to come in is a newline or a cursor
// movement or anything, then we should _not_ wrap this line
// here.
}
Remove unneeded VT-specific control character handling (#4289) ## Summary of the Pull Request This PR removes all of the VT-specific functionality from the `WriteCharsLegacy` function that dealt with control characters, since those controls are now handled in the state machine when in VT mode. It also removes most of the control character handling from the `Terminal::_WriteBuffer` method for the same reason. ## References This is a followup to PR #4171 ## PR Checklist * [x] Closes #3971 * [x] 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 * [x] 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: https://github.com/microsoft/terminal/issues/780#issuecomment-570287435 ## Detailed Description of the Pull Request / Additional comments There are four changes to the `WriteCharsLegacy` implementation: 1. The `TAB` character had special case handling in VT mode which is now no longer required. This fixes a bug in the Python REPL editor (when run from a cmd shell in Windows Terminal), which would prevent you tabbing past the end of the line. It also fixes #3971. 2. Following on from point 1, the `WC_NONDESTRUCTIVE_TAB` flag could also now be removed. It only ever applied in VT mode, in which case the `TAB` character isn't handled in `WriteCharsLegacy`, so there isn't a need for a non-destructive version. 3. There used to be special case handling for a `BS` character at the beginning of the line when in VT mode, and that is also no longer required. This fixes an edge-case bug which would prevent a glyph being output for code point 8 when `ENABLE_PROCESSED_OUTPUT` was disabled. 4. There was quite a lot of special case handling for control characters in the "end-of-line wrap" implementation, which is no longer required. This fixes a bug which would prevent "low ASCII" characters from wrapping when output at the end of a line. Then in the `Terminal::_WriteBuffer` implementation, I've simply removed all control character handling, except for `LF`. The Terminal is always in VT mode, so the control characters are always handled by the state machine. The exception for the `LF` character is simply because it doesn't have a proper implementation yet, so it still passes the character through to `_WriteBuffer`. That will get cleaned up eventually, but I thought that could wait for a later PR. Finally, with the removal of the VT mode handling in `WriteCharsLegacy`, there was no longer a need for the `SCREEN_INFORMATION::InVTMode` method to be publicly accessible. That has now been made private. ## Validation Steps Performed I've only tested manually, making sure the conhost and Windows Terminal still basically work, and confirming that the above-mentioned bugs are fixed by these changes.
2020-01-29 20:18:46 +01:00
_AdjustCursorPosition(proposedCursorPosition);
}
Remove unneeded VT-specific control character handling (#4289) ## Summary of the Pull Request This PR removes all of the VT-specific functionality from the `WriteCharsLegacy` function that dealt with control characters, since those controls are now handled in the state machine when in VT mode. It also removes most of the control character handling from the `Terminal::_WriteBuffer` method for the same reason. ## References This is a followup to PR #4171 ## PR Checklist * [x] Closes #3971 * [x] 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 * [x] 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: https://github.com/microsoft/terminal/issues/780#issuecomment-570287435 ## Detailed Description of the Pull Request / Additional comments There are four changes to the `WriteCharsLegacy` implementation: 1. The `TAB` character had special case handling in VT mode which is now no longer required. This fixes a bug in the Python REPL editor (when run from a cmd shell in Windows Terminal), which would prevent you tabbing past the end of the line. It also fixes #3971. 2. Following on from point 1, the `WC_NONDESTRUCTIVE_TAB` flag could also now be removed. It only ever applied in VT mode, in which case the `TAB` character isn't handled in `WriteCharsLegacy`, so there isn't a need for a non-destructive version. 3. There used to be special case handling for a `BS` character at the beginning of the line when in VT mode, and that is also no longer required. This fixes an edge-case bug which would prevent a glyph being output for code point 8 when `ENABLE_PROCESSED_OUTPUT` was disabled. 4. There was quite a lot of special case handling for control characters in the "end-of-line wrap" implementation, which is no longer required. This fixes a bug which would prevent "low ASCII" characters from wrapping when output at the end of a line. Then in the `Terminal::_WriteBuffer` implementation, I've simply removed all control character handling, except for `LF`. The Terminal is always in VT mode, so the control characters are always handled by the state machine. The exception for the `LF` character is simply because it doesn't have a proper implementation yet, so it still passes the character through to `_WriteBuffer`. That will get cleaned up eventually, but I thought that could wait for a later PR. Finally, with the removal of the VT mode handling in `WriteCharsLegacy`, there was no longer a need for the `SCREEN_INFORMATION::InVTMode` method to be publicly accessible. That has now been made private. ## Validation Steps Performed I've only tested manually, making sure the conhost and Windows Terminal still basically work, and confirming that the above-mentioned bugs are fixed by these changes.
2020-01-29 20:18:46 +01:00
cursor.EndDeferDrawing();
}
Remove unneeded VT-specific control character handling (#4289) ## Summary of the Pull Request This PR removes all of the VT-specific functionality from the `WriteCharsLegacy` function that dealt with control characters, since those controls are now handled in the state machine when in VT mode. It also removes most of the control character handling from the `Terminal::_WriteBuffer` method for the same reason. ## References This is a followup to PR #4171 ## PR Checklist * [x] Closes #3971 * [x] 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 * [x] 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: https://github.com/microsoft/terminal/issues/780#issuecomment-570287435 ## Detailed Description of the Pull Request / Additional comments There are four changes to the `WriteCharsLegacy` implementation: 1. The `TAB` character had special case handling in VT mode which is now no longer required. This fixes a bug in the Python REPL editor (when run from a cmd shell in Windows Terminal), which would prevent you tabbing past the end of the line. It also fixes #3971. 2. Following on from point 1, the `WC_NONDESTRUCTIVE_TAB` flag could also now be removed. It only ever applied in VT mode, in which case the `TAB` character isn't handled in `WriteCharsLegacy`, so there isn't a need for a non-destructive version. 3. There used to be special case handling for a `BS` character at the beginning of the line when in VT mode, and that is also no longer required. This fixes an edge-case bug which would prevent a glyph being output for code point 8 when `ENABLE_PROCESSED_OUTPUT` was disabled. 4. There was quite a lot of special case handling for control characters in the "end-of-line wrap" implementation, which is no longer required. This fixes a bug which would prevent "low ASCII" characters from wrapping when output at the end of a line. Then in the `Terminal::_WriteBuffer` implementation, I've simply removed all control character handling, except for `LF`. The Terminal is always in VT mode, so the control characters are always handled by the state machine. The exception for the `LF` character is simply because it doesn't have a proper implementation yet, so it still passes the character through to `_WriteBuffer`. That will get cleaned up eventually, but I thought that could wait for a later PR. Finally, with the removal of the VT mode handling in `WriteCharsLegacy`, there was no longer a need for the `SCREEN_INFORMATION::InVTMode` method to be publicly accessible. That has now been made private. ## Validation Steps Performed I've only tested manually, making sure the conhost and Windows Terminal still basically work, and confirming that the above-mentioned bugs are fixed by these changes.
2020-01-29 20:18:46 +01:00
void Terminal::_AdjustCursorPosition(const COORD proposedPosition)
{
#pragma warning(suppress : 26496) // cpp core checks wants this const but it's modified below.
auto proposedCursorPosition = proposedPosition;
auto& cursor = _buffer->GetCursor();
const Viewport bufferSize = _buffer->GetSize();
// If we're about to scroll past the bottom of the buffer, instead cycle the
// buffer.
Throttle scrollbar updates in TermControl to ~one per 8ms (#4608) In addition to the below (original) description, this commit introduces a ThrottledFunc template that can throttle _any_ function. It applies that type to muffle updates to the scrollbar. --- Redo #3531 but without the bug that it caused (#3622) which is why it was reverted. I'm sorry if I explain this badly. If you don't understand a part, make sure to let me know and I will explain it better. ### Explanation How it worked before: `Terminal` signals that viewport changed -> `TermControl::_TerminalScrollPositionChanged` gets called on the terminal thread -> it dispatches work for later to be ran the UI thread to updates the scrollbar's values Why it's bad: * If we have many viewport changes, it will create a long stack of operations to run. Instead, we should just update the scroll bar with the most recent information that we know. * Imagine if the rate that the work gets pushed on the UI thread is greater than the rate that it can handle: it might freeze? * No need to be real time, we can wait just a little bit (8ms) to accumulate viewport changes before we actually change the scroll bar's value because it appears to be expensive (see perf below). Now: `Terminal` signals that viewport changed -> `TermControl::_TerminalScrollPositionChanged` gets called on the terminal thread -> it tells the `ScrollBarUpdater` about a new update -> the `ScrollBarUpdater` only runs one job (I don't know if that's the right term) on the UI thread at a time. If a job is already running but hasn't updated the scroll bar yet, it changes the setting in the already existing job to update the scroll bar with the new values. A job "waits" some time before doing the update to throttle updates because we don't need real time scroll bar updates. -> eventually, it updates the scroll bar If the user scrolls when a scroll bar update is pending, we keep the scroll bar's Maximum and Minimum but let the user choose its new Value with the `CancelPendingValueChange` method. ### Note Also I changed a little bit the code from the Terminal to notify the TermControl less often when possible. I tried to scroll with the scroll bar, with the mouse wheel. I tried to scroll while content is being outputted. I tried to reproduce the crash from #2248 without success (good). Co-authored-by: Leonard Hecker <leonard@hecker.io> Closes #3622
2020-06-12 21:51:37 +02:00
SHORT rowsPushedOffTopOfBuffer = 0;
const auto newRows = std::max(0, proposedCursorPosition.Y - bufferSize.Height() + 1);
Throttle scrollbar updates in TermControl to ~one per 8ms (#4608) In addition to the below (original) description, this commit introduces a ThrottledFunc template that can throttle _any_ function. It applies that type to muffle updates to the scrollbar. --- Redo #3531 but without the bug that it caused (#3622) which is why it was reverted. I'm sorry if I explain this badly. If you don't understand a part, make sure to let me know and I will explain it better. ### Explanation How it worked before: `Terminal` signals that viewport changed -> `TermControl::_TerminalScrollPositionChanged` gets called on the terminal thread -> it dispatches work for later to be ran the UI thread to updates the scrollbar's values Why it's bad: * If we have many viewport changes, it will create a long stack of operations to run. Instead, we should just update the scroll bar with the most recent information that we know. * Imagine if the rate that the work gets pushed on the UI thread is greater than the rate that it can handle: it might freeze? * No need to be real time, we can wait just a little bit (8ms) to accumulate viewport changes before we actually change the scroll bar's value because it appears to be expensive (see perf below). Now: `Terminal` signals that viewport changed -> `TermControl::_TerminalScrollPositionChanged` gets called on the terminal thread -> it tells the `ScrollBarUpdater` about a new update -> the `ScrollBarUpdater` only runs one job (I don't know if that's the right term) on the UI thread at a time. If a job is already running but hasn't updated the scroll bar yet, it changes the setting in the already existing job to update the scroll bar with the new values. A job "waits" some time before doing the update to throttle updates because we don't need real time scroll bar updates. -> eventually, it updates the scroll bar If the user scrolls when a scroll bar update is pending, we keep the scroll bar's Maximum and Minimum but let the user choose its new Value with the `CancelPendingValueChange` method. ### Note Also I changed a little bit the code from the Terminal to notify the TermControl less often when possible. I tried to scroll with the scroll bar, with the mouse wheel. I tried to scroll while content is being outputted. I tried to reproduce the crash from #2248 without success (good). Co-authored-by: Leonard Hecker <leonard@hecker.io> Closes #3622
2020-06-12 21:51:37 +02:00
if (proposedCursorPosition.Y >= bufferSize.Height())
Remove unneeded VT-specific control character handling (#4289) ## Summary of the Pull Request This PR removes all of the VT-specific functionality from the `WriteCharsLegacy` function that dealt with control characters, since those controls are now handled in the state machine when in VT mode. It also removes most of the control character handling from the `Terminal::_WriteBuffer` method for the same reason. ## References This is a followup to PR #4171 ## PR Checklist * [x] Closes #3971 * [x] 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 * [x] 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: https://github.com/microsoft/terminal/issues/780#issuecomment-570287435 ## Detailed Description of the Pull Request / Additional comments There are four changes to the `WriteCharsLegacy` implementation: 1. The `TAB` character had special case handling in VT mode which is now no longer required. This fixes a bug in the Python REPL editor (when run from a cmd shell in Windows Terminal), which would prevent you tabbing past the end of the line. It also fixes #3971. 2. Following on from point 1, the `WC_NONDESTRUCTIVE_TAB` flag could also now be removed. It only ever applied in VT mode, in which case the `TAB` character isn't handled in `WriteCharsLegacy`, so there isn't a need for a non-destructive version. 3. There used to be special case handling for a `BS` character at the beginning of the line when in VT mode, and that is also no longer required. This fixes an edge-case bug which would prevent a glyph being output for code point 8 when `ENABLE_PROCESSED_OUTPUT` was disabled. 4. There was quite a lot of special case handling for control characters in the "end-of-line wrap" implementation, which is no longer required. This fixes a bug which would prevent "low ASCII" characters from wrapping when output at the end of a line. Then in the `Terminal::_WriteBuffer` implementation, I've simply removed all control character handling, except for `LF`. The Terminal is always in VT mode, so the control characters are always handled by the state machine. The exception for the `LF` character is simply because it doesn't have a proper implementation yet, so it still passes the character through to `_WriteBuffer`. That will get cleaned up eventually, but I thought that could wait for a later PR. Finally, with the removal of the VT mode handling in `WriteCharsLegacy`, there was no longer a need for the `SCREEN_INFORMATION::InVTMode` method to be publicly accessible. That has now been made private. ## Validation Steps Performed I've only tested manually, making sure the conhost and Windows Terminal still basically work, and confirming that the above-mentioned bugs are fixed by these changes.
