JetBrains.Platform.Text.Protocol Represents an undoable selection change. If you register this record in the undo stack, then when the undo action is executed, the selection is changed to the value, and a reversed is placed in the redo stack. Depending on the mode of operation, this record could be processed either on the backend or the frontend. Represents an undoable selection change. If you register this record in the undo stack, then when the undo action is executed, the selection is changed to the value, and a reversed is placed in the redo stack. Depending on the mode of operation, this record could be processed either on the backend or the frontend. Simple constructor for caret position Simple constructor for caret position with virtual offset Enumerable instead of FrugalLocalList Special factory method to create action record, which set caret to provided position only when Undo part is being executed. On Redo and initial execution caret will be stationary. Special factory method to create action record, which set caret to provided position only when Undo part is being executed. On Redo and initial execution caret will be stationary.

After positioning the caret in the document, scrolls the text control to reveal the caret position. This affects the visual location of the caret after scrolling.

Does not do any scrolling if the new caret position is already visible. If the caret gets too close to the edge, would scroll it away. This mode avoids the effect.

When moving within the active text control, ensures the caret is not too close to the edge with minimal amount of scrolling. When jumping to another text control, places the caret at the predefined position.

When moving within the active text control, works almost as , but avoids scrolling the caret away from a too-near edge if this means scrolling against the caret movement direction. When jumping to another text control, places the caret at the predefined position consistent with generic scrolling in the given direction within the same text control.

Fired at the end of document change processing when invalidated items are automatically removed from container due to document change.

Represents an () in a text document.

An offset from the specific character to the beginning of the document, calculated as the number of bytes in the character string preceding this character when encoded with UTF-16LE variable-length encoding, divided by two. This matches the indexing of late Windows NT LPCWSTR and .NET CLI System.String in-memory string representations.

Limits the current value. Returns a copy that's guaranteed to be in the [, ) limits.

Subtracts one from the current value. Does not modify the original object.

Adds one to the current value. Does not modify the original object.

Indicates whether this instance and a specified object are equal.

true if and this instance are the same type and represent the same value; otherwise, false. Another object to compare to. 2

Returns the hash code for this instance.

A 32-bit signed integer that is the hash code for this instance. 2

Compares the current instance with another object of the same type.

A 32-bit signed integer that indicates the relative order of the objects being compared. The return value has these meanings: Value Meaning Less than zero This instance is less than . Zero This instance is equal to . Greater than zero This instance is greater than . An object to compare with this instance. is not the same type as this instance. 2

Compares the current object with another object of the same type.

A 32-bit signed integer that indicates the relative order of the objects being compared. The return value has the following meanings: Value Meaning Less than zero This object is less than the parameter.Zero This object is equal to . Greater than zero This object is greater than . An object to compare with this object.

Returns the for this instance.

The enumerated constant that is the of the class or value type that implements this interface. 2

Converts the value of this instance to an equivalent Boolean value using the specified culture-specific formatting information.

A Boolean value equivalent to the value of this instance. An interface implementation that supplies culture-specific formatting information. 2

Converts the value of this instance to an equivalent 8-bit unsigned integer using the specified culture-specific formatting information.

An 8-bit unsigned integer equivalent to the value of this instance. An interface implementation that supplies culture-specific formatting information. 2

Converts the value of this instance to an equivalent Unicode character using the specified culture-specific formatting information.

A Unicode character equivalent to the value of this instance. An interface implementation that supplies culture-specific formatting information. 2

Converts the value of this instance to an equivalent using the specified culture-specific formatting information.

A instance equivalent to the value of this instance. An interface implementation that supplies culture-specific formatting information. 2

Converts the value of this instance to an equivalent number using the specified culture-specific formatting information.

A number equivalent to the value of this instance. An interface implementation that supplies culture-specific formatting information. 2

Converts the value of this instance to an equivalent double-precision floating-point number using the specified culture-specific formatting information.

A double-precision floating-point number equivalent to the value of this instance. An interface implementation that supplies culture-specific formatting information. 2

Converts the value of this instance to an equivalent 16-bit signed integer using the specified culture-specific formatting information.

An 16-bit signed integer equivalent to the value of this instance. An interface implementation that supplies culture-specific formatting information. 2

Converts the value of this instance to an equivalent 32-bit signed integer using the specified culture-specific formatting information.

An 32-bit signed integer equivalent to the value of this instance. An interface implementation that supplies culture-specific formatting information. 2

Converts the value of this instance to an equivalent 64-bit signed integer using the specified culture-specific formatting information.

