JetBrains.ReSharper.Psi A strongly-typed resource class, for looking up localized strings, etc. Returns the cached ResourceManager instance used by this class. Overrides the current thread's CurrentUICulture property for all resource lookups using this strongly typed resource class. Looks up a localized resource of type System.IO.UnmanagedMemoryStream similar to System.IO.MemoryStream. Options for presentating aliases. Present name of the alias. 'using A = C{int}' -> "A" Present the aliased symbol. 'using A = C{int}' -> "`C{int}`" Present name of the alias + '=' + aliased symbol. 'using A = C{int}' -> "A = `C{int}`" Present name of alias + '(' + aliased symbol + ')'. 'using A = C{int}' -> "A (`C{int}`)" Present name of alias + '(' + kind of aliased symbol + aliased symbol + ')'. 'using A = C{int}' -> "A (type `C{int}`)" 'using A = B.C' -> "A (namespace `B.C`)" Enroll listener of . After commit (successful or interrupted), client must register itself again manually to participate in a next commit. Can be called in any thread. Remove from queue of registered commit listeners. Can be called in any thread. Use this class for caching values that depend on PSI. It would NOT get invalidated on whitespace changes. Get value regardless of timestamp Use this class for caching values that depend on document or tree offsets. It gets invalidated even on whitespace changes and it isn't contained in scopes. Cookie to detect / assert that certain operations are invoked / must not be invoked during cached update. Load data from persistent storage. If cache requires write lock to store data, then load into temporary object and later merge it in method Merge loaded data into cache data structures. Executes under write lock Executes under write lock Executes under write lock Executes under write lock (probably) Synchronously update caches. Happens either when or is called. Every implementation should maintain list of dirty files by itself, usually by monitoring or invocations. For performance reasons, implementation could postpone it's update if invoked under if the cache is irrelevant to reference resolve PsiCaches notifies invalidating caches on any async file change (build/merge cycle). Use it when you need to invalidate your cache synchronously (in main thread under read-lock). Enumerates all source files from solution Obtains by index or return null if none matches 3index Return index of . Before use event must be thrown by . index Timestamp of file for which last caches were built Callback for to update for Gets persistent map for with key Gets persistent map for with key Creates map that will be dropped after program ends. Only purpose is to put some data from memory to hard drive. Indicates that we process initial file system snapshot (JobUpdateAllProjectFiles) Indicates that initial file system snapshot processing completed (JobUpdateAllProjectFiles) Indicates that caches have completed initialization and initial update (exactly - Save Job) after start-up Indicates if the async cache updater is idle right now (no foreground jobs) Indicates if the async cache updater doesn't have jobs except IBackgroundJob-s Whether PsiCaches are started. Changes from false to true once. can't become true while is false. Optimization for refactorings and long transactions. If not null then during SyncUpdate we update not all dirty files but only this file. Use through . Should be used with care (!!!) because SyncUpdate leaves caches dirty. This is persistent property (saved into db and survives ReSharper process exits). It's `false` on very first start of the solution and becomes `true` when first `SaveCachesJob` completes. The only ways it becomes `false` again is to invalidate caches. Check if cache contains dirty files due to document or PSI changes. This does not affect async cache update due to external changes () Synchronously wait until caches are ready, with a popup progress window. Must be called under a read lock. Identifies the caller of the utility, so that he could understand the reason of the delay. Defines whether to wait only primary caches or all caches. See and Title of the Cancel button, or Null if the Cancel button should be disabled. True if the caches are ready, False if user-cancelled. ! Use with care on MainThread. Temporary sets during action. If some code inside calls SyncUpdate then only caches for are rebuild. After action ends, proceed regular SyncUpdate that update caches for all dirty files Motivation: NET-CR-841, RSRP-467531 The problem with multiple contexts files during rename is that on each reference bind we: 1) In PsiFiles.HandleTransactionAction drop all psiFiles for sourceFiles from other contexts. 2) In PsiCaches.Execute on each document change we get all psiSourceFiles for projectFile and mark them as dirty in SymbolCache.OnDocumentChange 3) In SymbolCache.SyncUpdate we rebuild caches for all dirty files. That means we parse psiFiles for other contexts. So we group references from one psiSourceFiles and run honest SymbolCache.SyncUpdate only after group finished. During each reference bind SyncUpdate only current psiSourceFile (which doesn't mean reparse because we haven't dropped psiFiles). the only file that will be sync-updated when SyncUpdate is called inside action Action to execute under with special SyncUpdate filter Initializes the cache Release all data associated with cache. Cache won't receive any events any more. Declarations cache provides access to cached information about types and namespaces. Depending on its scope declarations cache may provide information about types and namespaces in whole solution including libraries in whole solution excluding libraries in project in project and its referenced libraries and projects in library and its referenced libraries Following code demonstrates use of IDeclarationsCache indexer to get information about given class. bool ClassIsSealed(string qualifiedName, ISolution solution) { DeclarationsCacheScope scope = DeclarationsCacheScope.SolutionScope(solution, false); IDeclarationsCache cache = PsiManager.GetInstance(solution).GetDeclarationsCache(scope); IDeclaredElementInfo info = cache[qualifiedName]; Logger.Assert(info.GetSymbolType() == DeclaredElementType.CLASS); return info.IsSealed(); } Checks if given typeElement is accepted by the scope of the cache Returns cached information about namespace or types with given qualified name Takes cached information about namespace or type with given qualified name and returns array with cached information about its nested namespaces and types (or just types if qualified name is a typename). Full qualified name of namespace or type Returns information about default namespace if empty string is passed. null should not be passed (throws NullReferenceException). Returns list of short names of all namespaces and types. Short name of a type is its declared name. Short name of a namespace is last part of its name without dot, e.g. in the following code namespace A.B { } short name of namespace `A.B` is `B`. Lexicographically sorted array of all short names of namespaces and types. Returns list of short names of all namespaces and types. Short name of a type is its declared name. Short name of a namespace is last part of its name without dot, e.g. in the following code namespace A.B { } short name of namespace `A.B` is `B`. List of all short names of namespaces and types. Returns array of cached information about namespaces and types having given short name. (GetAllShortNames for explanation of meaning of 'short name' term). Empty array is returned if no elements with given name are found. Empty array is returned if null is passed. Returns array of cached information about types that might directly inherit type with given short name. Type A is supposed that it might directly inherit type B if it contains reference that might resolve to B in its extends list. Reference might resolve to type B if its name is short name of type B. GetPossibleInheritors takes aliases into consideration. using X = B; class A : B {} Cached information for type A will be returned in calls to GetPossibleInheritors for type B. (And for type X if any exists in this declarations cache scope). (GetAllShortNames for explanation of meaning of 'short name' term). short name of type short type name should not be null Enumerates all possible names of the type members. The result could contains duplicate names, or names which do not match to any member Enumerates all the type elements. Exposed mostly for Import Completion feature. May perform additional checks based on alternative names of type elements. Allows retrieving source files containing identifiers with given names. Returns array of project items corresponding to all source code files containing identifiers which text is equal to given word Returns empty array if empty array is passed or no such files found. Does not expect null to be passed (throws exception) Returns array of project items corresponding to all source code files containing identifiers which text is equal to any of given words Returns empty array if empty array is passed or no such files found. Does not expect null to be passed (throws exception) Check if file contains given word Tokenize Tokenize file name. Gets a collection of items that contain all sub-words in provided query. The search query which could contain multiple sub-words. The collection of items that contains all sub-words from provided . Gets a collection of items that contain all sub-words in provided queriss. The search queries which could contain multiple sub-words each. The collection of items that contains all sub-words from provided . Determines whether provided item could contain all sub-words from provided text. The source file