2020-01-29 20:18:46 +01:00
{
for (auto dy = 0; dy < newRows; dy++)
{
Remove unneeded VT-specific control character handling (#4289) ## Summary of the Pull Request This PR removes all of the VT-specific functionality from the `WriteCharsLegacy` function that dealt with control characters, since those controls are now handled in the state machine when in VT mode. It also removes most of the control character handling from the `Terminal::_WriteBuffer` method for the same reason. ## References This is a followup to PR #4171 ## PR Checklist * [x] Closes #3971 * [x] 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 * [x] 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: https://github.com/microsoft/terminal/issues/780#issuecomment-570287435 ## Detailed Description of the Pull Request / Additional comments There are four changes to the `WriteCharsLegacy` implementation: 1. The `TAB` character had special case handling in VT mode which is now no longer required. This fixes a bug in the Python REPL editor (when run from a cmd shell in Windows Terminal), which would prevent you tabbing past the end of the line. It also fixes #3971. 2. Following on from point 1, the `WC_NONDESTRUCTIVE_TAB` flag could also now be removed. It only ever applied in VT mode, in which case the `TAB` character isn't handled in `WriteCharsLegacy`, so there isn't a need for a non-destructive version. 3. There used to be special case handling for a `BS` character at the beginning of the line when in VT mode, and that is also no longer required. This fixes an edge-case bug which would prevent a glyph being output for code point 8 when `ENABLE_PROCESSED_OUTPUT` was disabled. 4. There was quite a lot of special case handling for control characters in the "end-of-line wrap" implementation, which is no longer required. This fixes a bug which would prevent "low ASCII" characters from wrapping when output at the end of a line. Then in the `Terminal::_WriteBuffer` implementation, I've simply removed all control character handling, except for `LF`. The Terminal is always in VT mode, so the control characters are always handled by the state machine. The exception for the `LF` character is simply because it doesn't have a proper implementation yet, so it still passes the character through to `_WriteBuffer`. That will get cleaned up eventually, but I thought that could wait for a later PR. Finally, with the removal of the VT mode handling in `WriteCharsLegacy`, there was no longer a need for the `SCREEN_INFORMATION::InVTMode` method to be publicly accessible. That has now been made private. ## Validation Steps Performed I've only tested manually, making sure the conhost and Windows Terminal still basically work, and confirming that the above-mentioned bugs are fixed by these changes.
2020-01-29 20:18:46 +01:00
_buffer->IncrementCircularBuffer();
proposedCursorPosition.Y--;
Throttle scrollbar updates in TermControl to ~one per 8ms (#4608) In addition to the below (original) description, this commit introduces a ThrottledFunc template that can throttle _any_ function. It applies that type to muffle updates to the scrollbar. --- Redo #3531 but without the bug that it caused (#3622) which is why it was reverted. I'm sorry if I explain this badly. If you don't understand a part, make sure to let me know and I will explain it better. ### Explanation How it worked before: `Terminal` signals that viewport changed -> `TermControl::_TerminalScrollPositionChanged` gets called on the terminal thread -> it dispatches work for later to be ran the UI thread to updates the scrollbar's values Why it's bad: * If we have many viewport changes, it will create a long stack of operations to run. Instead, we should just update the scroll bar with the most recent information that we know. * Imagine if the rate that the work gets pushed on the UI thread is greater than the rate that it can handle: it might freeze? * No need to be real time, we can wait just a little bit (8ms) to accumulate viewport changes before we actually change the scroll bar's value because it appears to be expensive (see perf below). Now: `Terminal` signals that viewport changed -> `TermControl::_TerminalScrollPositionChanged` gets called on the terminal thread -> it tells the `ScrollBarUpdater` about a new update -> the `ScrollBarUpdater` only runs one job (I don't know if that's the right term) on the UI thread at a time. If a job is already running but hasn't updated the scroll bar yet, it changes the setting in the already existing job to update the scroll bar with the new values. A job "waits" some time before doing the update to throttle updates because we don't need real time scroll bar updates. -> eventually, it updates the scroll bar If the user scrolls when a scroll bar update is pending, we keep the scroll bar's Maximum and Minimum but let the user choose its new Value with the `CancelPendingValueChange` method. ### Note Also I changed a little bit the code from the Terminal to notify the TermControl less often when possible. I tried to scroll with the scroll bar, with the mouse wheel. I tried to scroll while content is being outputted. I tried to reproduce the crash from #2248 without success (good). Co-authored-by: Leonard Hecker <leonard@hecker.io> Closes #3622
2020-06-12 21:51:37 +02:00
rowsPushedOffTopOfBuffer++;
// Update our selection too, so it doesn't move as the buffer is cycled
if (_selection)
{
// If the start of the selection is above 0, we can reduce both the start and end by 1
if (_selection->start.Y > 0)
{
_selection->start.Y -= 1;
_selection->end.Y -= 1;
}
else
{
// The start of the selection is at 0, if the end is greater than 0, then only reduce the end
if (_selection->end.Y > 0)
{
_selection->start.X = 0;
_selection->end.Y -= 1;
}
else
{
// Both the start and end of the selection are at 0, clear the selection
_selection.reset();
}
}
}
}
// manually erase our pattern intervals since the locations have changed now
_patternIntervalTree = {};
Remove unneeded VT-specific control character handling (#4289) ## Summary of the Pull Request This PR removes all of the VT-specific functionality from the `WriteCharsLegacy` function that dealt with control characters, since those controls are now handled in the state machine when in VT mode. It also removes most of the control character handling from the `Terminal::_WriteBuffer` method for the same reason. ## References This is a followup to PR #4171 ## PR Checklist * [x] Closes #3971 * [x] 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 * [x] 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: https://github.com/microsoft/terminal/issues/780#issuecomment-570287435 ## Detailed Description of the Pull Request / Additional comments There are four changes to the `WriteCharsLegacy` implementation: 1. The `TAB` character had special case handling in VT mode which is now no longer required. This fixes a bug in the Python REPL editor (when run from a cmd shell in Windows Terminal), which would prevent you tabbing past the end of the line. It also fixes #3971. 2. Following on from point 1, the `WC_NONDESTRUCTIVE_TAB` flag could also now be removed. It only ever applied in VT mode, in which case the `TAB` character isn't handled in `WriteCharsLegacy`, so there isn't a need for a non-destructive version. 3. There used to be special case handling for a `BS` character at the beginning of the line when in VT mode, and that is also no longer required. This fixes an edge-case bug which would prevent a glyph being output for code point 8 when `ENABLE_PROCESSED_OUTPUT` was disabled. 4. There was quite a lot of special case handling for control characters in the "end-of-line wrap" implementation, which is no longer required. This fixes a bug which would prevent "low ASCII" characters from wrapping when output at the end of a line. Then in the `Terminal::_WriteBuffer` implementation, I've simply removed all control character handling, except for `LF`. The Terminal is always in VT mode, so the control characters are always handled by the state machine. The exception for the `LF` character is simply because it doesn't have a proper implementation yet, so it still passes the character through to `_WriteBuffer`. That will get cleaned up eventually, but I thought that could wait for a later PR. Finally, with the removal of the VT mode handling in `WriteCharsLegacy`, there was no longer a need for the `SCREEN_INFORMATION::InVTMode` method to be publicly accessible. That has now been made private. ## Validation Steps Performed I've only tested manually, making sure the conhost and Windows Terminal still basically work, and confirming that the above-mentioned bugs are fixed by these changes.