An 64-bit signed integer equivalent to the value of this instance. An interface implementation that supplies culture-specific formatting information. 2

Converts the value of this instance to an equivalent 8-bit signed integer using the specified culture-specific formatting information.

An 8-bit signed integer equivalent to the value of this instance. An interface implementation that supplies culture-specific formatting information. 2

Converts the value of this instance to an equivalent single-precision floating-point number using the specified culture-specific formatting information.

A single-precision floating-point number equivalent to the value of this instance. An interface implementation that supplies culture-specific formatting information. 2

Converts the value of this instance to an equivalent using the specified culture-specific formatting information.

A instance equivalent to the value of this instance. An interface implementation that supplies culture-specific formatting information. 2

Converts the value of this instance to an of the specified that has an equivalent value, using the specified culture-specific formatting information.

An instance of type whose value is equivalent to the value of this instance. The to which the value of this instance is converted. An interface implementation that supplies culture-specific formatting information. 2

Converts the value of this instance to an equivalent 16-bit unsigned integer using the specified culture-specific formatting information.

An 16-bit unsigned integer equivalent to the value of this instance. An interface implementation that supplies culture-specific formatting information. 2

Converts the value of this instance to an equivalent 32-bit unsigned integer using the specified culture-specific formatting information.

An 32-bit unsigned integer equivalent to the value of this instance. An interface implementation that supplies culture-specific formatting information. 2

Converts the value of this instance to an equivalent 64-bit unsigned integer using the specified culture-specific formatting information.

An 64-bit unsigned integer equivalent to the value of this instance. An interface implementation that supplies culture-specific formatting information. 2

Indicates whether the current object is equal to another object of the same type.

true if the current object is equal to the parameter; otherwise, false. An object to compare with this object.

Formats the value of the current instance using the specified format.

A containing the value of the current instance in the specified format. The specifying the format to use.-or- null to use the default format defined for the type of the implementation. The to use to format the value.-or- null to obtain the numeric format information from the current locale setting of the operating system. 2

Represents the () in a text document, which is basically a document offset. However, this position is often derived from a , which is in the text control Line-Column coordinate system, and not all column indices map into document offsets. If the column index is beyond the EOL of that line, then it's said to be in the Virtual Space. is the offset into the Virtual Space from the nearest valid document .

Number of virtual columns past the last real document position. Neutral means that it is equally presented in document and text control columns (, ).

Creates an offset that represents some point within the text control's virtual space.

Creates an offset that corresponds to a real document offset, without any shift to the virtual space.

For logical positions over actual document characters this is zero. For logical positions over the Virtual Space (column indices beyond the end of line), the number of virtual space characters from the EOL.

Gets whether the point isn't over document chars, but somewhere in the virtual space of the text control. This means that and are both above zero.

For logical positions over actual document characters, the document offset. For logical positions over the Virtual Space (column indices beyond the end of line), the EOL offset.

Limits the value of . Returns a copy that's guaranteed to be in the [, ) limits.

Indicates whether this instance and a specified object are equal.

true if and this instance are the same type and represent the same value; otherwise, false. Another object to compare to. 2

Returns the hash code for this instance.

A 32-bit signed integer that is the hash code for this instance. 2

Makes a normalized range, meaning it will always go forward and have a nonnegative length, but its ends might be flipped. .

Makes an unnormalized range, meaning it might be going backwards if is less than 0.

Makes an unnormalized range, meaning it might be going backwards if is less than . See .

On an instance, describes if/how to paint a marker on the Error Stripe. For registered highlighters, derived from .

If this highlighter, when applied to a document, is to be shown on the error stripe, defines the ID of another whose foreground color gives the error stripe mark color. When not set, a NULL is used here (and never an empty string).

Default value if not , otherwise the (possibly still default) marker kind.

Groups of .

markers only.

, , . Just common markers. Errors and warnings and suggestions from the Daemon.

markers only.

Any .

The sort of a marker this highlighter makes on the Error Stripe (“Marker Bar”).

A common marker which is not from code analysis (Daemon) and not a highlighted usage. Example: T O D O item. Your custom markers probably go here. Lowest priority of the mall. : .

A code analysis (Daemon) marker which is not an error nor a warning, but still important enough to be shown on the Error Stripe as a marker and navigated to. Usually, a suggestion. Lowest priority of the Daemon markers. : .

A warning-level code analysis (Daemon) marker. Wins over suggestion, loses to error. : .

An error-level code analysis (Daemon) marker. Wins over suggestion and warning. : .

A highlight-usages marker. When present, has the highest priority because it's context-specific. However, an error stripe can split a marker into two sub-parts to simultaneously display a usage marker and a code analysis marker. : .