to inspect. The value which could contain multiple sub-words. true if contains all sub-words from provided ; otherwise, false. Gets sub-words from provided text. The value which could contain multiple sub-words. A collection of sub-words from provided . Gets sub-words of the file name at provided path. The file system path that identifies the file name to be inpected for sub-words. A collection of sub-words from the file name provided by . Determines whether provided item may contain any words in the provided word list. Unlike the , this method doesn't split input queries into subwords, and instead it acts like which uses queries as is. Gets a collection of items that contain all words in the provided word list. The word is not necessary should contain only letters, it is valid to query entities like "asp-most-beautiful-language" Gets a collection of items that contain any words in the provided word list. The word is not necessary should contain only letters, it is valid to query entities like "asp-most-beautiful-language" Queries with words to search for. The type of the requested primary language. The individual description of the job. Not included into the status bar cache notifier. Common ID for the group of jobs. It's the group progress that is displayed in the cache notifier, not jobs'. The group ID is displayed as the cache notifier main text (the only text of the status-bar cache notifier). Runs the task. The should be used only if the task requires additional progress display (like smooth execution progress or more text messages). Normally, for short-time tasks, the executing engine will handle the progress. Mark that job can be executed even if ProjectModelUpdateStates.IsInHardUpdateState is true Mark that job should affect PrimaryIdle property Do not include libraries Only libraries directly referenced from source projects Libraries referenced from source projects and their transitive closure All libraries in code model For modules that can't provide add/remove events on main thread in ChangeManager, like CppExternalModule files that are being created during better-not-interruptable parsing The additional (verbose comment) text for the notifier. This property can be set from any thread. The main (title) text for the notifier. This property can be set from any thread. Whether the cache notifier is current active. This means it should be visible on the screen, in either form, if additional conditions are met (eg VS window is active). This property can be set from any thread. The progress value for the cache notifier, must be in the [0…1] range. This property can be set from any thread. Indicates the readiness of primary caches at current moment of time. See You can rely on the value of this property only if it is requested under read lock and without further HideReadLock Indicates the readiness of caches at current moment of time. See You can rely on the value of this property only if it is requested under read lock and without further HideReadLock Sets the ownership over the notifier, so that it could not be modified without a token. Synchronously wait until caches are ready, with a popup progress window. Identifies the caller of the utility, so that he could understand the reason of the delay. Defines whether to wait only primary caches or all caches. See and . Title of the Cancel button, or Null if the Cancel button should be disabled. True if the caches are ready (but possible with dirty files), False if user-cancelled. Recalculates data under and returns data to be assigned under write lock. Psi modules to recalculate data for Data to be assigned under write lock. Null if nothing to do Stores data calculated by and return change to fire Merged results of Change to fire Cache is a that listens events from document model and file system. This map is stored in field. Normally, you should implement only method. Cache has following lifecycle: On solution load stage () method is invoked on PoolThread with no lock and its return value is used by calling on MainThread + WriteLock. For each 'Add' file system change that method is invoked to see whether file is already contained in with timestamp == modificationTimeUtc. If not method is invoked on PoolThread + ReadLock and its result is used as a parameter for method invoked on MainThread + WriteLock. is updated later by caches subsystem. Also, on startup, all files are seeded by + ! => + procedure. After startup finished method is called on MainThread + ReadLock 'Delete' file system change leads to method that simply remove element from and from 'Modify' file system change is decomposed to Remove + Add events that leads to and + method invocations. Document changes in memory leads to immediate that push file into state. While state is dirty, method always return false. After somebody asked for and PSI for this file is built, is executed. In base implementation it's just +. In base implementation is updated on ( is not used), , and methods. Also right-access methods (, , ) removes file from A trie node where this element was added to in addition to its primary node. All nested types belong to the primary node, including ones added via alternative names. Represents scope of declarations cache. Scope can be one of : whole solution including libraries - covers types and namespaces in solution whole solution excluding libraries - covers types and namespaces in source code of solution project - covers types and namespaces in source code of given project project and its referenced libraries and projects- covers types and namespaces in given project and its references including transitive ones library and its referenced libraries - covers types and namespaces in given library and libraries referenced by it See for example of use. CLRTypeName => {ForwardedTo, ForwardedFrom} Namespace => {ForwardedTo, ForwardedFrom} For every assembly, stores set of types' namespaces forwarded to it For every assembly, store assembly names from which types are forwarded For every assembly, store type names which are forwarded Caching factories for ClrTypeName and AssemblyNameInfos Find all assemblies which have this namespace forwarded to another assembly Inheritors index implementation which could be incrementally updated. Multithreaded access is guarded by read-write lock Indicates that a massive bulk operation is in progress Override if you want to participate in "Code annotations for parameter declarations" inlay hints to help user with seeing the inherited code annotations (like [NotNull]/[CanBeNull]) Indexes of containing function parameters, arguments to which should be passed to interpolated string handler constructor. -1 index corresponds to receiver. Annotated parameter isn't of interpolated string handler type. Attribute uses "" which stands for receiver but containing function is static. Attribute uses 'null' as parameter name. Attribute refers to non-existing parameter. Attribute refers to the same parameter which annotated with [InterpolatedStringHandlerArgument]. Valid but doesn't refer to any additional parameters. Valid and refers to additional parameters. Valid and refers to additional parameters but some of them are placed after interpolated string handler parameter forcing user to use named arguments in correct order. Core entity in settings detection. Wraps , have methods to add cases into the autodetect session. Also provides access to settings, supports adding cases with Synthetic settings (for now only with ) When case is added through this context, it is immediately added to autodetect session. Use with care, as in detection we don't want to be dependant on any other existing settings values, the only place where this method is used now is during modifiers order detection, to get the initial order of modifiers and then sort it Adds a possibility to change node of the detection context, this interface is supposed to be used in file processors, or other infrastructure code where we need to change node of the context. As the other approach we could make DetectionContext immutable, but this requires creating new DetectionContext for each new node, because DetectionContext must contain currently processed node, and this leads to many not needed allocations during detect, so it is better to update the node inplace. This interface is not supposed to be used in syntax-style inspectors, where the node is not meant to be changed Updates the Stores settings which require a complex way to detect them (like modifiers order from syntax styles) Updates state of one of the settings detection results which corresponds to the given set of settings, if there is no such a detection result, will create it and then update Represents the part of the setting which consists of several "checkboxes", e.g. for setting CSharpCodeStyleSettingsAccessors.USE_THIS_FOR there will be 4 synthetic settings with value true or false each of which corresponds to each member (Field, Property, Method, Event). The value of parent setting can be restored if we have any number of those synthetic settings. The reason for creation of such synthetic settings is that algorithms for calculating optimal value for the setting can't process Flag enums, but they can easily handle settings with true or false values, so instead of trying to calculate optimal value for parent setting we will calculate optimal values foreach synthetic setting and then merge them to a parent setting value. Also in detection optimization it gives us more options to control the state of detected cases for each synthetic setting (e.g. separately for Fields, Methods, Events, Properties). Helper methods to work with Ensures that there's a line break before a given node. If a line break already exists does nothing