2020-01-29 20:18:46 +01:00
}
// Update Cursor Position
cursor.SetPosition(proposedCursorPosition);
Remove unneeded VT-specific control character handling (#4289) ## Summary of the Pull Request This PR removes all of the VT-specific functionality from the `WriteCharsLegacy` function that dealt with control characters, since those controls are now handled in the state machine when in VT mode. It also removes most of the control character handling from the `Terminal::_WriteBuffer` method for the same reason. ## References This is a followup to PR #4171 ## PR Checklist * [x] Closes #3971 * [x] 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 * [x] 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: https://github.com/microsoft/terminal/issues/780#issuecomment-570287435 ## Detailed Description of the Pull Request / Additional comments There are four changes to the `WriteCharsLegacy` implementation: 1. The `TAB` character had special case handling in VT mode which is now no longer required. This fixes a bug in the Python REPL editor (when run from a cmd shell in Windows Terminal), which would prevent you tabbing past the end of the line. It also fixes #3971. 2. Following on from point 1, the `WC_NONDESTRUCTIVE_TAB` flag could also now be removed. It only ever applied in VT mode, in which case the `TAB` character isn't handled in `WriteCharsLegacy`, so there isn't a need for a non-destructive version. 3. There used to be special case handling for a `BS` character at the beginning of the line when in VT mode, and that is also no longer required. This fixes an edge-case bug which would prevent a glyph being output for code point 8 when `ENABLE_PROCESSED_OUTPUT` was disabled. 4. There was quite a lot of special case handling for control characters in the "end-of-line wrap" implementation, which is no longer required. This fixes a bug which would prevent "low ASCII" characters from wrapping when output at the end of a line. Then in the `Terminal::_WriteBuffer` implementation, I've simply removed all control character handling, except for `LF`. The Terminal is always in VT mode, so the control characters are always handled by the state machine. The exception for the `LF` character is simply because it doesn't have a proper implementation yet, so it still passes the character through to `_WriteBuffer`. That will get cleaned up eventually, but I thought that could wait for a later PR. Finally, with the removal of the VT mode handling in `WriteCharsLegacy`, there was no longer a need for the `SCREEN_INFORMATION::InVTMode` method to be publicly accessible. That has now been made private. ## Validation Steps Performed I've only tested manually, making sure the conhost and Windows Terminal still basically work, and confirming that the above-mentioned bugs are fixed by these changes.
2020-01-29 20:18:46 +01:00
// Move the viewport down if the cursor moved below the viewport.
Throttle scrollbar updates in TermControl to ~one per 8ms (#4608) In addition to the below (original) description, this commit introduces a ThrottledFunc template that can throttle _any_ function. It applies that type to muffle updates to the scrollbar. --- Redo #3531 but without the bug that it caused (#3622) which is why it was reverted. I'm sorry if I explain this badly. If you don't understand a part, make sure to let me know and I will explain it better. ### Explanation How it worked before: `Terminal` signals that viewport changed -> `TermControl::_TerminalScrollPositionChanged` gets called on the terminal thread -> it dispatches work for later to be ran the UI thread to updates the scrollbar's values Why it's bad: * If we have many viewport changes, it will create a long stack of operations to run. Instead, we should just update the scroll bar with the most recent information that we know. * Imagine if the rate that the work gets pushed on the UI thread is greater than the rate that it can handle: it might freeze? * No need to be real time, we can wait just a little bit (8ms) to accumulate viewport changes before we actually change the scroll bar's value because it appears to be expensive (see perf below). Now: `Terminal` signals that viewport changed -> `TermControl::_TerminalScrollPositionChanged` gets called on the terminal thread -> it tells the `ScrollBarUpdater` about a new update -> the `ScrollBarUpdater` only runs one job (I don't know if that's the right term) on the UI thread at a time. If a job is already running but hasn't updated the scroll bar yet, it changes the setting in the already existing job to update the scroll bar with the new values. A job "waits" some time before doing the update to throttle updates because we don't need real time scroll bar updates. -> eventually, it updates the scroll bar If the user scrolls when a scroll bar update is pending, we keep the scroll bar's Maximum and Minimum but let the user choose its new Value with the `CancelPendingValueChange` method. ### Note Also I changed a little bit the code from the Terminal to notify the TermControl less often when possible. I tried to scroll with the scroll bar, with the mouse wheel. I tried to scroll while content is being outputted. I tried to reproduce the crash from #2248 without success (good). Co-authored-by: Leonard Hecker <leonard@hecker.io> Closes #3622
2020-06-12 21:51:37 +02:00
bool updatedViewport = false;
const auto scrollAmount = std::max(0, proposedCursorPosition.Y - _mutableViewport.BottomInclusive());
if (scrollAmount > 0)
Remove unneeded VT-specific control character handling (#4289) ## Summary of the Pull Request This PR removes all of the VT-specific functionality from the `WriteCharsLegacy` function that dealt with control characters, since those controls are now handled in the state machine when in VT mode. It also removes most of the control character handling from the `Terminal::_WriteBuffer` method for the same reason. ## References This is a followup to PR #4171 ## PR Checklist * [x] Closes #3971 * [x] 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 * [x] 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: https://github.com/microsoft/terminal/issues/780#issuecomment-570287435 ## Detailed Description of the Pull Request / Additional comments There are four changes to the `WriteCharsLegacy` implementation: 1. The `TAB` character had special case handling in VT mode which is now no longer required. This fixes a bug in the Python REPL editor (when run from a cmd shell in Windows Terminal), which would prevent you tabbing past the end of the line. It also fixes #3971. 2. Following on from point 1, the `WC_NONDESTRUCTIVE_TAB` flag could also now be removed. It only ever applied in VT mode, in which case the `TAB` character isn't handled in `WriteCharsLegacy`, so there isn't a need for a non-destructive version. 3. There used to be special case handling for a `BS` character at the beginning of the line when in VT mode, and that is also no longer required. This fixes an edge-case bug which would prevent a glyph being output for code point 8 when `ENABLE_PROCESSED_OUTPUT` was disabled. 4. There was quite a lot of special case handling for control characters in the "end-of-line wrap" implementation, which is no longer required. This fixes a bug which would prevent "low ASCII" characters from wrapping when output at the end of a line. Then in the `Terminal::_WriteBuffer` implementation, I've simply removed all control character handling, except for `LF`. The Terminal is always in VT mode, so the control characters are always handled by the state machine. The exception for the `LF` character is simply because it doesn't have a proper implementation yet, so it still passes the character through to `_WriteBuffer`. That will get cleaned up eventually, but I thought that could wait for a later PR. Finally, with the removal of the VT mode handling in `WriteCharsLegacy`, there was no longer a need for the `SCREEN_INFORMATION::InVTMode` method to be publicly accessible. That has now been made private. ## Validation Steps Performed I've only tested manually, making sure the conhost and Windows Terminal still basically work, and confirming that the above-mentioned bugs are fixed by these changes.
2020-01-29 20:18:46 +01:00
{
Throttle scrollbar updates in TermControl to ~one per 8ms (#4608) In addition to the below (original) description, this commit introduces a ThrottledFunc template that can throttle _any_ function. It applies that type to muffle updates to the scrollbar. --- Redo #3531 but without the bug that it caused (#3622) which is why it was reverted. I'm sorry if I explain this badly. If you don't understand a part, make sure to let me know and I will explain it better. ### Explanation How it worked before: `Terminal` signals that viewport changed -> `TermControl::_TerminalScrollPositionChanged` gets called on the terminal thread -> it dispatches work for later to be ran the UI thread to updates the scrollbar's values Why it's bad: * If we have many viewport changes, it will create a long stack of operations to run. Instead, we should just update the scroll bar with the most recent information that we know. * Imagine if the rate that the work gets pushed on the UI thread is greater than the rate that it can handle: it might freeze? * No need to be real time, we can wait just a little bit (8ms) to accumulate viewport changes before we actually change the scroll bar's value because it appears to be expensive (see perf below). Now: `Terminal` signals that viewport changed -> `TermControl::_TerminalScrollPositionChanged` gets called on the terminal thread -> it tells the `ScrollBarUpdater` about a new update -> the `ScrollBarUpdater` only runs one job (I don't know if that's the right term) on the UI thread at a time. If a job is already running but hasn't updated the scroll bar yet, it changes the setting in the already existing job to update the scroll bar with the new values. A job "waits" some time before doing the update to throttle updates because we don't need real time scroll bar updates. -> eventually, it updates the scroll bar If the user scrolls when a scroll bar update is pending, we keep the scroll bar's Maximum and Minimum but let the user choose its new Value with the `CancelPendingValueChange` method. ### Note Also I changed a little bit the code from the Terminal to notify the TermControl less often when possible. I tried to scroll with the scroll bar, with the mouse wheel. I tried to scroll while content is being outputted. I tried to reproduce the crash from #2248 without success (good). Co-authored-by: Leonard Hecker <leonard@hecker.io> Closes #3622
2020-06-12 21:51:37 +02:00
const auto newViewTop = std::max(0, proposedCursorPosition.Y - (_mutableViewport.Height() - 1));
Remove unneeded VT-specific control character handling (#4289) ## Summary of the Pull Request This PR removes all of the VT-specific functionality from the `WriteCharsLegacy` function that dealt with control characters, since those controls are now handled in the state machine when in VT mode. It also removes most of the control character handling from the `Terminal::_WriteBuffer` method for the same reason. ## References This is a followup to PR #4171 ## PR Checklist * [x] Closes #3971 * [x] 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 * [x] 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: https://github.com/microsoft/terminal/issues/780#issuecomment-570287435 ## Detailed Description of the Pull Request / Additional comments There are four changes to the `WriteCharsLegacy` implementation: 1. The `TAB` character had special case handling in VT mode which is now no longer required. This fixes a bug in the Python REPL editor (when run from a cmd shell in Windows Terminal), which would prevent you tabbing past the end of the line. It also fixes #3971. 2. Following on from point 1, the `WC_NONDESTRUCTIVE_TAB` flag could also now be removed. It only ever applied in VT mode, in which case the `TAB` character isn't handled in `WriteCharsLegacy`, so there isn't a need for a non-destructive version. 3. There used to be special case handling for a `BS` character at the beginning of the line when in VT mode, and that is also no longer required. This fixes an edge-case bug which would prevent a glyph being output for code point 8 when `ENABLE_PROCESSED_OUTPUT` was disabled. 4. There was quite a lot of special case handling for control characters in the "end-of-line wrap" implementation, which is no longer required. This fixes a bug which would prevent "low ASCII" characters from wrapping when output at the end of a line. Then in the `Terminal::_WriteBuffer` implementation, I've simply removed all control character handling, except for `LF`. The Terminal is always in VT mode, so the control characters are always handled by the state machine. The exception for the `LF` character is simply because it doesn't have a proper implementation yet, so it still passes the character through to `_WriteBuffer`. That will get cleaned up eventually, but I thought that could wait for a later PR. Finally, with the removal of the VT mode handling in `WriteCharsLegacy`, there was no longer a need for the `SCREEN_INFORMATION::InVTMode` method to be publicly accessible. That has now been made private. ## Validation Steps Performed I've only tested manually, making sure the conhost and Windows Terminal still basically work, and confirming that the above-mentioned bugs are fixed by these changes.