Transport for the gutter marks logic over the text control, serving both IDE Glyph Margin and Jet Action Margin, placed in by backend when ready to serve.

Text control transport has a bag for per-text-control extensions, like this class, and that's the ID for putting this class in the bag.

The tooltip manager wants a tooltip on front end, a request goes backwards to calculate text for the tooltip.

A list of request-response sessions for getting gutter marks in specific text view lines layout, kept by independent clients. Normally, should be only one session per text control, but as formally one can be from Jet Action Margin and one from VS Glyph Margin, let them be independent to avoid crosstalk. Created by frontend when it has a new view of either kind which needs gutter marks.

Mouse down on front, asked backward to do something (like show drop down menu on mousedown), or tell it needs a click.

Contains gutter marks that is currently on mouse hovering

Query popup over gutter mark of whatever kind. // TODO: when we think up a transport for popup window contexts, this could just originate on backend, and be attached to the context correctly on frontend

When True, gutter marks are only put on the Jet Action Margin. IDE Glyph Margin is kept clean.. When False, gutter marks are only put on the IDE Glyph Margin, and the Jet Action Margin is hidden. Applies mainly to VS.

Most recent request for info on lines, sent to backend. The response (directed forward) is expected in .

Last response to some , maybe to some prev value in that prop. Must be sorted by start offset in text view coordinates (line identity). The view is designed to use an older value when the latest one is not available yet, so this isn't a problem if the response here is not for the freshest request, provided that a response to the fresh request is given later. The view will paint with previous info in the period.

Custom option that can be passed to backend along with .

Doc ranges covered by clicked glyph position. Actual displayed icon which were clicked, to show as menu title. TODO: transport popup window contexts correctly. Front knows how to show the menu correctly, and should provide context here.

Doc ranges covered by clicked glyph position.

Actual displayed icon which were clicked, to show as menu title.

TODO: transport popup window contexts correctly. Front knows how to show the menu correctly, and should provide context here.

Range in the specific version of the visual text buffer of the text view. Not incl line break.

Indicates if outdated. Visual buffer snapshot (in which the range is taken) version number.

Executes on main thread under readlock

Is terminated when hover is finished,on main thread under readlock

Icon to display on gutter. If the implementation is based on menu items, it should elect the best icon here.

If more than one gutter mark icon lays up claim on the same text view line, this is used to determine whose icon wins and is actually displayed. The other icons won't be immediately visible. NULL means lowest priority. Note that this priority only affects choosing which icon to show if there're multiple candidates, but does not affect sorting of menu items yielded by when displaying a menu, those items got their own anchors.

Tooltip for the gutter mark icon. If the implementation is based on menu items, it should elect relevant text here.

Fills in the bulb menu items when opening a menu on the gutter mark. A NULL or empty list means there will be a gutter mark with a tooltip, but no clickable actions on it. If more than one gutter mark sits on the same text view line, their actions are combined on the menu; those with no actions are not represented on the menu.

Mouse hover handler the gutter mark icon. Is null if hover handling is not needed

A gutter mark is rendered to the left of the text view line and displays information such as base/drived types, unit testing glyph, etc. Instances of are defined per kind of similar gutter marks, not per gutter mark instance. A specific gutter mark instance is made by creating a highlighter in the markup model, and is represented by an description of a gutter mark obtained from for that specific highlighter.

Default info that is used is no specific 'IGutterMarkInfo' is passed. Should be not-null for daemon marks.

Marker interface used to hide inheritance gutter marks when VS inheritance margin is on.

NOTE: blittable, serializers not required.

Keeps the connected ranges-highlighters sets independent, in case there're multiple players.

Ranges for which highlighters are needed on front, filled by front based on the current viewport. Must be sorted. Must not overlap. NOTE: why set: we probably want to exclude massive collapsed regions. // TODO: really exclude

The highlighters collected by back, for rendering on front.

A set of highlighters with a snapshot of their positions for a certain document revision. A typed wrapper which has structural (deep) equality.

Transport model part for an adornment. NOTE: non-blittable, contains text.

For IntraText adorns used for Virtual Indentation: indent presented in the form of IntraText adorns

Joins together boolean props for .

Virtual indenting presentation form. Intra text adornments that should reserve space, but don't be drawn.

Virtual indenting presentation form. Intra text adornments that should reserve space, and show real characters count under range.

Implies that the adornment logically does NOT relate to the Inlay Hints.
If true general Inlay Hints options will NOT be applied to the adornment during rendering.

Marks that the adornment does not require Ctrl to be pressed down to respond to clicks

Icon's docking position relating to the adorn's text.