Node type Node before which a line break will be inserted Formatter profile for generated whitespace The number of line breaks to be inserted. This parameter overrides default formatter behavior and should only be used if you want to ensure a line break where formatter doesn't insert it by default. Node parameter Ensures that there's a line break after a given node. If a line break already exists does nothing Node type Node after which a line break will be inserted Formatter profile for generated whitespace The number of line breaks to be inserted. This parameter overrides default formatter behavior and should only be used if you want to ensure a line break where formatter doesn't insert it by default. Node parameter Format the whole file Format the given subtree Format the given range in the tree. The bounding nodes are random, not siblings Format the given range in the tree. The bounding nodes are random, not siblings Format code during WritePSI action Format code during WritePSI action Format code during WritePSI action Format code during WritePSI action Format code during WritePSI action For plugging our formatter into frontend Specific code style settings (for specific language, for example) should be marked by this attribute to be included in the code style settings container The component-like holder for all code style settings Get the copyright header of the file. Empty string means no copyright info Get if the copyright notice should be embraced in region. Empty region name means no embracement Code style settings are pre user Code style settings are shared in solution Code style settings are serialized to external file Name for color , if exists Create and start a new task under the task barrier. Represents the base class for all constant values. The hierarchy of constant values is semi-closed. The closed part is defined by values of the enum. Constant values of predefined types can be created via factory methods e.g. , or . Hierarchy of error values is extensible: all types of bad values must be derived from . Hierarchy of non compile-time values is extensible: all types of non compile-time values must be derived from . Special types of constant values should be derived from and their should return . Creating constant value instance from value can be done via . A default value for a given type can be created via . Constant values calculated from literals are called 'pure' and can be created from ordinary constant value by calling method. For pure constant values the returns true. Constant values calculated from references to named constants can be created by calling method. Checks whether the current instance is not compile time constant or an error value. Non compile-time constant. Error during constant computation. `null` value without type or of reference, , pointer, or unknown type. value. value. value. value. value. value. value. value. Native integer value limited to 32-bit. Native unsigned integer value limited to 32-bit. value. value. value. value. value. value. Enum value. value (allowed in Visual Basic). Not null value of type. Type value. Special language-specific value. Phantom edge represents edge which actually doesn't present in the control flow graph. It is used in "what-if" analysis, such as control flow statement redundancy. Elements connected by such edges do not take part in definite assignment / data flow analysis and are reported by the compiler as dead code. However elements reachable by these edges can contribute information required for some analyzers such as nullable reference types analysis e.g. 'true ? x : y'. While 'y' is unreachable and should be reported as dead code it contributes to the resulting nullability of the conditional expression Unreachable edge represents an edge which is present in dead code analysis but should not be processed in definite assignment / data flow analysis as it's statically known to be unreachable by the compiler's analysis Compiler skipped edge represents an edge which will be excluded during compilation e.g. to conditional invocations with unsatisfied conditions these edges are required to process invariants enforced by conditional invocations in e.g. nullable analysis Base element of control flow tree. Getters use class for performance reasons to not to box enumerator Deep value analysis (nullability/types tracking) is turned off. This mode is useful when you only need very basic analysis, like compiler is doing to check not initialized variables or out-parameters initialization. This mode enables deep value analysis to infer and check nullability issues, track real types of variables and so on. In optimistic mode we assume code entities to have unknown nullability, until having explicit [NotNull]/[CanBeNull] annotations. The same as mode, but assuming code entities to be [CanBeNull] by default. The same as mode, but assuming code entities to be [NotNull] by default. Fixup graph into canonical representation. Prepare reachable sequence for inspection and analyze reachability of elements Should be called from any inheritor constructor Represents the PSI "view" of some document range in the collection of s. Represents the PSI "view" of the collection of s. Represents the PSI "view" of some document range in the collection of s. * You should not generally rely on the order of elements in exposed collections, especially when is constructed from multiple s. However, for single case the following order is guaranteed: 1) nodes from injected PSIs (from the most nested to most outer one), 2) nodes from secondary PSIs in undefined order, 3) nodes from primary PSI. Distinct collection of PSI languages in 's context. Containing nodes in 's document range. * May contains two nodes with the same parent, if range is zero-length and located exactly in between of those two nodes. This also means that returned nodes are not in pairwise parent-child relationship even for from single . * This collection will not contain nodes with ranges contained in or intersecting with non-zero 's range. * This collection DO contains nodes from secondary PSI without valid document range - some features work great without valid document range (for example, 'rearrange code' needs to find block in order to work in Razor @{ code blocks }, but corresponding C#'s IBlock from secondary PSI has no proper document range). So it is user's responsibility to use . * Enumerating this collection do not allocates enumerators. Provides access to current document PSI view. This view is bound to one or more IPsiSourceFiles and NOT associated to caret offset or selection range in text editor. You can choose the source files set and observe them at any range. Provides access to current editor PSI view. This view is bound to one or more IPsiSourceFiles associated with editor's current document and associated to specified caret offset/selection range of current text editor. This constant requires both ISolution and ITextControl to exists. For getting data ONLY. You MUST NOT create data rules for this constant. You MUST always use instead when supplying data. Use the constant when you need all declared elements from all contexts (i.e. in case of multiple target frameworks) Use this constant if some declared element is should be accessible when you're in the editor when caret is over it's "signature". Use the constant when you need all declared elements from all contexts (i.e. in case of multiple target frameworks) Returns the declared element as the declared elements data constant: a list of one non-Null item, or Null. Prevents getting collections with null elements inside. In other words, makes it valid for the data rule. Makes sure the declared elements collection does not contain Null items and is non-empty. Filters out Nulls, returns Null instead of empty collections. In other words, makes it valid for the data rule. Represents for presentations. Provides an ability to keep presentation when the element becomes invalid A presenter that returns strings containing declared element entity kind. A presenter that returns strings containing declared element entity kind with "in" preposition. A presenter that returns strings containing declared element entity kind with "from" preposition. A presenter that returns strings containing declared element entity kind and declared element name. A presenter that returns strings containing declared element entity kind and declared element name in quotes. A presenter that returns strings containing declared element name. A presenter that returns strings containing declared element name or containing type name if constructor has no name in the target language. A presenter that returns strings containing declared element name. A presenter that returns strings containing declared element entity kind and declared element qualified name. Designed for and . A presenter that returns strings containing declared element qualified name. Designed for and . Tries to cut a long string with by replacing some of the closing parameters with "…". "Long" string containing a presentation of a . If cutting succeeds the altered str contains a "shortened" presentation of the . Marking of a . If cutting succeeds the altered str contains a "shortened" presentation of the . true if succeeds to shorten a string. This class describes positions of different components of the string presenting . This class is used as out parameter in calls to methods. See methods for examples of use. The declared element entity kind range in a string presenting . The declared element access rights range in a string presenting . The declared element modifier words range in a string presenting . The declared element type (for variables) or return type (for functions) range in a string