2020-01-29 20:18:46 +01:00
if (newViewTop != _mutableViewport.Top())
{
_mutableViewport = Viewport::FromDimensions({ 0, gsl::narrow<short>(newViewTop) },
_mutableViewport.Dimensions());
Throttle scrollbar updates in TermControl to ~one per 8ms (#4608) In addition to the below (original) description, this commit introduces a ThrottledFunc template that can throttle _any_ function. It applies that type to muffle updates to the scrollbar. --- Redo #3531 but without the bug that it caused (#3622) which is why it was reverted. I'm sorry if I explain this badly. If you don't understand a part, make sure to let me know and I will explain it better. ### Explanation How it worked before: `Terminal` signals that viewport changed -> `TermControl::_TerminalScrollPositionChanged` gets called on the terminal thread -> it dispatches work for later to be ran the UI thread to updates the scrollbar's values Why it's bad: * If we have many viewport changes, it will create a long stack of operations to run. Instead, we should just update the scroll bar with the most recent information that we know. * Imagine if the rate that the work gets pushed on the UI thread is greater than the rate that it can handle: it might freeze? * No need to be real time, we can wait just a little bit (8ms) to accumulate viewport changes before we actually change the scroll bar's value because it appears to be expensive (see perf below). Now: `Terminal` signals that viewport changed -> `TermControl::_TerminalScrollPositionChanged` gets called on the terminal thread -> it tells the `ScrollBarUpdater` about a new update -> the `ScrollBarUpdater` only runs one job (I don't know if that's the right term) on the UI thread at a time. If a job is already running but hasn't updated the scroll bar yet, it changes the setting in the already existing job to update the scroll bar with the new values. A job "waits" some time before doing the update to throttle updates because we don't need real time scroll bar updates. -> eventually, it updates the scroll bar If the user scrolls when a scroll bar update is pending, we keep the scroll bar's Maximum and Minimum but let the user choose its new Value with the `CancelPendingValueChange` method. ### Note Also I changed a little bit the code from the Terminal to notify the TermControl less often when possible. I tried to scroll with the scroll bar, with the mouse wheel. I tried to scroll while content is being outputted. I tried to reproduce the crash from #2248 without success (good). Co-authored-by: Leonard Hecker <leonard@hecker.io> Closes #3622
2020-06-12 21:51:37 +02:00
updatedViewport = true;
}
}
// If the viewport moved, or we circled the buffer, we might need to update
// our _scrollOffset
if (updatedViewport || newRows != 0)
Throttle scrollbar updates in TermControl to ~one per 8ms (#4608) In addition to the below (original) description, this commit introduces a ThrottledFunc template that can throttle _any_ function. It applies that type to muffle updates to the scrollbar. --- Redo #3531 but without the bug that it caused (#3622) which is why it was reverted. I'm sorry if I explain this badly. If you don't understand a part, make sure to let me know and I will explain it better. ### Explanation How it worked before: `Terminal` signals that viewport changed -> `TermControl::_TerminalScrollPositionChanged` gets called on the terminal thread -> it dispatches work for later to be ran the UI thread to updates the scrollbar's values Why it's bad: * If we have many viewport changes, it will create a long stack of operations to run. Instead, we should just update the scroll bar with the most recent information that we know. * Imagine if the rate that the work gets pushed on the UI thread is greater than the rate that it can handle: it might freeze? * No need to be real time, we can wait just a little bit (8ms) to accumulate viewport changes before we actually change the scroll bar's value because it appears to be expensive (see perf below). Now: `Terminal` signals that viewport changed -> `TermControl::_TerminalScrollPositionChanged` gets called on the terminal thread -> it tells the `ScrollBarUpdater` about a new update -> the `ScrollBarUpdater` only runs one job (I don't know if that's the right term) on the UI thread at a time. If a job is already running but hasn't updated the scroll bar yet, it changes the setting in the already existing job to update the scroll bar with the new values. A job "waits" some time before doing the update to throttle updates because we don't need real time scroll bar updates. -> eventually, it updates the scroll bar If the user scrolls when a scroll bar update is pending, we keep the scroll bar's Maximum and Minimum but let the user choose its new Value with the `CancelPendingValueChange` method. ### Note Also I changed a little bit the code from the Terminal to notify the TermControl less often when possible. I tried to scroll with the scroll bar, with the mouse wheel. I tried to scroll while content is being outputted. I tried to reproduce the crash from #2248 without success (good). Co-authored-by: Leonard Hecker <leonard@hecker.io> Closes #3622
2020-06-12 21:51:37 +02:00
{
const auto oldScrollOffset = _scrollOffset;
// scroll if...
// - no selection is active
// - viewport is already at the bottom
const bool scrollToOutput = !IsSelectionActive() && _scrollOffset == 0;
_scrollOffset = scrollToOutput ? 0 : _scrollOffset + scrollAmount + newRows;
// Clamp the range to make sure that we don't scroll way off the top of the buffer
_scrollOffset = std::clamp(_scrollOffset,
0,
_buffer->GetSize().Height() - _mutableViewport.Height());
// If the new scroll offset is different, then we'll still want to raise a scroll event
updatedViewport = updatedViewport || (oldScrollOffset != _scrollOffset);
}
// If the viewport moved, then send a scrolling notification.
if (updatedViewport)
{
Throttle scrollbar updates in TermControl to ~one per 8ms (#4608) In addition to the below (original) description, this commit introduces a ThrottledFunc template that can throttle _any_ function. It applies that type to muffle updates to the scrollbar. --- Redo #3531 but without the bug that it caused (#3622) which is why it was reverted. I'm sorry if I explain this badly. If you don't understand a part, make sure to let me know and I will explain it better. ### Explanation How it worked before: `Terminal` signals that viewport changed -> `TermControl::_TerminalScrollPositionChanged` gets called on the terminal thread -> it dispatches work for later to be ran the UI thread to updates the scrollbar's values Why it's bad: * If we have many viewport changes, it will create a long stack of operations to run. Instead, we should just update the scroll bar with the most recent information that we know. * Imagine if the rate that the work gets pushed on the UI thread is greater than the rate that it can handle: it might freeze? * No need to be real time, we can wait just a little bit (8ms) to accumulate viewport changes before we actually change the scroll bar's value because it appears to be expensive (see perf below). Now: `Terminal` signals that viewport changed -> `TermControl::_TerminalScrollPositionChanged` gets called on the terminal thread -> it tells the `ScrollBarUpdater` about a new update -> the `ScrollBarUpdater` only runs one job (I don't know if that's the right term) on the UI thread at a time. If a job is already running but hasn't updated the scroll bar yet, it changes the setting in the already existing job to update the scroll bar with the new values. A job "waits" some time before doing the update to throttle updates because we don't need real time scroll bar updates. -> eventually, it updates the scroll bar If the user scrolls when a scroll bar update is pending, we keep the scroll bar's Maximum and Minimum but let the user choose its new Value with the `CancelPendingValueChange` method. ### Note Also I changed a little bit the code from the Terminal to notify the TermControl less often when possible. I tried to scroll with the scroll bar, with the mouse wheel. I tried to scroll while content is being outputted. I tried to reproduce the crash from #2248 without success (good). Co-authored-by: Leonard Hecker <leonard@hecker.io> Closes #3622
2020-06-12 21:51:37 +02:00
_NotifyScrollEvent();
}
if (rowsPushedOffTopOfBuffer != 0)
Remove unneeded VT-specific control character handling (#4289) ## Summary of the Pull Request This PR removes all of the VT-specific functionality from the `WriteCharsLegacy` function that dealt with control characters, since those controls are now handled in the state machine when in VT mode. It also removes most of the control character handling from the `Terminal::_WriteBuffer` method for the same reason. ## References This is a followup to PR #4171 ## PR Checklist * [x] Closes #3971 * [x] 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 * [x] 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: https://github.com/microsoft/terminal/issues/780#issuecomment-570287435 ## Detailed Description of the Pull Request / Additional comments There are four changes to the `WriteCharsLegacy` implementation: 1. The `TAB` character had special case handling in VT mode which is now no longer required. This fixes a bug in the Python REPL editor (when run from a cmd shell in Windows Terminal), which would prevent you tabbing past the end of the line. It also fixes #3971. 2. Following on from point 1, the `WC_NONDESTRUCTIVE_TAB` flag could also now be removed. It only ever applied in VT mode, in which case the `TAB` character isn't handled in `WriteCharsLegacy`, so there isn't a need for a non-destructive version. 3. There used to be special case handling for a `BS` character at the beginning of the line when in VT mode, and that is also no longer required. This fixes an edge-case bug which would prevent a glyph being output for code point 8 when `ENABLE_PROCESSED_OUTPUT` was disabled. 4. There was quite a lot of special case handling for control characters in the "end-of-line wrap" implementation, which is no longer required. This fixes a bug which would prevent "low ASCII" characters from wrapping when output at the end of a line. Then in the `Terminal::_WriteBuffer` implementation, I've simply removed all control character handling, except for `LF`. The Terminal is always in VT mode, so the control characters are always handled by the state machine. The exception for the `LF` character is simply because it doesn't have a proper implementation yet, so it still passes the character through to `_WriteBuffer`. That will get cleaned up eventually, but I thought that could wait for a later PR. Finally, with the removal of the VT mode handling in `WriteCharsLegacy`, there was no longer a need for the `SCREEN_INFORMATION::InVTMode` method to be publicly accessible. That has now been made private. ## Validation Steps Performed I've only tested manually, making sure the conhost and Windows Terminal still basically work, and confirming that the above-mentioned bugs are fixed by these changes.