Affects adornment's presentation

Defines adornment's placement in an ascending order from left to right. For an INTER-line adornment, defines its placement in between the lines. For an INTRA-text adornment, defines its placement in between the adjacent characters. If an INTRA-text adorn is placed on a document offset N, then it's laid out in between characters at offsets N-1 and N, ordered by in ascending order. INTER-line adorns vertical order is defined by . The higher the value the lower the adorn is placed. For an INTRA-text adorn there are also special anchoring effects based on : When <0 — adornment is “anchored” to the N-1th character and prefers to stay together with it when laying out text. When >=0 — adornment is “anchored” to the Nth character and prefers to stay together with it when laying out text. Negative zero might be treated as if it were a positive zero. Avoid using zero values here for clarity.

The position of an adornment relative to the text it is attached to. AFTER_LINE_END position prevents the caret to be moved to the related adornment right side. AFTER_LINE_END position does NOT imply that the related adornment will be moved to the line end even if its offset is not located at the line end. Keep it relevant when creating a highlighter.

Requests tooltip for an adorn, offsets are adjusted to the current backend sync version.

Asks for a click reaction on the adornment, offsets are adjusted to the current backend sync version.

Transport for an individual adornment highlighter (data + range). NOTE: non-blittable, contains text in Data.

For all adorns: background color. For adorns with text: text foreground color.

Doc offset in some document version; the modification stamp of the document must be defined externally outside this object together with a set of adorns.

Keeps the connected ranges-highlighters sets independent, in case there're multiple players.

Front tells when to send push-to-hint-only adorns for rendering.

The highlighters collected by back, for rendering on front.

todo: should not be in protocol project, backend component, but currently used on frontend in adornments

Inserts a space the size of a single editor's line height above the adornment.

Inserts a space the size of a single editor's line height below the adornment.

If not null - used for calculating adorn's indent from the editor's left edge using characters width.

0.5 <= IconSizeScaleFactor <= 5

If true places adorn above line.
If false places adorn below line. Inserts a space the size of a single editor's line height above the adornment. Inserts a space the size of a single editor's line height below the adornment. If not null - used for calculating adorn's indent from the editor's left edge using characters width. 0.5 <= IconSizeScaleFactor <= 5 If true places adorn above line.
If false places adorn below line. Inserts a space the size of a single editor's line height above the adornment. Inserts a space the size of a single editor's line height below the adornment. If not null - used for calculating adorn's indent from the editor's left edge using characters width. 0.5 <= IconSizeScaleFactor <= 5

If true places adorn above line.
If false places adorn below line.

Inserts a space the size of a single editor's line height above the adornment.

Inserts a space the size of a single editor's line height below the adornment.

If not null - used for calculating adorn's indent from the editor's left edge using characters width.

0.5 <= IconSizeScaleFactor <= 5

For an intra text adornment, adds its full position description (document offset before which the adorn is inserted + intra placement details between the chars).

Doc offset in some document version; the modification stamp of the document must be defined externally outside this object together with a set of adorns.

For an intra text adornment, adds its full position description (document offset before which the adorn is inserted + intra placement details between the chars).

Doc offset in some document version; the modification stamp of the document must be defined externally outside this object together with a set of adorns.

Very similar to but with all the user customizations in the current IDE taken into account.

Lightweight IHighlighterCustomization, with different impl on frontend and backend

Defines the identity hash function for attribute IDs which are passed thru transport.

Values for , . Defines priorities for highlighters against other highlighters, but in a tricky fashion.

A tooltip on a highlighter over a text range in a text editor.

A tooltip on an error stripe mark representing a highlighter.

Text to be shown in the IDE status bar (if the highlighter is severe enough, and we decide to put its text to the Status Bar Messenger).

A tooltip over a glyph on a gutter margin of some sort, if a highlighter produces a gutter mark.

A tooltip dumped directly into a quickdoc

A bulb menu item text when gutter marks are included with the bulb menu

Gets the highlighter attributes, as defined by the highlighter registration. NO user customizations are reflected here. Use to take user customizations into account.

Gets a tooltip for the highlighter. NULL if a tooltip of this kind should not be shown.

Where this tooltip is intended to be shown. Generally, a tooltip provider does not care and would just return a tooltip, ignoring this value. The tooltip rich text, or NULL if the tooltip of this kind does not need to be shown.

If NotRecyclable is true, highlighter will be removed if its range is intersected by changes range, otherwise it will not be removed.

Gets if this attr ID is known. Otherwise calling other methods would register an error. Ideally, this should not exist, but now the highlighter attr IDs are hard coded and they sometimes violate zoning and presence of other package considerations.

Looks up the original attribute ID by its hash made with . Throws if not found.