presenting . The declared element scalar type range in a string presenting . E.g. in the string "int [] x" the scalar type range of "x" is the range of "int". Range of the explicit interface qualification of the member. The declared element name range in a string presenting . The declared element parameter list range in a string presenting . The declared element type parameter list range in a string presenting . An array of the declared element type parameter ranges in a string presenting . An array of the declared element parameter ranges in a string presenting . The declared element container (type or namespace) range in a string presenting . //TODO: what it returns for parameter? The range of the name of the container. The range of the value of a constant or a local constant in a string presenting . This class describes position of parameter-related components of the string presenting . An instance of this class is stored in the marking per each parameter of the formatted . The parameter range in a string presenting . The parameter scoped modifier range in a string presenting . The parameter kind range in a string presenting . The parameter type range in a string presenting . The range of scalar type of the parameter type in a string presenting . E.g. in the string "void Foo(int [] x)" the scalar type range of parameter "x" is the range of "int". The parameter name range in a string presenting . The parameter default value range in a string presenting . Configures presentation of for . class C { private static readonly DeclaredElementPresenterStyle ourStyle = new() { ShowName = NameStyle.SHORT, ShowParameterTypes = true, ShowParameterNames = true } public RichText Display(IDeclaredElement element, ITreeNode context) { return DeclaredElementPresenter.Format(context.Language, ourStyle, element); } } General presentation flags Indicates if we should show access rights of a (if any) Indicates if we should show the modifiers other than access rights . Indicates if we should show parameters for a . Indicates if we should show accessors in a form of `get_PropertyName` Indicates if the string should contain constructor's name in MSDN compatible form ("#ctor" instead of ".ctor") Indicates if the constructor is presented as a containing type name or using a language specific name (like 'New' in VB.NET) Indicates if the result string should contain element entity kind Indicates if the result string should contain element name Indicates if type parameters should be shown Indicates if the explicit interface qualification should be shown before type member name This option is incompatible with == . This option also disables the access rights presentation for languages that do not support access rights specification for explicit interface implementations Indicates if the result string should contain element name (including signature, return type, etc.) enclosed in quotes Indicates where and how type of a variable return type of a function should be inserted into result Indicates if the parameter list of a function should contain parameter names Indicates if the parameter list of a function should contain parameter types Indicates if the parameter list of a parameterless function should contain a special hint like (/*no parameters*/) Indicates if types should be fully qualified. Does not affect . To specify qualified names for use property. Indicates where the container type of a type member should be shown. Affects only . Indicates if the containing type or namespace of a should be shown and where it should be shown. Affects only . Indicates how the container type of a type member should be shown. Affects only . Indicates if the containing namespace of a should be shown and where it should be shown. Affects only . Indicates if the containing method of a should be shown and where it should be shown. Affects only . Configures how aliases should be presented. Sets the output string length limit. Returns interface for computing output string length Delegate for custom length checker. Length check is used to replace long parameter lists with "…". is capable of calculating length itself (. However, if there are other length considerations (e.g. UI limitations) this class may be used. Indicates if the value of a constants is to be shown. Affects constants and parameters default value. Use \u00A0 symbol in some contexts for better text wrapping (for example, between type and a name of parameters) Text styles for colorization. Presentation style for types. Styles currently used for parameter info Type of the declared element. Each type should be singleton Presentable name of the declared element Image of the declared element. Should not generally be used directly, use PsiIconManager.GetImage(). Default declared element presenter Returns presentation of this element is sound for given language type Represents type constructor. Note: ShortName of all implementations should return Default constructor - the constructor that can be invoked without any arguments provided: public C() { } public C(int x = 0) { } public C(params int[] xs) { } Parameterless constructor - the constructor w/o any parameters: public C() { } Implicit constructor - the constructor that is not directly declared in a source code: public class C { // has implicit 'public C() { }' } Default value type constructor - the parameterless struct constructor that is provided by the runtime and initialized all the struct fields with zeroes. Get the language on which this element is declared Potentially VERY expensive method, use with care. For example, in the case of a it will parse all the files with classes declared in the namespace. Get the set of source files which contains the declaration of this element Checks if there are declarations of this element in given Get the owning module of this Reference to explicit interface implementation Interface which contains implemented member Name of implemented member Resolve In C#2.0 for field, check if this field is fixed-size buffer. Returns null if this is not fixed-size buffer, otherwise returns it's size If the field is fixed-size buffer, then its type is a pointer to its element Return 'true' if this implicit type member can be replaced with the explicit member of the same signature, like with Equals/GetHashCode/PrintMembers/EqualityContract members in C# records. Returns the declared element that "declares" this implicit type member, if any. For example, for C# positional properties in records it returns the corresponding parameter of primary constructor. Indicates if the interface can be instantiated using object creation expression Gets value indicating whether this method is defined as a classic extension method (). To check if an invocation actually invokes extension method in extension form (like instance method), use helper method over instance you get from invocation expression node. Represents named tuple component element as an alias of the field of specific tuple type instantiation. The property returns component name instead of aliased field's name. The property returns component type instead of aliased field's type. Two instances of are considered equal if their names, indicies and containing types are equal. Index of component within tuple type. Type of component considering specific tuple type instantiation. Containing tuple type. Aliased field of System.ValueTuple`N type. Creates pointer to the current instance of named tuple component. Presents type member, that could be overridden. Currently overridable members are method, property, indexer and event. SuperMember for class (struct) overridable member "foo" is class member that is overridden by "foo" or interface member that is implemented by "foo". Interface member cannot have supermembers. If "A" is a class, "I" is an interface, both "A" and "I" contains member "foo", class "B" extends "A" and implements "I", but "B" does not contain member "foo", then "I.foo" is QuasiSuperMember for "A.foo". public interface I { void Foo(); } public class A { public void Foo() {} // "I.foo" is QuasiSuperMember for "A.foo", if class B exists. } public class B : A, I { // If we remove class B, "A.foo" becomes just a simple method without any supermembers/QuasiSuperMembers. } Let "A.foo" is supermember for "B.foo". If there is no "C.foo" such that "C.foo" is supermember for "B.foo" and "A.foo" is supermember foe "C.foo", then "A.foo" is immediate supermember for "B.foo". All interface members are immediate supermembers. All quasisupermembers are immediate (quasi)supermembers. If "A.foo" is supermember for "B.foo" and "A.foo" has no supermembers, then "A.foo" is root supermember for "B.foo". All interface members are root supermembers. All quasisupermembers are root (quasi)supermembers. Returns true if this member is explicit implementation. See Returns the list of the members which are implemented explicitly Returns true if this member can implement interface member implicitly Language-dependent property of IOverridableMember. Reflection implementations should return true. Finds all immediate supermembers for this member. if true, returns quasisupermembers too. optional progress indicator as the operation might take significant time Check if member has supers (inheritors) if true, returns quasi super-members too. Finds all root supermembers for this member. if true, returns quasi super-members too. Finds all supermembers for this member. if true, returns quasisupermembers too. Checks if the given member is supermember Gets kind of argument passing. Indicates whether the parameter is declared as a parameter array or as a parameter collection. Indicates whether the parameter is declared as a parameter array. Indicates whether the parameter is declared as a