2020-01-29 20:18:46 +01:00
{
Render row-by-row instead of invalidating entire screen (#5185) ## Summary of the Pull Request Adjusts DirectX renderer to use `til::bitmap` to track invalidation regions. Uses special modification to invalidate a row-at-a-time to ensure ligatures and NxM glyphs continue to work. ## References Likely helps #1064 ## PR Checklist * [x] Closes #778 * [x] I work here. * [x] Manual testing performed. See Performance traces in #778. * [x] Automated tests for `til` changes. * [x] Am core contributor. And discussed with @DHowett-MSFT. ## Detailed Description of the Pull Request / Additional comments - Applies `til::bitmap` as the new invalidation scheme inside the DirectX renderer and updates all entrypoints for collecting invalidation data to coalesce into this structure. - Semi-permanently routes all invalidations through a helper method `_InvalidateRectangle` that will expand any invalidation to cover the entire line. This ensures that ligatures and NxM glyphs will continue to render appropriately while still allowing us to dramatically reduce the number of lines drawn overall. In the future, we may come up with a tighter solution than line-by-line invalidation and can modify this helper method appropriately at that later date to further scope the invalid region. - Ensures that the `experimental.retroTerminalEffects` feature continues to invalidate the entire display on start of frame as the shader is applied at the end of the frame composition and will stack on itself in an amusing fashion when we only redraw part of the display. - Moves many member variables inside the DirectX renderer into the new `til::size`, `til::point`, and `til::rectangle` methods to facilitate easier management and mathematical operations. Consequently adds `try/catch` blocks around many of the already-existing `noexcept` methods to deal with mathematical or casting failures now detected by using the support classes. - Corrects `TerminalCore` redraw triggers to appropriately communicate scrolling circumstances to the renderer so it can optimize the draw regions appropriately. - Fixes an issue in the base `Renderer` that was causing overlapping scroll regions due to behavior of `Viewport::TrimToViewport` modifying the local. This fix is "good enough" for now and should go away when `Viewport` is fully migrated to `til::rectangle`. - Adds multiplication and division operators to `til::rectangle` and supporting tests. These operates will help scale back and forth between a cell-based rectangle and a pixel-based rectangle. They take special care to ensure that a pixel rectangle being divided downward back to cells will expand (with the ceiling division methods) to cover a full cell when even one pixel inside the cell is touched (as is how a redraw would have to occur). - Blocks off trace logging of invalid regions if no one is listening to optimize performance. - Restores full usage of `IDXGISwapChain1::Present1` to accurately and fully communicate dirty and scroll regions to the underlying DirectX framework. This additional information allows the framework to optimize drawing between frames by eliminating data transfer of regions that aren't modified and shuffling frames in place. See [Remarks](https://docs.microsoft.com/en-us/windows/win32/api/dxgi1_2/nf-dxgi1_2-idxgiswapchain1-present1#remarks) for more details. - Updates `til::bitmap` set methods to use more optimized versions of the setters on the `dynamic_bitset<>` that can bulk fill bits as the existing algorithm was noticeably slow after applying the "expand-to-row" helper to the DirectX renderer invalidation. - All `til` import hierarchy is now handled in the parent `til.h` file and not in the child files to prevent circular imports from happening. We don't expect the import of any individual library file, only the base one. So this should be OK for now. ## Validation Steps Performed - Ran `cmatrix`, `cmatrix -u0`, and `cacafire` after changes were made. - Made a bunch of ligatures with `Cascadia Code` in the Terminal before/after the changes and confirmed they still ligate. - Ran `dir` in Powershell and fixed the scrolling issues - Clicked all over the place and dragged to make sure selection works. - Checked retro terminal effect manually with Powershell.
2020-04-13 22:09:02 +02:00
// We have to report the delta here because we might have circled the text buffer.
// That didn't change the viewport and therefore the TriggerScroll(void)
// method can't detect the delta on its own.
Throttle scrollbar updates in TermControl to ~one per 8ms (#4608) In addition to the below (original) description, this commit introduces a ThrottledFunc template that can throttle _any_ function. It applies that type to muffle updates to the scrollbar. --- Redo #3531 but without the bug that it caused (#3622) which is why it was reverted. I'm sorry if I explain this badly. If you don't understand a part, make sure to let me know and I will explain it better. ### Explanation How it worked before: `Terminal` signals that viewport changed -> `TermControl::_TerminalScrollPositionChanged` gets called on the terminal thread -> it dispatches work for later to be ran the UI thread to updates the scrollbar's values Why it's bad: * If we have many viewport changes, it will create a long stack of operations to run. Instead, we should just update the scroll bar with the most recent information that we know. * Imagine if the rate that the work gets pushed on the UI thread is greater than the rate that it can handle: it might freeze? * No need to be real time, we can wait just a little bit (8ms) to accumulate viewport changes before we actually change the scroll bar's value because it appears to be expensive (see perf below). Now: `Terminal` signals that viewport changed -> `TermControl::_TerminalScrollPositionChanged` gets called on the terminal thread -> it tells the `ScrollBarUpdater` about a new update -> the `ScrollBarUpdater` only runs one job (I don't know if that's the right term) on the UI thread at a time. If a job is already running but hasn't updated the scroll bar yet, it changes the setting in the already existing job to update the scroll bar with the new values. A job "waits" some time before doing the update to throttle updates because we don't need real time scroll bar updates. -> eventually, it updates the scroll bar If the user scrolls when a scroll bar update is pending, we keep the scroll bar's Maximum and Minimum but let the user choose its new Value with the `CancelPendingValueChange` method. ### Note Also I changed a little bit the code from the Terminal to notify the TermControl less often when possible. I tried to scroll with the scroll bar, with the mouse wheel. I tried to scroll while content is being outputted. I tried to reproduce the crash from #2248 without success (good). Co-authored-by: Leonard Hecker <leonard@hecker.io> Closes #3622
2020-06-12 21:51:37 +02:00
COORD delta{ 0, -rowsPushedOffTopOfBuffer };
Render row-by-row instead of invalidating entire screen (#5185) ## Summary of the Pull Request Adjusts DirectX renderer to use `til::bitmap` to track invalidation regions. Uses special modification to invalidate a row-at-a-time to ensure ligatures and NxM glyphs continue to work. ## References Likely helps #1064 ## PR Checklist * [x] Closes #778 * [x] I work here. * [x] Manual testing performed. See Performance traces in #778. * [x] Automated tests for `til` changes. * [x] Am core contributor. And discussed with @DHowett-MSFT. ## Detailed Description of the Pull Request / Additional comments - Applies `til::bitmap` as the new invalidation scheme inside the DirectX renderer and updates all entrypoints for collecting invalidation data to coalesce into this structure. - Semi-permanently routes all invalidations through a helper method `_InvalidateRectangle` that will expand any invalidation to cover the entire line. This ensures that ligatures and NxM glyphs will continue to render appropriately while still allowing us to dramatically reduce the number of lines drawn overall. In the future, we may come up with a tighter solution than line-by-line invalidation and can modify this helper method appropriately at that later date to further scope the invalid region. - Ensures that the `experimental.retroTerminalEffects` feature continues to invalidate the entire display on start of frame as the shader is applied at the end of the frame composition and will stack on itself in an amusing fashion when we only redraw part of the display. - Moves many member variables inside the DirectX renderer into the new `til::size`, `til::point`, and `til::rectangle` methods to facilitate easier management and mathematical operations. Consequently adds `try/catch` blocks around many of the already-existing `noexcept` methods to deal with mathematical or casting failures now detected by using the support classes. - Corrects `TerminalCore` redraw triggers to appropriately communicate scrolling circumstances to the renderer so it can optimize the draw regions appropriately. - Fixes an issue in the base `Renderer` that was causing overlapping scroll regions due to behavior of `Viewport::TrimToViewport` modifying the local. This fix is "good enough" for now and should go away when `Viewport` is fully migrated to `til::rectangle`. - Adds multiplication and division operators to `til::rectangle` and supporting tests. These operates will help scale back and forth between a cell-based rectangle and a pixel-based rectangle. They take special care to ensure that a pixel rectangle being divided downward back to cells will expand (with the ceiling division methods) to cover a full cell when even one pixel inside the cell is touched (as is how a redraw would have to occur). - Blocks off trace logging of invalid regions if no one is listening to optimize performance. - Restores full usage of `IDXGISwapChain1::Present1` to accurately and fully communicate dirty and scroll regions to the underlying DirectX framework. This additional information allows the framework to optimize drawing between frames by eliminating data transfer of regions that aren't modified and shuffling frames in place. See [Remarks](https://docs.microsoft.com/en-us/windows/win32/api/dxgi1_2/nf-dxgi1_2-idxgiswapchain1-present1#remarks) for more details. - Updates `til::bitmap` set methods to use more optimized versions of the setters on the `dynamic_bitset<>` that can bulk fill bits as the existing algorithm was noticeably slow after applying the "expand-to-row" helper to the DirectX renderer invalidation. - All `til` import hierarchy is now handled in the parent `til.h` file and not in the child files to prevent circular imports from happening. We don't expect the import of any individual library file, only the base one. So this should be OK for now. ## Validation Steps Performed - Ran `cmatrix`, `cmatrix -u0`, and `cacafire` after changes were made. - Made a bunch of ligatures with `Cascadia Code` in the Terminal before/after the changes and confirmed they still ligate. - Ran `dir` in Powershell and fixed the scrolling issues - Clicked all over the place and dragged to make sure selection works. - Checked retro terminal effect manually with Powershell.
2020-04-13 22:09:02 +02:00
_buffer->GetRenderTarget().TriggerScroll(&delta);
Remove unneeded VT-specific control character handling (#4289) ## Summary of the Pull Request This PR removes all of the VT-specific functionality from the `WriteCharsLegacy` function that dealt with control characters, since those controls are now handled in the state machine when in VT mode. It also removes most of the control character handling from the `Terminal::_WriteBuffer` method for the same reason. ## References This is a followup to PR #4171 ## PR Checklist * [x] Closes #3971 * [x] 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 * [x] 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: https://github.com/microsoft/terminal/issues/780#issuecomment-570287435 ## Detailed Description of the Pull Request / Additional comments There are four changes to the `WriteCharsLegacy` implementation: 1. The `TAB` character had special case handling in VT mode which is now no longer required. This fixes a bug in the Python REPL editor (when run from a cmd shell in Windows Terminal), which would prevent you tabbing past the end of the line. It also fixes #3971. 2. Following on from point 1, the `WC_NONDESTRUCTIVE_TAB` flag could also now be removed. It only ever applied in VT mode, in which case the `TAB` character isn't handled in `WriteCharsLegacy`, so there isn't a need for a non-destructive version. 3. There used to be special case handling for a `BS` character at the beginning of the line when in VT mode, and that is also no longer required. This fixes an edge-case bug which would prevent a glyph being output for code point 8 when `ENABLE_PROCESSED_OUTPUT` was disabled. 4. There was quite a lot of special case handling for control characters in the "end-of-line wrap" implementation, which is no longer required. This fixes a bug which would prevent "low ASCII" characters from wrapping when output at the end of a line. Then in the `Terminal::_WriteBuffer` implementation, I've simply removed all control character handling, except for `LF`. The Terminal is always in VT mode, so the control characters are always handled by the state machine. The exception for the `LF` character is simply because it doesn't have a proper implementation yet, so it still passes the character through to `_WriteBuffer`. That will get cleaned up eventually, but I thought that could wait for a later PR. Finally, with the removal of the VT mode handling in `WriteCharsLegacy`, there was no longer a need for the `SCREEN_INFORMATION::InVTMode` method to be publicly accessible. That has now been made private. ## Validation Steps Performed I've only tested manually, making sure the conhost and Windows Terminal still basically work, and confirming that the above-mentioned bugs are fixed by these changes.