Gets if available.

An ID. Silently skips if invalid. Entry point for container extension point for overriding part catalog set. FullPartCatalogSet cannot be used everywhere because of RSRP-498541

Used in high contrast mode only.

List of colors, delimited by ':' In case of VS2003 first COLORINDEX color will be used

Obsolete. Use ForegroundColor, DarkForegroundColor, BackgroundColor and DarkBackgroundColor to define effect colors.

A kind of the Error Stripe Marker, if this highlighter is to be shown as a marker on the error stripe when applied to a document or a text control. This is defined by set to being non-NULL. You may omit setting this value explicitly if the default () fits you, which is commonly OK. For a non-error-stripe highlighter, this MUST not be set to any non-default value. This value is considered on the original highlighter in the document or text control which has non-NULL and NOT on the highlighter pointed to by .

If this highlighter, when applied to a document or a text control, is to be shown on the error stripe, defines the ID of another whose foreground color gives the error stripe marker color. is considered on the highlighter which has a , but not on the highlighter pointed to by this ID.

Optional string which will be shown in a corresponding Color Scheme Page. Two slashes is a default IJ convention for nesting. e.g Foo//Bar, Foo//Baz will create two children under Foo

Property for deciding whether to delete highlighter immediately if it intersects with modified range or push it through highlighter-reusing-machinery (thingy which eliminates blinking). If the highlighter is produced by a long-running stage consider setting this to false for better UX ALL CARET DEPENDENT HIGHLIGHTERS MUST BE NOT RECYCLABLE!!! Crucial for Rider

If NotRecyclable is true, highlighter will be removed if its range is intersected by changes range, otherwise it will not be removed.

[Rider specific] Property for deciding whether to send updates of this highlighter to the frontend.

[Rider specific] Name of the corresponding default attributeId in R# or TextAttributesKey in IDEA platform. Used for keeping default IDEA's attributes hierarchy in settings for highlighter representation. e.g. You can configure SOME_LANGUAGE_KEYWORD highlighter color for only one particular language, or configure the colors for keywords in all languages by changing DEFAULT_KEYWORD. e.g CPP_CLASS_FIELD_ATTRIBUTE falls back to FIELD_IDENTIFIER_ATTRIBUTE e.g AnalysisHighlightingAttributeIds.ERROR falls back to CodeInsightColors.ERRORS_ATTRIBUTES in IDEA

[Rider specific] Unique id which will be used in Rider instead of one specified in Making impossible to distinguish these highlighters. It's necessary for keep default IDEA's behavior (e.g. IDEA has the same Ids for errors and warnings across all languages).

[VS specific] Generate VS ClassificationTypeDefinition and ClassificationFormatDefinition for this highlighter.

[VS specific] Used for ClassificationFormatDefinition generation.

[VS specific] A predefined VS classification type to be used as a base type for the classification definition of this highlighter.

[VS specific] Allows to define custom Order(After="..") attribute for generated VS classification format definitions.

[VS specific] Allows to define custom Order(Before="..") attribute for generated VS classification format definitions.

[Rider specific] A unique identifier corresponding to a group the highlighter belongs to

See: Microsoft.CodeAnalysis.Classification.ClassificationTypeNames Assembly: Microsoft.CodeAnalysis.Workspaces, Version=4.2.0.0, Culture=neutral, PublicKeyToken=31bf3856ad364e35 C:\Program Files\Microsoft Visual Studio\2022\Professional\Common7\IDE\CommonExtensions\Microsoft\VBCSharp\LanguageServices\Core\Microsoft.CodeAnalysis.Workspaces.dll

Additive classifications types supply additional context to other classifications.

Predefined VS classification types. Type: Microsoft.VisualStudio.Language.StandardClassification.PredefinedClassificationTypeNames Assembly: Microsoft.VisualStudio.Language.StandardClassification, Version=10.0.0.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a

VS built-in highlighting IDs, or to be precise names of well-known IClassificationTypes in Visual Studio.

Use DumpVsTextViewClassificationTagsAction internal action to know the other ones.

Error stripe might have one or more indicators above the marker bar. Or none.

Mouse down on front, asked backward to do something (like show drop down menu on mousedown), or tell it needs a click.

// TODO: when we think up a transport for popup window contexts, this could just originate on backend, and be attached to the context correctly on frontend

Defines ordering for indicators on the error stripe view, in ascending order.

Bag for custom properties.

Used by rd reflection serializers.

for this error stripe.

Kind of this marker, might be if this is a highligh-usages highlighter (?).

HighlighterAttribute ID for the source of the primary color. As we only list existing markers (not empty marker slots), this must never be NULL or empty.