parameter collection. Expensive: checks the type of the parameter. Indicates whether the parameter belongs to a property setter. Indicates whether the parameter has default value or has annotation. Indicates whether the parameter is declared as vararg (`__arglist` in C#). Gets effective of parameter. Gets default value of parameter. Gets owner of parameter. Represents a function-like construct that can be invoked. Set of formal parameters of the invocable Return type of the invocable Signature of the invocable Declaration that contains parameter declarations for this element (can be delegate declaration for its Invoke method, property declaration for its accessor etc.) Logical path to some file Path to the actual physical file May differs from logic path in XAML universal apps and WinRT resource file references. Returns random file at specified path Returns project items at specified path Returns Path.ExistsDirectory by default Get the owning module of this Returns true if struct not all fields are returned in property. This could happen since private fields are not loaded through metadata Get the CLR name of type element. Immediate supertypes of the type Namespace containing this type Get the names of all type members. This set is approximate (i.e. it can contain extra names). Constructor and nested type names are not included in this list. It is much faster than getting all members itself Return single source files where the type is defined. Returns null in case it is defined in multiple files. Returns true if the type is (not necessary immediate) descendant of another type returns all members of this type (including nested types!) with given name Returns all type parameters of this type and all container types in right-to-left, nested class first order. In the following example: class A<X,Y> { class B<Z,W> {} } for the class B the result is [ W,Z,Y,X ] Returns all type parameters of this type and all container types in left-to-right, outer class first order. In the following example: class A<X,Y> { class B<Z,W> {} } for the class B the result is [ X, Y, Z, W ] Returns a type parameter by the index in a global type parameters list. Returns null if exceeds the length of a global type parameters list.
Example: for nested type A<T0, T1>.B<T2> the type parameter at index 0 is T0. Gets containing type of type member. Gets the accessibility domain of the type member, taking into account the accessibility of the containing types. Finds the members in supertypes which are hidden by this member. Gets the hide policy of type member. Gets kind of extension member. Calculates a hash of type member for solution-wide analysis caches. Local index. For A{T0, T1}.B{T2} index of the type parameter T2 is 0, not 2. The variance annotation, if any, of the type parameter declaration. Type parameters may be declared as covariant ('out'), contravariant ('in'), or neither (invariant). Indicates whether the type parameter has value type constraint ('struct'). 'unmanaged' constraint assumes value type constraint as well. Indicates whether the type parameter has reference type constraint ('class' or 'class?'). Indicates whether the type parameter has unmanaged type constraint ('unmanaged'). Indicates whether the type parameter has default (parameterless) constructor constraint ('new()'). Indicates whether the type parameter has 'notnull' constraints. Indicates whether the type parameter allows by-ref like types. Gets the nullability of type parameter based on its constraints. Gets the nullability of type parameter based on its constraints with explicit substitution for type constraints. Indicates whether type parameter has any type constraints. Gets the types specified as constraints for the type parameter. Gets the constraint flags combining all the non-type constraints specified for the type parameter. Gets the nullable annotation of the type parameter based on the nullable context of its declaration. Returns if nullable context is enabled at type parameter declaration; otherwise returns Gets the owner of type parameter. List of local type parameters (e.g. for nested type A<T0, T1>.B<T2> this list contains only T2) Identifier of this member in XML documentation file Unknown because nullable context is disabled. * no constraints and nullable context is disabled at type parameter declaration position * 'T : new()' when nullable context is disabled at type parameter declaration position * 'T : class' when nullable context is disabled at 'class' position Notnullable because type parameter constrained to notnullable value types only (doesn't allow 'Nullable{T}') * 'T : struct' * 'T : unmanaged' * 'T : struct, ISomeInterface' * 'T : struct, ISomeInterface?' Notnullable because type parameter constrained to notnullable value types (doesn't allow 'Nullable{T}') and notnullable reference types. 'T : notnull' 'T : notnull, ISomeInterface' 'T : notnull, ISomeInterface?' Nullable because type parameter accepts any value types (including nullable value types) and any reference types. * when there is no constraints and nullable context is enabled at type parameter declaration position * 'T : new()' when nullable context is enabled at type parameter declaration position Notnullable because type parameter constrained to notnullable reference types only. 'T : class' when nullable context is enabled at 'class' position Nullable because type parameter constrained to any reference types including nullable ones. 'T : class?' Notnullable because type parameter constrained with notnullable super type. 'T : BaseClassOrInterface' when nullable context is enabled at 'BaseClassOrInterface' position 'T : TNotNullableTypeParameter' 'T : class?, BaseClassOrInterface' when nullable context is enabled at 'BaseClassOrInterface' position 'T : class?, TNotNullableTypeParameter' Nullable because type parameter constrained with nullable super type. * 'T : BaseClassOrInterface?' * 'T : TNullableTypeParameter' Invariant Contra-variant, in Co-variant, out Represents an evidence of a type decoration with the specific decoration. Creates a value without checking that contains in its set of decorations. This API is internal and must be used carefully. Use method to ensure safety. Creates a value if the contains the decoration of the type in its ; otherwise returns null. A type containing the in its . A decoration attached to the . Calculates a hash of type member declaration for solution-wide analysis caches. Syntax-based, must not use resolve. Alias is not resharper specific, for example it came from editorconfig.org If this flag is used, then alias with prefix "resharper_" are not generated for non-language specific property Alias can be used to set several properties at once If this flag is omitted, then two aliases are generated: propertyname and keyprefix_propertyname If this flag is used, then only alias with a key prefix is generated (like vb_xxxx) Alias is not resharper specific, for example it came from Roslyn .NET conventions If this flag is used, then alias with prefix "resharper_" are not generated for language specific property (csharp_xxx) Don't generate alias with key prefix (no csharp_xxx alias) Return null if index type is enum and EditorConfig should just use default enum enumeration Allows entries from a settings key to be specified in .editorconfig Indicates how declared element entity kind is presented: Enum member nameDescription or example NONEDo not show entity kind (default value) NORMALnamespace NORMAL_IN_BRACKETS(in namespace) INin namespace FROMfrom namespace LANGUAGE_KEYWORDclass Base class for constraint which accepts all types assignable from a given type Constraint implementation which accepts all types assignable to a given type Interface for constraint which accepts all types assignable to a given type Base class for constraint which accepts all types assignable to a given type Represents an expectation based on expression context Completion Mode = Ignore tree after expression BlackHole Mode = Ignore only this expression Precise Mode = Take full tree Get all possible types which could present at expression's place If there are more than 1 type, then the first is more preferable. For example: Foo(int); Foo(string); ........ Foo({caret}""); expected types at caret will be (string, int), but the string is first Returns type intersection, but if intersection is empty then returns composition Represents an expectation constraint Checks if a type matches the constraint Gets some default appropriate type Gets some default appropriate types Gets array of expected types if can be performed efficiently null - if ineffective, array of types - if fast An invocation target candidate for invocation context Accepts only types which can be enumerated Accepts only types which can be enumerated Accepts types which inherit from the specified type Joins multiple constraints (a type matches joined constraint if it matches at least one of nested constraints) Constraint which accepts only reference types Constraint which accepts only super types of a given type Kind of extension member. Kind of extension member. Non-extension member. Classic C# 3-style extension method. The first parameter is a this-parameter. C# 14-style instance extension method. Member of . Unlike doesn't have this-parameter, a receiver parameter is coming from containing . C# 14-style static extension method. Member of . Extends a receiver type itself, not objects of a receiver type. C# 14-style instance extension property. Member of . C# 14-style static extension property. Member of . Extends a receiver type itself, not objects of a receiver type. C# 14-style extension operator. Member of . Check by syntax. Shouldn't resolve here to prevent deadlocks! keys are normalized qnames