2020-01-29 20:18:46 +01:00
}
// Firing the CursorPositionChanged event is very expensive so we try not to do that when
// the cursor does not need to be redrawn.
if (!cursor.IsDeferDrawing())
{
_NotifyTerminalCursorPositionChanged();
}
}
void Terminal::UserScrollViewport(const int viewTop)
{
Render row-by-row instead of invalidating entire screen (#5185) ## Summary of the Pull Request Adjusts DirectX renderer to use `til::bitmap` to track invalidation regions. Uses special modification to invalidate a row-at-a-time to ensure ligatures and NxM glyphs continue to work. ## References Likely helps #1064 ## PR Checklist * [x] Closes #778 * [x] I work here. * [x] Manual testing performed. See Performance traces in #778. * [x] Automated tests for `til` changes. * [x] Am core contributor. And discussed with @DHowett-MSFT. ## Detailed Description of the Pull Request / Additional comments - Applies `til::bitmap` as the new invalidation scheme inside the DirectX renderer and updates all entrypoints for collecting invalidation data to coalesce into this structure. - Semi-permanently routes all invalidations through a helper method `_InvalidateRectangle` that will expand any invalidation to cover the entire line. This ensures that ligatures and NxM glyphs will continue to render appropriately while still allowing us to dramatically reduce the number of lines drawn overall. In the future, we may come up with a tighter solution than line-by-line invalidation and can modify this helper method appropriately at that later date to further scope the invalid region. - Ensures that the `experimental.retroTerminalEffects` feature continues to invalidate the entire display on start of frame as the shader is applied at the end of the frame composition and will stack on itself in an amusing fashion when we only redraw part of the display. - Moves many member variables inside the DirectX renderer into the new `til::size`, `til::point`, and `til::rectangle` methods to facilitate easier management and mathematical operations. Consequently adds `try/catch` blocks around many of the already-existing `noexcept` methods to deal with mathematical or casting failures now detected by using the support classes. - Corrects `TerminalCore` redraw triggers to appropriately communicate scrolling circumstances to the renderer so it can optimize the draw regions appropriately. - Fixes an issue in the base `Renderer` that was causing overlapping scroll regions due to behavior of `Viewport::TrimToViewport` modifying the local. This fix is "good enough" for now and should go away when `Viewport` is fully migrated to `til::rectangle`. - Adds multiplication and division operators to `til::rectangle` and supporting tests. These operates will help scale back and forth between a cell-based rectangle and a pixel-based rectangle. They take special care to ensure that a pixel rectangle being divided downward back to cells will expand (with the ceiling division methods) to cover a full cell when even one pixel inside the cell is touched (as is how a redraw would have to occur). - Blocks off trace logging of invalid regions if no one is listening to optimize performance. - Restores full usage of `IDXGISwapChain1::Present1` to accurately and fully communicate dirty and scroll regions to the underlying DirectX framework. This additional information allows the framework to optimize drawing between frames by eliminating data transfer of regions that aren't modified and shuffling frames in place. See [Remarks](https://docs.microsoft.com/en-us/windows/win32/api/dxgi1_2/nf-dxgi1_2-idxgiswapchain1-present1#remarks) for more details. - Updates `til::bitmap` set methods to use more optimized versions of the setters on the `dynamic_bitset<>` that can bulk fill bits as the existing algorithm was noticeably slow after applying the "expand-to-row" helper to the DirectX renderer invalidation. - All `til` import hierarchy is now handled in the parent `til.h` file and not in the child files to prevent circular imports from happening. We don't expect the import of any individual library file, only the base one. So this should be OK for now. ## Validation Steps Performed - Ran `cmatrix`, `cmatrix -u0`, and `cacafire` after changes were made. - Made a bunch of ligatures with `Cascadia Code` in the Terminal before/after the changes and confirmed they still ligate. - Ran `dir` in Powershell and fixed the scrolling issues - Clicked all over the place and dragged to make sure selection works. - Checked retro terminal effect manually with Powershell.
2020-04-13 22:09:02 +02:00
// we're going to modify state here that the renderer could be reading.
auto lock = LockForWriting();
const auto clampedNewTop = std::max(0, viewTop);
Search - add search box control and implement search experience (#3590) <!-- 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)? --> This is the PR for feature Search: #605 This PR includes the newly introduced SearchBoxControl in TermControl dir, which is the search bar for the search experience. And the codes that enable Search in Windows Terminal. <!-- Other than the issue solved, is this relevant to any other issues/existing PRs? --> The PR that migrates the Conhost search module: https://github.com/microsoft/terminal/pull/3279 Spec (still actively updating): https://github.com/microsoft/terminal/pull/3299 <!-- Please review the items on the PR checklist before submitting--> ## PR Checklist * [x] Closes #605 * [ ] 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 --> These functionalities are included in the search experience. 1. Search in Terminal text buffer. 2. Automatic wrap-around. 3. Search up or down switch by clicking different buttons. 4. Search case sensitively/insensitively by clicking a button. S. Move the search box to the top/bottom by clicking a button. 6. Close by clicking 'X'. 7. Open search by ctrl + F. When the searchbox is open, the user could still interact with the terminal by clicking the terminal input area. While I already have the search functionalities, currently there are still some known to-do works and I will keep updating my PR: 1. Optimize the search box UI, this includes: 1) Theme adaptation. The search box background and font color should change according to the theme, 2) Add background. Currently the elements in search box are all transparent. However, we need a background. 3) Move button should be highlighted once clicked. 2. Accessibility: search process should be able to performed without mouse. Once the search box is focused, the user should be able to navigate between all interactive elements on the searchbox using keyboard. <!-- Describe how you validated the behavior. Add automated tests wherever possible, but list manual validation steps taken as well --> To test: 1. checkout this branch. 2. Build the project. 3. Start Windows Terminal and press Ctrl+F 4. The search box should appear on the top right corner.
2019-12-17 16:52:37 +01:00
const auto realTop = ViewStartIndex();
const auto newDelta = realTop - clampedNewTop;
// if viewTop > realTop, we want the offset to be 0.
_scrollOffset = std::max(0, newDelta);
Render row-by-row instead of invalidating entire screen (#5185) ## Summary of the Pull Request Adjusts DirectX renderer to use `til::bitmap` to track invalidation regions. Uses special modification to invalidate a row-at-a-time to ensure ligatures and NxM glyphs continue to work. ## References Likely helps #1064 ## PR Checklist * [x] Closes #778 * [x] I work here. * [x] Manual testing performed. See Performance traces in #778. * [x] Automated tests for `til` changes. * [x] Am core contributor. And discussed with @DHowett-MSFT. ## Detailed Description of the Pull Request / Additional comments - Applies `til::bitmap` as the new invalidation scheme inside the DirectX renderer and updates all entrypoints for collecting invalidation data to coalesce into this structure. - Semi-permanently routes all invalidations through a helper method `_InvalidateRectangle` that will expand any invalidation to cover the entire line. This ensures that ligatures and NxM glyphs will continue to render appropriately while still allowing us to dramatically reduce the number of lines drawn overall. In the future, we may come up with a tighter solution than line-by-line invalidation and can modify this helper method appropriately at that later date to further scope the invalid region. - Ensures that the `experimental.retroTerminalEffects` feature continues to invalidate the entire display on start of frame as the shader is applied at the end of the frame composition and will stack on itself in an amusing fashion when we only redraw part of the display. - Moves many member variables inside the DirectX renderer into the new `til::size`, `til::point`, and `til::rectangle` methods to facilitate easier management and mathematical operations. Consequently adds `try/catch` blocks around many of the already-existing `noexcept` methods to deal with mathematical or casting failures now detected by using the support classes. - Corrects `TerminalCore` redraw triggers to appropriately communicate scrolling circumstances to the renderer so it can optimize the draw regions appropriately. - Fixes an issue in the base `Renderer` that was causing overlapping scroll regions due to behavior of `Viewport::TrimToViewport` modifying the local. This fix is "good enough" for now and should go away when `Viewport` is fully migrated to `til::rectangle`. - Adds multiplication and division operators to `til::rectangle` and supporting tests. These operates will help scale back and forth between a cell-based rectangle and a pixel-based rectangle. They take special care to ensure that a pixel rectangle being divided downward back to cells will expand (with the ceiling division methods) to cover a full cell when even one pixel inside the cell is touched (as is how a redraw would have to occur). - Blocks off trace logging of invalid regions if no one is listening to optimize performance. - Restores full usage of `IDXGISwapChain1::Present1` to accurately and fully communicate dirty and scroll regions to the underlying DirectX framework. This additional information allows the framework to optimize drawing between frames by eliminating data transfer of regions that aren't modified and shuffling frames in place. See [Remarks](https://docs.microsoft.com/en-us/windows/win32/api/dxgi1_2/nf-dxgi1_2-idxgiswapchain1-present1#remarks) for more details. - Updates `til::bitmap` set methods to use more optimized versions of the setters on the `dynamic_bitset<>` that can bulk fill bits as the existing algorithm was noticeably slow after applying the "expand-to-row" helper to the DirectX renderer invalidation. - All `til` import hierarchy is now handled in the parent `til.h` file and not in the child files to prevent circular imports from happening. We don't expect the import of any individual library file, only the base one. So this should be OK for now. ## Validation Steps Performed - Ran `cmatrix`, `cmatrix -u0`, and `cacafire` after changes were made. - Made a bunch of ligatures with `Cascadia Code` in the Terminal before/after the changes and confirmed they still ligate. - Ran `dir` in Powershell and fixed the scrolling issues - Clicked all over the place and dragged to make sure selection works. - Checked retro terminal effect manually with Powershell.
2020-04-13 22:09:02 +02:00
// We can use the void variant of TriggerScroll here because
// we adjusted the viewport so it can detect the difference
// from the previous frame drawn.