Secondary color, if we show two colors per marker (e.g. error and usages).

The slot for which this marker has been allocated, as a fraction in the set of all slots (slot index divided by total number of slots). Keeping the fraction allows to stretch the markers more or less realistically when marker bar changes its height and before new markers have been calculated.

The doc range which corresponds to the slot for which the marker were created, at the moment the marker were created.

Asks for a tooltip on the marker which were created for a slot on this text range. NOTE that if the marker bar view is nontransparent, it's OK to fire this event for an empty slot (or for a specific document position even).

Asks for a click reaction on the marker which were created for a slot on this text range. NOTE that if the marker bar view is nontransparent, it's OK to fire this event for an empty slot (or for a specific document position even).

Error stripe marker slots layout. Gives a set of document offsets which split document into slices which contribute into corresponding markers. Even if VS operates on lines (when in non-word-wrap mode), this still holds document offsets, even though rounded to line boundaries. Each int is the end of the slot, non-inclusively. The first slot starts at 0 (inclusively), the next slot starts (inclusively) where the prev one ends (end non-inclusively). Populated by frontend based on the text view buffer layout against its scrollbar, and scrollbar height.

Backend analyzes highlighters in the document markup and sends which markers should be displayed on the error stripe, based on layout of the document.

Error stripe slots correspond either to the raw offset of the character (e.g. in VS Word Wrap mode), or to line numbers (in non-wrap mode, so all stuff from one line gets into the same slot).

Why IDE in Text Control solution? Temporary measure until we have proper separation for protocol entities in IDE project to pass zone validation

Get or set the readonly state of the text control. The implementation and side effects depend on text control kind (see remarks section).

Standalone SyntaxEditor text controls. Readonly is a property of a document. So if the document is open in several text contols or views, they will all be affected. Visual Studio. Readonly is a property of a text view, so if you have two views in the same tab (splitted views), one of them could be readonly while another one is editable.

Get or set the tab size used by the text control.

This option exists since VS12.

This option exists since VS16.

(DK) A temporary property, duplicates VerticalScrollbarMode. TODO redesign in 192.

Common zone for serializable stuff for text controls support

The active end of the selection (the one which moves when you do Shift+Arrow; where the caret is). When selection is represented by a range with a “start” and an “end”, this is the “end”. If asked and in doubt, use this one (that's why it's the default value for the type).

The anchor end of the selection (the one which remains in place when you do Shift+Arrow; the selection is drawn from the anchor to the current location of the caret). When selection is represented by a range with a “start” and an “end”, this is the “start”.

A view model which allows placing a text control that you have created in the backend into UI. How to use: (1) Basic scenario is when you need to just define text and coloring, which can be fully provided thru an that you get by calling . Then create a with and use as any other view model to have a view created for it. (2) If you need -specific features with your editor, such as caret operations, do the same and then use text controls in the live collection. They will appear as soon as there's one or more views running. (3) If you definitely need a single instance, or must be able to use it even before the views are created, then create an with , fill in the object with its , and use in the view model. A proper text control view will be created automatically when this viewmodel is presented.

When a text control view is created on front end and a text control gets connected to it, either an existing or a new one, it gets registered in this collection for the lifetime of the view. See docs on for details.

The document for which an editor view (one ore more text controls) will be created. You should lock an to have this stable ID for at least the lifetime of the view model. If you supply a user-created text control, this must be its document ID.

The type of the window where the text control is hosted.

Accessor for creator factory methods.

Overrides for text control preferences. Baseline values (those left NULL) are taken either from or from the underlying IDE defaults.

Optional. If you already have created a backend text control, pass it here, and the first text control view created on front end will be connected to it. Secondary views creation will be disabled in this case (so no splitter functionality). will have this as the single entry for the lifetime when the view is actually created and connected. If NULL, then front end will create text controls for every text view (there might be more than one since user is allowed to split the original text view, and each split part is a text view which needs a text control), and advertise in collection. How to choose whether to supply: • If your feature does not care about API at all (supplies text and coloring via ), then create this object with just a (e.g. with ). • If your feature wants to interact with the resulting text controls, use the collection to bind to each text control. // TODO: example • If your feature needs strictly one single text control, or needs it available before spawning the view, create it beforehand with overloads and supply

Do not translate Tab and Shift+Tab so that parent form could use them to navigate between controls.

Allows creation of secondary views, such as by splitting the main view into parts. All of the views will be in . Normally, the behavior is as follows: if is NULL, secondary views can be created; if it's defined, then it's supposed to be connected to the only view. This flag overrides the behavior in the latter case. This isn't supposed to be needed in normal features.