Implementations should override this method to bind their declarations to their declared elements Represents a query interface for retrieving extension members. Provides methods for filtering and enumerating extension members based on various criteria. Retrieves a collection of all extension method names available in the current query context. Determines whether there are extension members with any of the specified names in the query context. Filters extension members by the specified name. Filters extension members by the specified kinds. Filters query to include only extension members that may be applicable to the specified receiver type. Note that the resulting query may include extension members that are not applicable to the specified receiver type. This operation performs filtering by caches built only with syntactic information. For exact filtering, use . Filters query to include only extension members defined in modules referenced by the specified module. Filters extension members by the specified containing namespaces. Filters extension members by the specified containing types. Returns a collection of type members that satisfy the conditions defined by the current query. Base class for language-dependent cached namespace decls A node of cache tree Root of a cached declarations tree for a file Root of a cached declarations tree for a file Base class for language-dependent cached type declarations Mustn't create PSI. Better to don't make this property lazy, cause it is used in ICache.Merge(). If you invoke GetDeclarations() inside, it could lead to creating PSI and OperationCanceledException. Scalar indexer by integer type, e.g. 'array[0]' Temporarily disables code formatting. Temporarily disables expensive stages of code formatter such as indenting, wrapping, and aligning. Stub implementation of For documentation see If null - enumerate all annotations files Exposes APIs to check presence of references to declared elements in files. Implementations must be annotated with . Kind of search filter. Checks whether the search filter is available for the specified . Gets an object for the specified that will be used to check files with . Checks whether the can contain references to the declared element with . The is an object created by . Kinds of search filters. Filters files by language. Filters files by caches. Filters files by more complicated and possibly slower checks. Get the predicate to check if the given node should be checked for text occurrence Indicates that the marked class provides service for generated document Indicates that the marked class provides service for generated document Solution wide and cross language anonymous type implementation. True if further search is not needed True if further search is not needed A searcher which runs on the original search domain and then can create additional search domain and search for the second phase. Creates additional searcher and search domain for it. Language service or some technology service should implement the searcher factory to allow the finder look into files with specific language or related to specific technology Creates searchers of implicit usages for given declared elements. Enumerate words which should occur in the file to contain reference to . Return null or in enumeration to switch word filtering off Checks whether the is a special name and shouldn't be used for filtering files by words. Checks whether the declared element can be referenced from injected PSI. Get the declared elements which should be searched for along with the given element Providers a way to extend search result with additional find results, which are not required to have IReference in source code The first example of such result: show field initializer as write usage in find result Get the derived find request. When "Find Usages' found reference, this can return few more declared elements which should be found. Collection of declared elements which should be searched, predicate if their result should be consumed, and boolean value indicating if original find result is relevant, i.e. should be consumed Provider could override "Goto declaration" behaviour, by providing preferred elements The bool value indicates whether the original navigation target is relevant, out should be omitted Provider could transform navigation targets through this method. If not transformed, return the original collection or null. If transformed, return new one Provides search domain for specified declared element. Search domains from different factories are merged. Search domain for specified declared element or empty domain if not applicable Methods to modify PSI tree with reference rebinding and applying modification services. Recursively applies PSI modification service to a specified subtree. Warning: should not be child of Create the copy of given a node under sandbox. All references are fake resolved. Extensions for NotNull version of Accounts for _every_ PSI modification inside this PSI-file, advancing the current epoch Accounts for the _version_ of this PSI-file, taking into account only _committed_ (i.e. those not being rollbacked) PSI-modifications Set to the given only if file has no yet any Search for the child with given role Cached value of local node offset in parent. The value is valid only when parent. is valid -1 means unknown zero or more: offset relative to parent less than -1: absolute offset in buffer, -2 means zero Finds the deepest sub-node implementing node which contains the given range internally Specialization of with additional info (if available) about unexpected token type and expected types set for code completion. All possible languages of documents generated by this service. Is language generated by this service Additional checks, whether the file is handled by this service. Generates text of a secondary document. Temporary applied result of incremental (or full) reparse is provided. The process of generated document commit (in the case of primary document incremental reparse) can be overridden in this method. Returns null if full regeneration is required. This method is not allowed to do destructive changes due to interruptibility! Called from transaction manager when generated document is changed. If method returns false current transaction fails. Called from transaction manager when primary document is changed Custom navigation range, for instance for page declarations in ASP Checks that generated file is not corrupted Here provider may store list of nodes with injects, status of generation (which nodes already has injected PSI generated), information about changes delayed until commit, etc. Returns the default language generated by this provider. This is not the only language if is implemented! Use to always get all the generated languages for this provider. Should fail fast in case when it's impossible to inject Here we may support (incremental?) generation. processImmediately - whether we should process change immediately or can leave it until commit. Here we may support (incremental?) regeneration. processImmediately - whether we should process change immediately or can leave it until commit. Use this property to get all the languages this can generate. only returns the default generated language in this case. Don't forget to add transaction actions if you implement this Performs an auxiliary modification of a newly inserted PSI node. The standard language specific services are returned from the property. An additional services can be added by the method. Determines the priority of the service. Checks whether the service is applicable to the given insertion operation. The original node to be inserted into a PSI tree. Indicates whether belongs to a physical tree. The copy of inserted into a PSI tree. Indicates whether belongs to a physical tree. Optionally creates a modification context of the newly inserted PSI node which is later passed to a method call for every node in the inserted subtree. The original node to be inserted into PSI tree. The current reference context codec. The copy of inserted into PSI tree. An implementation must not modify neither of the parameters. Performs a modification of a single node if necessary. The node from inserted subtree. The modification context created by method. The same node or a modified node or value null if was deleted. The PSI modification engine automatically recurses into the whole subtree so an implementation of this method may not process descendant nodes unless it skips processed subtrees. Checks whether the service should be applied to descendant nodes after application to the . The processed node. The modification context created by the method. Cleans up the context created by the method. The result of applying all applicable modification services. The modification context created by method. Handles creating of a copy of a the node with the reference resolve information (). The method is invoked after copying the reference resolve information by . An implementation must not modify neither of the parameters. Configures order in which PSI modification services are applied. All modification services are applied after the code formatting and reference decoding. A service must be applied before standard language specific ones. A standard language specific service returned from (e.g. name mangling service). A service must be applied after standard language specific ones. Part of the PSI write interface. Provides the functionality to re-bind references when the subtree is inserted into physical tree Encode the references from the tree into the tree Create copy with preserving identity of ResolveResult Occurs during encoding when reference identity is set Occurs during decoding