_buffer->GetRenderTarget().TriggerScroll();
}
int Terminal::GetScrollOffset() noexcept
{
return _VisibleStartIndex();
}
void Terminal::_NotifyScrollEvent() noexcept
try
{
if (_pfnScrollPositionChanged)
{
const auto visible = _GetVisibleViewport();
const auto top = visible.Top();
const auto height = visible.Height();
const auto bottom = this->GetBufferHeight();
_pfnScrollPositionChanged(top, height, bottom);
}
}
CATCH_LOG()
void Terminal::_NotifyTerminalCursorPositionChanged() noexcept
{
if (_pfnCursorPositionChanged)
{
try
{
_pfnCursorPositionChanged();
}
CATCH_LOG();
}
}
void Terminal::SetWriteInputCallback(std::function<void(std::wstring&)> pfn) noexcept
{
_pfnWriteInput.swap(pfn);
}
void Terminal::SetWarningBellCallback(std::function<void()> pfn) noexcept
{
_pfnWarningBell.swap(pfn);
}
Split `TermControl` into a Core, Interactivity, and Control layer (#9820) ## Summary of the Pull Request Brace yourselves, it's finally here. This PR does the dirty work of splitting the monolithic `TermControl` into three components. These components are: * `ControlCore`: This encapsulates the `Terminal` instance, the `DxEngine` and `Renderer`, and the `Connection`. This is intended to everything that someone might need to stand up a terminal instance in a control, but without any regard for how the UX works. * `ControlInteractivity`: This is a wrapper for the `ControlCore`, which holds the logic for things like double-click, right click copy/paste, selection, etc. This is intended to be a UI framework-independent abstraction. The methods this layer exposes can be called the same from both the WinUI TermControl and the WPF control. * `TermControl`: This is the UWP control. It's got a Core and Interactivity inside it, which it uses for the actual logic of the terminal itself. TermControl's main responsibility is now By splitting into smaller pieces, it will enable us to * write unit tests for the `Core` and `Interactivity` bits, which we desparately need * Combine `ControlCore` and `ControlInteractivity` in an out-of-proc core process in the future, to enable tab tearout. However, we're not doing that work quite yet. There's still lots of work to be done to enable that, thought this is likely the biggest portion. Ideally, this would just be methods moved wholesale from one file to another. Unfortunately, there are a bunch of cases where that didn't work as well as expected. Especially when trying to better enforce the boundary between the classes. We've got a couple tests here that I've added. These are partially examples, and partially things I ran into while implementing this. A bunch of things from #7001 can go in now that we have this. This PR is gonna be a huge pain to review - 38 files with 3,730 additions and 1,661 deletions is nothing to scoff at. It will also conflict 100% with anything that's targeting `TermControl`. I'm hoping we can review this over the course of the next week and just be done with it, and leave plenty of runway for 1.9 bugs in post. ## References * In pursuit of #1256 * Proc Model: #5000 * https://github.com/microsoft/terminal/projects/5 ## PR Checklist * [x] Closes #6842 * [x] Closes https://github.com/microsoft/terminal/projects/5#card-50760249 * [x] Closes https://github.com/microsoft/terminal/projects/5#card-50760258 * [x] I work here * [x] Tests added/passed * [n/a] Requires documentation to be updated ## Detailed Description of the Pull Request / Additional comments * I don't love the names `ControlCore` and `ControlInteractivity`. Open to other names. * I added a `ICoreState` interface for "properties that come from the `ControlCore`, but consumers of the `TermControl` need to know". In the future, these will all need to be handled specially, because they might involve an RPC call to retrieve the info from the core (or cache it) in the window process. * I've added more `EventArgs` to make more events proper `TypedEvent`s. * I've changed how the TerminalApp layer requests updated TaskbarProgress state. It doesn't need to pump TermControl to raise a new event anymore. * ~~Something that snuck into this branch in the very long history is the switch to `DCompositionCreateSurfaceHandle` for the `DxEngine`. @miniksa wrote this originally in 30b8335, I'm just finally committing it here. We'll need that in the future for the out-of-proc stuff.~~ * I reverted this in c113b65d9. We can revert _that_ commit when we want to come back to it. * I've changed the acrylic handler a decent amount. But added tests! * All the `ThrottledFunc` things are left in `TermControl`. Some might be able to move down into core/interactivity, but once we figure out how to use a different kind of Dispatcher (because a UI thread won't necessarily exist for those components). * I've undoubtably messed up the merging of the locking around the appearance config stuff recently ## Validation Steps Performed I've got a rolling list in https://github.com/microsoft/terminal/issues/6842#issuecomment-810990460 that I'm updating as I go.
2021-04-27 17:50:45 +02:00
void Terminal::SetTitleChangedCallback(std::function<void(std::wstring_view)> pfn) noexcept
{
_pfnTitleChanged.swap(pfn);
}
void Terminal::SetTabColorChangedCallback(std::function<void(const std::optional<til::color>)> pfn) noexcept
{
_pfnTabColorChanged.swap(pfn);
}
Split `TermControl` into a Core, Interactivity, and Control layer (#9820) ## Summary of the Pull Request Brace yourselves, it's finally here. This PR does the dirty work of splitting the monolithic `TermControl` into three components. These components are: * `ControlCore`: This encapsulates the `Terminal` instance, the `DxEngine` and `Renderer`, and the `Connection`. This is intended to everything that someone might need to stand up a terminal instance in a control, but without any regard for how the UX works. * `ControlInteractivity`: This is a wrapper for the `ControlCore`, which holds the logic for things like double-click, right click copy/paste, selection, etc. This is intended to be a UI framework-independent abstraction. The methods this layer exposes can be called the same from both the WinUI TermControl and the WPF control. * `TermControl`: This is the UWP control. It's got a Core and Interactivity inside it, which it uses for the actual logic of the terminal itself. TermControl's main responsibility is now By splitting into smaller pieces, it will enable us to * write unit tests for the `Core` and `Interactivity` bits, which we desparately need * Combine `ControlCore` and `ControlInteractivity` in an out-of-proc core process in the future, to enable tab tearout. However, we're not doing that work quite yet. There's still lots of work to be done to enable that, thought this is likely the biggest portion. Ideally, this would just be methods moved wholesale from one file to another. Unfortunately, there are a bunch of cases where that didn't work as well as expected. Especially when trying to better enforce the boundary between the classes. We've got a couple tests here that I've added. These are partially examples, and partially things I ran into while implementing this. A bunch of things from #7001 can go in now that we have this. This PR is gonna be a huge pain to review - 38 files with 3,730 additions and 1,661 deletions is nothing to scoff at. It will also conflict 100% with anything that's targeting `TermControl`. I'm hoping we can review this over the course of the next week and just be done with it, and leave plenty of runway for 1.9 bugs in post. ## References * In pursuit of #1256 * Proc Model: #5000 * https://github.com/microsoft/terminal/projects/5 ## PR Checklist * [x] Closes #6842 * [x] Closes https://github.com/microsoft/terminal/projects/5#card-50760249 * [x] Closes https://github.com/microsoft/terminal/projects/5#card-50760258 * [x] I work here * [x] Tests added/passed * [n/a] Requires documentation to be updated ## Detailed Description of the Pull Request / Additional comments * I don't love the names `ControlCore` and `ControlInteractivity`. Open to other names. * I added a `ICoreState` interface for "properties that come from the `ControlCore`, but consumers of the `TermControl` need to know". In the future, these will all need to be handled specially, because they might involve an RPC call to retrieve the info from the core (or cache it) in the window process. * I've added more `EventArgs` to make more events proper `TypedEvent`s. * I've changed how the TerminalApp layer requests updated TaskbarProgress state. It doesn't need to pump TermControl to raise a new event anymore. * ~~Something that snuck into this branch in the very long history is the switch to `DCompositionCreateSurfaceHandle` for the `DxEngine`. @miniksa wrote this originally in 30b8335, I'm just finally committing it here. We'll need that in the future for the out-of-proc stuff.~~ * I reverted this in c113b65d9. We can revert _that_ commit when we want to come back to it. * I've changed the acrylic handler a decent amount. But added tests! * All the `ThrottledFunc` things are left in `TermControl`. Some might be able to move down into core/interactivity, but once we figure out how to use a different kind of Dispatcher (because a UI thread won't necessarily exist for those components). * I've undoubtably messed up the merging of the locking around the appearance config stuff recently ## Validation Steps Performed I've got a rolling list in https://github.com/microsoft/terminal/issues/6842#issuecomment-810990460 that I'm updating as I go.
2021-04-27 17:50:45 +02:00
void Terminal::SetCopyToClipboardCallback(std::function<void(std::wstring_view)> pfn) noexcept
{
_pfnCopyToClipboard.swap(pfn);
}
void Terminal::SetScrollPositionChangedCallback(std::function<void(const int, const int, const int)> pfn) noexcept
{
_pfnScrollPositionChanged.swap(pfn);
}
void Terminal::SetCursorPositionChangedCallback(std::function<void()> pfn) noexcept
{
_pfnCursorPositionChanged.swap(pfn);
}
// Method Description:
// - Allows setting a callback for when the background color is changed
// Arguments:
// - pfn: a function callback that takes a color
void Terminal::SetBackgroundCallback(std::function<void(const til::color)> pfn) noexcept
{
_pfnBackgroundColorChanged.swap(pfn);
}
// Method Description:
// - Allows settings a callback for settings the taskbar progress indicator
// Arguments:
// - pfn: a function callback that takes 2 size_t parameters, one indicating the progress state
// and the other indicating the progress value
void Microsoft::Terminal::Core::Terminal::TaskbarProgressChangedCallback(std::function<void()> pfn) noexcept
{
_pfnTaskbarProgressChanged.swap(pfn);
}
void Terminal::_InitializeColorTable()
try
{
const gsl::span<COLORREF> tableView = { _colorTable.data(), _colorTable.size() };
// First set up the basic 256 colors
Standardize the color table order (#11602) ## Summary of the Pull Request In the original implementation, we used two different orderings for the color tables. The WT color table used ANSI order, while the conhost color table used a Windows-specific order. This PR standardizes on the ANSI color order everywhere, so the usage of indexed colors is consistent across both parts of the code base, which will hopefully allow more of the code to be shared one day. ## References This is another small step towards de-duplicating `AdaptDispatch` and `TerminalDispatch` for issue #3849, and is essentially a followup to the SGR dispatch refactoring in PR #6728. ## PR Checklist * [x] Closes #11461 * [x] CLA signed. * [x] Tests added/passed * [ ] Documentation updated. * [ ] Schema updated. * [x] I've discussed this with core contributors already. Issue number where discussion took place: #11461 ## Detailed Description of the Pull Request / Additional comments Conhost still needs to deal with legacy attributes using Windows color order, so those values now need to be transposed to ANSI colors order when creating a `TextAttribute` object. This is done with a simple mapping table, which also handles the translation of the default color entries, so it's actually slightly faster than the original code. And when converting `TextAttribute` values back to legacy console attributes, we were already using a mapping table to handle the narrowing of 256-color values down to 16 colors, so we just needed to adjust that table to account for the translation from ANSI to Windows, and then could make use of the same table for both 256-color and 16-color values. There are also a few places in conhost that read from or write to the color tables, and those now need to transpose the index values. I've addressed this by creating separate `SetLegacyColorTableEntry` and `GetLegacyColorTableEntry` methods in the `Settings` class which take care of the mapping, so it's now clearer in which cases the code is dealing with legacy values, and which are ANSI values. These methods are used in the `SetConsoleScreenBufferInfoEx` and `GetConsoleScreenBufferInfoEx` APIs, as well as a few place where color preferences are handled (the registry, shortcut links, and the properties dialog), none of which are particularly sensitive to performance. However, we also use the legacy table when looking up the default colors for rendering (which happens a lot), so I've refactored that code so the default color calculations now only occur once per frame. The plus side of all of this is that the VT code doesn't need to do the index translation anymore, so we can finally get rid of all the calls to `XTermToWindowsIndex`, and we no longer need a separate color table initialization method for conhost, so I was able to merge a number of color initialization methods into one. We also no longer need to translate from legacy values to ANSI when generating VT sequences for conpty. The one exception to that is the 16-color VT renderer, which uses the `TextColor::GetLegacyIndex` method to approximate 16-color equivalents for RGB and 256-color values. Since that method returns a legacy index, it still needs to be translated to ANSI before it can be used in a VT sequence. But this should be no worse than it was before. One more special case is conhost's secret _Color Selection_ feature. That uses `Ctrl`+Number and `Alt`+Number key sequences to highlight parts of the buffer, and the mapping from number to color is based on the Windows color order. So that mapping now needs to be transposed, but that's also not performance sensitive. The only thing that I haven't bothered to update is the trace logging code in the `Telemetry` class, which logs the first 16 entries in the color table. Those entries are now going to be in a different order, but I didn't think that would be of great concern to anyone. ## Validation Steps Performed A lot of unit tests needed to be updated to use ANSI color constants when setting indexed colors, where before they might have been expecting values in Windows order. But this replaced a wild mix of different constants, sometimes having to use bit shifting, as well as values mapped with `XTermToWindowsIndex`, so I think the tests are a whole lot clearer now. Only a few cases have been left with literal numbers where that seemed more appropriate. In addition to getting the unit tests working, I've also manually tested the behaviour of all the console APIs which I thought could be affected by these changes, and confirmed that they produced the same results in the new code as they did in the original implementation. This includes: - `WriteConsoleOutput` - `ReadConsoleOutput` - `SetConsoleTextAttribute` with `WriteConsoleOutputCharacter` - `FillConsoleOutputAttribute` and `FillConsoleOutputCharacter` - `ScrollConsoleScreenBuffer` - `GetConsoleScreenBufferInfo` - `GetConsoleScreenBufferInfoEx` - `SetConsoleScreenBufferInfoEx` I've also manually tested changing colors via the console properties menu, the registry, and shortcut links, including setting default colors and popup colors. And I've tested that the "Quirks Mode" is still working as expected in PowerShell. In terms of performance, I wrote a little test app that filled a 80x9999 buffer with random color combinations using `WriteConsoleOutput`, which I figured was likely to be the most performance sensitive call, and I think it now actually performs slightly better than the original implementation. I've also tested similar code - just filling the visible window - with SGR VT sequences of various types, and the performance seems about the same as it was before.