Type disambig for . Represents zero-based column indices in the text control view. This does not correspond directly to the document lines and columns because of the collapsed regions and other view transformations.

If the text control is hosted in a window, it requires special processing for Alt+F4, Esc, etc. Message filtering method depends on the window type.

Type disambig for . Represents zero-based line indices in the text control view. This does not correspond directly to the document lines and columns because of the collapsed regions and other view transformations.

Represents the document position in terms of Line and Column indices. Note that this has nothing to do with screen pixels, which cannot be operated on backend interfaces like .

Gets a new visual position located at the same row as the original one, but in the first column.

Returns a new position with the same row and an offset column. Does not change the current object.

A range between two s.

Normalizes the range so that its start position went before the end position.

Indicates whether this instance and a specified object are equal.

true if and this instance are the same type and represent the same value; otherwise, false. Another object to compare to. 2

Returns the hash code for this instance.

A 32-bit signed integer that is the hash code for this instance. 2

Indicates whether the current object is equal to another object of the same type.

true if the current object is equal to the parameter; otherwise, false. An object to compare with this object.

Sourcing options values from creation-time stuff. Should basically be done by all text controls upon creation. Apart from that, should be bound to the actual view. And called from user code to change either options.

Read-only. Has the effective value of the +.

Takes two sets of nullable preferences, lower-priority and higher-priority. Higher-priority takes precedence if non-NULL for a certain option, otherwise, lower-priority.

Applies when the text control is scrolled to reveal a certain position. Defines how the desired position will be accomodared inside the viewport area.

Scrolls so that the position is visible anywhere in the viewport. There're no other guarantees or heuristics applied.

Scrolls so that the position were at about the center of the viewport. The view is scrolled even if the position is already visible.

Scrolls so that the position were at about the top of the viewport. The view is scrolled even if the position is already visible.

Scrolls so that the position were in the upper half of the viewport. The view is scrolled even if the position is already visible.

Scrolls so that the position were in the lower half of the viewport. The view is scrolled even if the position is already visible.

Smart scrolling mode. Should be used in most cases. Does minimal scrolling (or none at all) required to place the position into the viewport not too close to its edges.

If the position is already visible in the view and is not too close to the edges, no scrolling is performed. This fits for navigating adjacent occurrences without the view jumping. Otherwise, places the position above the viewport center if it were above the viewport, and vice versa.

A flavor of for special cases. Relaxes the “not too close to the edge” constraint: would not scroll against the direction if the target is at least visible.

In case the text control does not support applying continuous selection, calculates the minimum spanning range and arranges its ends so that the first point is the anchor and the second is the active one. The anchor and active are determined from the given ranges in the same way as on .

The root for all protocolable stuff about text control. This is NOT a component, this is an extension for protocol. To import, take from either or depending on which side you think you are at.

Bag for custom properties. This is the way to extend basic text control transport root with transport classes for additional services around the text control.

Gets the list of transport text controls. These entities come from either side of the protocol. For all of them, backend services are created on back side, including putting an instance into automatically. On front end, it depends: * If frontend-originated, then some visual representation must already exist (e.g. IDE opened a tab). * If backend-originated, then it's intended either for: * * A user control inserted into UI — make an automation object then of the Text Control ID and place it into your view model, either a transport view model (like BeControls) from the back side, or a regular view model on the front side. A corresponding text control view would be spawned. * * A mock text control () — call to mock up a front part manually.

Use to access the Text Controls Transport Root when you're running on the back end.

Use to access the Text Controls Transport Root when you're running on the front end.

Ignores any virtual space on the range ends.

TMP cross-process root for text controls. Probably to be joined with the .

Whether the text control is visible on screen or is active. These are not exactly flags because they cannot be freely combined, later states imply all previous ones. Values can be arithmetically compared.

The text control is not visible on screen (and is not active).

The text control is currently visible on screen. It does not have keyboard focus, and its window frame is not active.

The text control is currently visible on screen (if you are in this state, is kind of implied). It does not have keyboard focus at the moment, but had it recently, and no other window frame has been activated since. This happens when you open a popup or a dialog over the text control: the text control itself loses focus, but no other window frame ("considerable piece" of IDE UI, like a tool window or a non-text-control editor tab) has been activated. Many features which are active while there is an active text control might consider this state to avoid deactivating/reactivating on every alt+enter popup, menu or refactoring shown.

The text control is currently visible on screen and has keyboard focus (if you are in this state, and are kind of implied).

Provide some sane default value for stuff which is not valid to be left zero.