after reference was bound Called once after encoding and gives the match table between local declarations Try to extract reference from given element which could be bind Element to extract reference from Array of references to the element itself (i.e. C in A.B.C), or empty array if no bindable reference found Decode the references when the clone is inserted into the real tree and returns the clone position in the tree Since generation of documents must be pure (due to interruptibility of commit operation) all destructive actions should be done here. Methods to modify PSI tree with commiting changes to document through PSI manager Represents the scope with custom PSI modification rules. Use Case - Trivial Context (no reference rebinding, no modification services): using (PsiModificationScope.CreateTrivial()) { ... } - or more verbosely - using (new PsiModificationScope().DisableReferenceRebinding().DisableModificationServices()) { ... } Use Case - Custom Modification Services using (var scope = new PsiModificationScope()) { scope.AddModificationService(service1) .AddModificationService(service2); ... } Use Case - Disable Specific Modification Service using (new PsiModificationScope().DisableModificationService<TService>) { ... } Map from declared elements in original subtree into declaration in clone subtree Get references which should be fake-bound along with the given one Base class for custom context codec factory. Every custom factory should be used only in 'using' statement Do chain remap Check that forwarded type assembly is referenced Check that forwarded type assembly is referenced Check that forwarded type assembly is referenced If both forwarded to and forwarded from types are in list, then only "forwarded from" should survive Check that forwarded type assembly is referenced If both forwarded to and forwarded from types are in list, then only "forwarded from" should survive Prefer symbols defined in given assembly if their FQN is the same Note: symbols from assembly with "Embed Interop Types = true" counted as locals Prefer symbols defined in given assembly if their FQN is the same Note: symbols from assembly with "Embed Interop Types = true" counted as locals Filters our static methods of class object Filters our static methods of class object Filter local variable declarations by their position in the tree. Base class Filter local variable declarations by their position in the tree. Base class Scope which can store symbol table collected BEFORE this scope Tag name in which this reference occurs Indicates scope which provides additional symbols after/before specific children Provides symbols which are visible inside scope Imports = Empty - 2 Import aliases = Empty - 1 Empty namespace level = 1 File level = Empty + default namespace length Override this method if your reference has multiple names Override this method if your reference has multiple names Filters symbol table with multiple filters If result is dynamic returns candidates, otherwise declared element. If ResolveErrorType is acceptable returns declared element, otherwise returns null. "Proper" declared type means that it does not refer to type parameter Returns null if for "no-problem" Indicates that the marked class provides service for generated document Indicates that the marked class provides service for generated document Incremental reparse for secondary psi. Result of incremental reparse Apply the reparse result to the file Tries to minimize changes. Result of minimization of incremental reparse results. Reparse gives us a change to a chameleon node, like take this block/html tag and change it to a newer one. But a change like that could be too big, for example when formatter inserts/deletes spaces in some language in Razor, the other language file is reparsed and too many nodes are invalidated, making calling a formatter from write PSI/QA/CA a risky thing resulting in exceptions. So we are trying to minimize changes, but not to an extreme. We find an array of sibling nodes which could be replaced with an array of some new sibling nodes. Parent should stay the same. Apply the reparse result to the file Called on the thread where is completed Result of reparse or newly created psi-file. Used to wait for caches and execute an action when caches are ready. Suggested use when all document are committed or Maybe used to commit all documents and wait for caches by but the method should be used with care Takes read lock and commits all document. Returns CommitCookie ready to wait for caches. Most likely, should not be used. Current solution CommitCookie with read lock taken and documents are committed Assert read lock taken and documents are committed. Current solution CommitCookie ready to wait for caches Assert read lock taken and documents are committed. Valid declared element CommitCookie ready to wait for caches Source files for which PSI is outdated Find common parts Ignore whitespaces, for code formatter may work weirdly with "\r\n\n" sequences Translates ranges in the primary tree to the ranges in actual documents Main thread, read lock. After execution is false. Use only with . Main thread, read lock. Think twice before sync committing in UI thread, might be performance critical. Consider using RequirementsManager or instead to execute an action asynchronously after committing the documents Main thread, read lock, no write lock Doesn't retry when commit is interrupted, use parameter interrupted to handle that case. Consider using if you need retry. Scheduler that executes it's task only when documents are committed on background thread under ReadLock. Try to execute when Action is executing in retrying manner until its: 1. Completed successfully 2. Completed with exception that is NOT 3. is terminated Cancellation token for action. If action throws OCE it's retried only when lifetimed is not terminated. Note that this event fires for physical changes in forks Timestamp that got incremented on every non-whitespace change that is not contained in nodes. TODO: looks like it may be incremented for non-physical nodes? Timestamp that got incremented when is incremented and also on physical node changes, including whitespace node changes. Not incremented for non-physical nodes. Used by 'CachedPsiValueWithOffsets{T}'. Timestamp that got incremented when is incremented and also on physical changes outside of nodes. Used by 'CachedPsiValue{T}'. Requests async commit and starts a task when it succeeds. Throws if commit is interrupted, the task is interrupted or lifetime is not alive. Similar to but without retry and the task is arbitrary Return typ IPsiFiles component Lifetime of the task Function that will create a task Task that evaluates on committed documents, throws and exception from the task if any Starts background read action on documents committed. Throws if commit is interrupted or lifetime is not alive IPsiFiles component Lifetime of the task Action to execute on background thread under read lock when documents are committed Task Starts background read action on documents committed. Throws if commit is interrupted or lifetime is not alive Return type IPsiFiles component Lifetime of the task Action to execute on background thread under read lock when documents are committed Task Starts main thread read action on documents committed. Throws if commit is interrupted or lifetime is not alive IPsiFiles component Lifetime of the task Action to execute on main thread under read lock when documents are committed Task Starts main read action on documents committed. Throws if commit is interrupted or lifetime is not alive Return type IPsiFiles component Lifetime of the task Action to execute on main thread under read lock when documents are committed Task Translates ranges from the secondary IFile to the ranges in the original IFile. Returns tuple of original and generated ranges, null if not successful Cookie to detect / assert that certain operations are invoked / must not be invoked during PSI files build process. Should be used to get primary or main code behind file If the psi is clean - execute right away. If not - wait observe once and execute Main thread, read lock, no write lock When async, automatically retries if commit is interrupted. Creates instances of which return only references satisfying the filter passed to . Returns true if item was added, false if existing usage was updates Imagine two adjacent nodes and a virtual line break between them. Indent ops generated at the end of the left node and those generated at the start of the right node would have the same offset. But we need to distinguish between them, and also place calculating of blank line indent and calculation of right node indent somewhere between them. This is especially actual if nodes belong to different languages, because indent ops for them are generated independently and need to be merged for example in Razor @{var a = 0;var b=1;} If indent op is generated at the end of some non-zero node If indent op is generated at the start/end of some zero-length node Also for calculation of blank line indents If indent op is generated at the start of some non-zero node. For calculations of indent If delayed indent op is generated at the start of some non-zero node. This interface is for dealing with a crazy situation when one whitespace interval belongs to two adjacent tokens. For example, in C# we could have such a situation with INTERPOLATED_STRING_RAW_MULTI_LINE_START and INTERPOLATED_STRING_RAW_TEXT or with INTERPOLATED_STRING_RAW_TEXT and INTERPOLATED_STRING_RAW_END. In this case we need to know how to split whitespaces between them. Show access rights of a (if any) Show the modifiers other than access rights Present element name (including signature, return type, etc.) enclosed in quotes Indicates if the value of a constants is to be shown Parameter list of a function should contain parameter names Parameter list of a function should contain parameter types Show parameters for a . Indicates how the container type of a type member should be shown Show accessors in a form of `get_PropertyName` Should contain constructor's name in MSDN compatible form ("#ctor" instead of ".ctor") Explicit interface qualification should be shown before type member name Use \u00A0 symbol in some contexts for better text wrapping (for example, between type and a name of parameters) Indicates if the parameter list of a parameterless function should contain a special hint like (/*no parameters*/) Indicates if the constructor is presented as a containing type name or using a language specific name (like 'New' in VB.NET) On MainThread, with readlock Parsing first line of generated file: #pragma checksum "..