2021-11-04 23:13:22 +01:00
Utils::InitializeColorTable(tableView);
// Then make sure all the values have an alpha of 255
Utils::SetColorTableAlpha(tableView, 0xff);
}
CATCH_LOG()
// Method Description:
// - Sets the cursor to be currently on. On/Off is tracked independently of
// cursor visibility (hidden/visible). On/off is controlled by the cursor
// blinker. Visibility is usually controlled by the client application. If the
// cursor is hidden, then the cursor will remain hidden. If the cursor is
// Visible, then it will immediately become visible.
// Arguments:
// - isVisible: whether the cursor should be visible
Show a double width cursor for double width characters (#5319) # Summary of the Pull Request This PR will allow the cursor to be double width when on top of a double width character. This required changing `IsCursorDoubleWidth` to check whether the glyph the cursor's on top of is double width. This code is exactly the same as the original PR that addressed this issue in #2932. That one got reverted at some point due to the crashes related to it, but due to a combination of Terminal having come further since that PR and other changes to address use-after-frees, some of the crashes may/may not be relevant now. The ones that seemed to be relevant/repro-able, I attempt to address in this PR. The `IsCursorDoubleWidth` check would fail during the `TextBuffer::Reflow` call inside of `Terminal::UserResize` occasionally, particularly when `newCursor.EndDeferDrawing()` is called. This is because when we tell the newCursor to `EndDefer`, the renderer will attempt to redraw the cursor. As part of this redraw, it'll ask if `IsCursorDoubleWidth`, and if the renderer managed to ask this before `UserResize` swapped out the old buffer with the new one from `Reflow`, the renderer will be asking the old buffer if its out-of-bounds cursor is double width. This was pretty easily repro'd using `cmatrix -u0` and resizing the window like a madman. As a solution, I've moved the Start/End DeferDrawing calls out of `Reflow` and into `UserResize`. This way, I can "clamp" the portion of the code where the newBuffer is getting created and reflowed and swapped into the Terminal buffer, and only allow the renderer to draw once the swap is done. This also means that ConHost's `ResizeWithReflow` needed to change slightly. In addition, I've added a WriteLock to `SetCursorOn`. It was mentioned as a fix for a crash in #2965 (although I can't repro), and I also figured it would be good to try to emulate where ConHost locks with regards to Cursor operations, and this seemed to be one that we were missing. ## PR Checklist * [x] Closes #2713 * [x] CLA signed * [x] Tests added/passed ## Validation Steps Performed Manual validation that the cursor is indeed chonky, added a test case to check that we are correctly saying that the cursor is double width (not too sure if I put it in the right place). Also open to other test case ideas and thoughts on what else I should be careful for since I am quite nervous about what other crashes might occur.
2020-04-15 21:23:06 +02:00
void Terminal::SetCursorOn(const bool isOn)
{
Show a double width cursor for double width characters (#5319) # Summary of the Pull Request This PR will allow the cursor to be double width when on top of a double width character. This required changing `IsCursorDoubleWidth` to check whether the glyph the cursor's on top of is double width. This code is exactly the same as the original PR that addressed this issue in #2932. That one got reverted at some point due to the crashes related to it, but due to a combination of Terminal having come further since that PR and other changes to address use-after-frees, some of the crashes may/may not be relevant now. The ones that seemed to be relevant/repro-able, I attempt to address in this PR. The `IsCursorDoubleWidth` check would fail during the `TextBuffer::Reflow` call inside of `Terminal::UserResize` occasionally, particularly when `newCursor.EndDeferDrawing()` is called. This is because when we tell the newCursor to `EndDefer`, the renderer will attempt to redraw the cursor. As part of this redraw, it'll ask if `IsCursorDoubleWidth`, and if the renderer managed to ask this before `UserResize` swapped out the old buffer with the new one from `Reflow`, the renderer will be asking the old buffer if its out-of-bounds cursor is double width. This was pretty easily repro'd using `cmatrix -u0` and resizing the window like a madman. As a solution, I've moved the Start/End DeferDrawing calls out of `Reflow` and into `UserResize`. This way, I can "clamp" the portion of the code where the newBuffer is getting created and reflowed and swapped into the Terminal buffer, and only allow the renderer to draw once the swap is done. This also means that ConHost's `ResizeWithReflow` needed to change slightly. In addition, I've added a WriteLock to `SetCursorOn`. It was mentioned as a fix for a crash in #2965 (although I can't repro), and I also figured it would be good to try to emulate where ConHost locks with regards to Cursor operations, and this seemed to be one that we were missing. ## PR Checklist * [x] Closes #2713 * [x] CLA signed * [x] Tests added/passed ## Validation Steps Performed Manual validation that the cursor is indeed chonky, added a test case to check that we are correctly saying that the cursor is double width (not too sure if I put it in the right place). Also open to other test case ideas and thoughts on what else I should be careful for since I am quite nervous about what other crashes might occur.
2020-04-15 21:23:06 +02:00
auto lock = LockForWriting();
_buffer->GetCursor().SetIsOn(isOn);
}
bool Terminal::IsCursorBlinkingAllowed() const noexcept
{
const auto& cursor = _buffer->GetCursor();
return cursor.IsBlinkingAllowed();
}
// Method Description:
// - Update our internal knowledge about where regex patterns are on the screen
// - This is called by TerminalControl (through a throttled function) when the visible
// region changes (for example by text entering the buffer or scrolling)
// - INVARIANT: this function can only be called if the caller has the writing lock on the terminal
void Terminal::UpdatePatternsUnderLock() noexcept
{
auto oldTree = _patternIntervalTree;
_patternIntervalTree = _buffer->GetPatterns(_VisibleStartIndex(), _VisibleEndIndex());
_InvalidatePatternTree(oldTree);
_InvalidatePatternTree(_patternIntervalTree);
}
// Method Description:
// - Clears and invalidates the interval pattern tree
// - This is called to prevent the renderer from rendering patterns while the
// visible region is changing
void Terminal::ClearPatternTree() noexcept
{
auto oldTree = _patternIntervalTree;
_patternIntervalTree = {};
_InvalidatePatternTree(oldTree);
}
// Method Description:
// - Returns the tab color
// If the starting color exits, it's value is preferred
const std::optional<til::color> Terminal::GetTabColor() const noexcept
{
return _startingTabColor.has_value() ? _startingTabColor : _tabColor;
}
Add support for the "blink" graphic rendition attribute (#7490) This PR adds support for the _blink_ graphic rendition attribute. When a character is output with this attribute set, it "blinks" at a regular interval, by cycling its color between the normal rendition and a dimmer shade of that color. The majority of the blinking mechanism is encapsulated in a new `BlinkingState` class, which is shared between the Terminal and Conhost implementations. This class keeps track of the position in the blinking cycle, which determines whether characters are rendered as normal or faint. In Windows Terminal, the state is stored in the `Terminal` class, and in Conhost it's stored in the `CONSOLE_INFORMATION` class. In both cases, the `IsBlinkingFaint` method is used to determine the current blinking rendition, and that is passed on as a parameter to the `TextAttribute::CalculateRgbColors` method when these classes are looking up attribute colors. Prior to calculating the colors, the current attribute is also passed to the `RecordBlinkingUsage` method, which keeps track of whether there are actually any blink attributes in use. This is used to determine whether the screen needs to be refreshed when the blinking cycle toggles between the normal and faint renditions. The refresh itself is handled by the `ToggleBlinkingRendition` method, which is triggered by a timer. In Conhost this is just piggybacking on the existing cursor blink timer, but in Windows Terminal it needs to have its own separate timer, since the cursor timer is reset whenever a key is pressed, which is not something we want for attribute blinking. Although the `ToggleBlinkingRendition` is called at the same rate as the cursor blinking, we actually only want the cells to blink at half that frequency. We thus have a counter that cycles through four phases, and blinking is rendered as faint for two of those four. Then every two cycles - when the state changes - a redraw is triggered, but only if there are actually blinking attributes in use (as previously recorded). As mentioned earlier, the blinking frequency is based on the cursor blink rate, so that means it'll automatically be disabled if a user has set their cursor blink rate to none. It can also be disabled by turning off the _Show animations in Windows_ option. In Conhost these settings take effect immediately, but in Windows Terminal they only apply when a new tab is opened. This PR also adds partial support for the `SGR 6` _rapid blink_ attribute. This is not used by DEC terminals, but was defined in the ECMA/ANSI standards. It's not widely supported, but many terminals just it implement it as an alias for the regular `SGR 5` blink attribute, so that's what I've done here too. ## Validation Steps Performed I've checked the _Graphic rendition test pattern_ in Vttest, and compared our representation of the blink attribute to that of an actual DEC VT220 terminal as seen on [YouTube]. With the right color scheme it's a reasonably close match. [YouTube]: https://www.youtube.com/watch?v=03Pz5AmxbE4&t=1m55s Closes #7388
2020-09-22 01:21:33 +02:00
BlinkingState& Terminal::GetBlinkingState() const noexcept
{
return _blinkingState;
}
// Method Description:
// - Gets the internal taskbar state value
// Return Value:
// - The taskbar state
const size_t Microsoft::Terminal::Core::Terminal::GetTaskbarState() const noexcept
{
return _taskbarState;
}
// Method Description:
// - Gets the internal taskbar progress value
// Return Value:
// - The taskbar progress
const size_t Microsoft::Terminal::Core::Terminal::GetTaskbarProgress() const noexcept
{
return _taskbarProgress;
}