This is the core implementation of this generation of text controls, friendly with cross-process back/front separation. This object is serializable into the Rd Protocol and is supposed to live on the protocol transport. On back side, it's wrapper with the classic interface for use in feature code. Normally, features only work with that interface. On frontends, there's supposed to be little or no code which works with our text control implementations but for bindings to the actual media of the host.

We wouldn't need this after we get rid of the TMP* sync calls.

Focus, visibility, and currentframedness states combined.

Sends backwards a request to execute the action handlers chain for the action identified by its ActionID. This is initiated either by user triggering a command in the text editor, or by emulating it with an call. If neither handler cancels execution, the last-chance handler would send forwards the signal so that the text control did the actual action execution in its implementation.

Sends backwards a request to execute the typing handlers chain for the given char. This is initiated either by user typing in the text editor, or by emulating it with an / call. If neither handler cancels execution, the last-chance handler would send forwards the signal so that the text control did the actual typing.

Records batch update requests, the key is just unique, and the value is the batch update request description.

After running the action handlers chain in response to and reaching the end, sends forward this signal for the editor implementation to do its own implementation execution. The frontend is supposed to know the active action execution session, so there's no data. This only could be sent in response to .

After running the typing handlers chain in response to and reaching the end, sends forward this signal for the editor implementation to do the actual typing. The frontend is supposed to know the active typing session, so there's no data. This only could be sent in response to .

In case a view is late to connect, make it a persistent prop rather than signal. A nonce allows for repeated calls.

PLAN: window frames should become protocolable objects.

Bag for custom properties. This is the way to extend basic text control transport with transport classes for additional services around the text control.

For .

Controls the behavior and layout of the text editor vertical scroll bar.

Standard vertical scroll bar without any annotations and enhancements.

Show annotations over vertical scroll bar. Annotations are: changes, marks, errors and caret position.

Show changes on vertical scroll bar. Requires Show Annotations to be enabled.

Show marks (breakpoints and bookmarks) on vertical scroll bar. Requires Show Annotations to be enabled.

Show errors on vertical scroll bar. Requires Show Annotations to be enabled.

Show caret position on vertical scroll bar. Requires Show Annotations to be enabled.

Use map mode for vertical scroll bar.

In map mode, when you click a location on the scroll bar, the cursor jumps to that location in the file instead of just scrolling up or down a page.

Lines of code are shown, in miniature, on the scroll bar. Requires Map Mode to be enabled.

Show a tooltip with a preview of the code. Requires Map Mode to be enabled.

There should be no components in this assembly, this project is indented to use only for protocol models. But our validator runs on every part attribute, so to satisfy it this zone is explicitly declared.

Whether opened text control should be getting keyboard focus or visibility. set for the file open. Null if unknown or don't care. In what type of tab the text control should be open. What type of is to be used. Set selection ranges by the same rules as (which puts caret at the active end — the last offset of the last range). NULL means dont care (IDE will use MRU values). If set, must have at least one range. Specially for cases when we do not need the instance of the opened but just need the effect of VS' opening it in a tab, without waiting for it to fully apply. Completes the task as soon as VS is instructed to open the editor, and MIGHT return NULL even if in the end an opens correctly. Skips validations related to finding the text control. This is the only way that works when the file being opened is in the middle of a rename transaction.

Whether opened text control should be getting keyboard focus or visibility.

set for the file open. Null if unknown or don't care.

In what type of tab the text control should be open.

What type of is to be used.

Set selection ranges by the same rules as (which puts caret at the active end — the last offset of the last range). NULL means dont care (IDE will use MRU values). If set, must have at least one range.

Specially for cases when we do not need the instance of the opened but just need the effect of VS' opening it in a tab, without waiting for it to fully apply. Completes the task as soon as VS is instructed to open the editor, and MIGHT return NULL even if in the end an opens correctly. Skips validations related to finding the text control. This is the only way that works when the file being opened is in the middle of a rename transaction.

Depending on option in Search and navigation options default may means either "normal tab" or "preview tab"

Never opens in a provisional tab.

Always opens in a provisional tab.

Don't care. Just cause the IDE to have the tab open, do not show or focus it if possible. The implementation MIGHT need to behave as if and .

Set keyboard focus to the newly opened editor tab, or show the existing tab and set keyboard focus to it. Implementation MUST fulfill this requirement. This is the only option which is strictly guaranteed by the implementation.

Make sure the newly opened or existing tab is visible on screen, but try not to steal keyboard focus from the current focused UI element. This might not always be possible, e.g. when keyboard focus is on a frame which would be hidden by the new tab, like in a sibling tab. The implementation MIGHT behave as if . The implementation MUST make sure the tab is visible on screen.