\..\..\foo\UserControl1.xaml" #ExternalChecksum("..\..\MainWindow.xaml","{406ea660-64cf-4c82-b6f0-42d48172a799}","AE919C97FD19B80A065D1673E984D9D7") Path to generated file (.g.cs) Relative path to a file if managed to fetch Checks if file or code region was generated by custom tool (uses R# user settings) Allows to check for generated code regions while traversing the PSI tree. Language-specific interface to check if file is generated Explicitly excluded from set of files to analyse Always run analysis Interface for a language construction that may have access rights modifiers Represents an array type. Gets the type of array elements. T[] -> T T[,] -> T T[][] -> T[] Gets the rank (number of dimensions) of the array. T[] -> 1 T[,] -> 2 T[][] -> 1 (since it's an array of arrays) Implement this interface if you want your class to participate in async commit process. See . Commit will happen after this client executes (time is up to 200ms + nAttempt*30ms, where nAttempt depends on whether PSI is in bad state - current transaction not null, write lock is held, etc.) When commit happens, invokes all registered for this commit and clears the registered queue, so the client must register itself again via the method to participate in the next async commit. Method is invoked on the main thread under ReadLock before actual happens. Returns action that should be executed after async commit. Caution: if returns null, no invocation will be performed in case of async commit interruption. Action that should be executed after actual async commit in successful continuation (main thread, read lock) Action that should be executed after actual async commit in failure continuation (main thread, read lock), if was thrown. Warning: will be called ONLY if returned not null action (and, of course, async commit process was interrupted). Calculate compile time constant value Contains to provide string presentation of. Returns the markup of the string with a presentation. Returns a string containing declared element text presentation made according to this presenter settings. This method is useful when additional processing is required for the returned string, e.g. as is done in the following method: RichText Foo(IMethod method) { DeclaredElementPresenterMarking marking; RichTextParameters rtp = new RichTextParameters(ourFont); // make rich text with declared element presentation RichText result = new RichText(ourInvocableFormatter.Format(method, out marking),rtp); // highlight name of declared element in rich text result.SetColors(SystemColors.HighlightText,SystemColors.Info,marking.NameRange.StartOffset,marking.NameRange.EndOffset); return result; } Returns language specific presentation for a given parameter kind Returns language specific presentation for a given access rights value Returns language specific presentation of kind for given declared element Interface for types which might have declaration (that is, correspond to some ). For example, "IDeclaredType" and "int" are declared types. Returns ITypeElement corresponding to this type. Can return 'null' when it cannot "resolve" this type. For example, this can happen when you get type of a variable and the corresponding type usage in the source code cannot be resolved (that is, the code is not compilable). ITypeElement corresponding to this type or null if the type cannot be resolved. Assembly name of the assembly where this type is defined, or null if unspecified Returns the corresponding natural type if possible. Returns a presentation string for the type. Use property to specify whether the type should be presented in fully qualified form. Note: do not use presentable name should not be parsed to restore any information about the original type, it's only intended for presentation purposes! Check if the type is resolved successfully (and it's substitution is resolved correctly). Check if this type is unknown, i.e. it can't have the correct presentation. Get the PSI services. Check if this type is valid, i.e. references to valid declared elements. Get the module where this type was built. Marker interface indicating that give expression type couldn't be cached Returns a presentation string for the in fully qualified form. Note: long presentable name should not be used for any purposes except presentation. For example, it should not be parsed to restore any information about the original type. Returns a presentation string for the type without full qualification. Returns a presentation string for the in fully qualified form. Note: long presentable name should not be used for any purposes except presentation. For example, it should not be parsed to restore any information about the original type. Returns a presentation string for the type without full qualification. Checks if an expression of this type can be implicitly converted to a given type by the rules of the . Checks if an expression of this type can be explicitly converted to a given type by the rules of the . Interface for unsafe code function pointer type. Creates and registers secondary document. Returns all possible secondary languages of the project file. Languages are sorted alphabetically by property Passes document change to the related documents Enumerable of services for specified project file type Checks that generated files are not corrupted Enumerable of services for source file Finds 2 adjacent non-whitespace nodes and calculates the size of an imaginary empty line indent between them. Node to start searching with. Choose node's parent if possible instead passing the most granular option. If true - moves towards left trying to find the second node. If false - towards right. Interface for a language construction that may have modifiers Checks if entity either abstract, virtual or override. That is to say, it can be overridden in derived classes This method iterates and checks all nodes that compose our group: grouped before, main pattern, grouped after. It returns false if one of the following is true: 1. Blank line is encountered before main pattern 2. groupedBeforeOnlyOnSingleLine and we encounter line break 3. We cannot reach main pattern: either we encounter unknown node (not grouped and not main) or end of nodes while searching for main 4. MultilinePermissibility doesn't match our multiline status 5. We encounter comment that enables (if we go right) or disables (if we go left) highlighting before main pattern, because usually we disable comment before our main pattern and enable after Base implementation of and of Specifies whether the current operation is solution wide, can be used to set different limits if we know that all solution files will be processed. Returns whether formatter can modify space between two elements. Currently it is prohibited if either element is fully or partially included from another document. Returns whether formatter can modify element. Currently it is prohibited if element is fully or partially included from another document. Replaces spaces between two nodes by specified new nodes. Trim all left sibling nodes. Trim all left sibling nodes. Skips node's zero-length prev siblings. TODO: Better get rid of this function. We should instead account for zero-length tokens when making indent (IsIndentNode should include zero-length tokens, etc). Skips node's zero-length prev siblings. TODO: Better get rid of this function. We should instead account for zero-length tokens when making indent (IsIndentNode should include zero-length tokens, etc). Calculates the end column of the given node. It is either - If node is multiline, the it the length of the first line - If the node is singleline, then it is the end position of the node Calculates the end column of the given node. It is either - If node is multiline, the it the length of the first line - If the node is singleline, then it is the end position of the node Calculates the end column of the given node. It is either - If node is multiline, the it the length of the first line - If the node is singleline, then it is the end position of the node Calculates the end column of the line with a given node Calculates the end column of the line with a given node -1: found nothing 0: found linebreak 1: found endNode1 2: found endNode2 -1: found nothing 0: found linebreak 1: found endNode1 2: found endNode2 -1: found nothing 0: found linebreak 1: found lastNode1 2: found lastNode2 Check if there are any of the given tokens set between two given nodes in the tree Compare the position of two given nodes in the tree. By default compare nodes' StartOffset Compare the position of two given nodes in the tree. By default compare nodes' StartOffset The same as Node() It should be changed to true... some day, when somebody would be ready to deal with gold updates and performance improvements If setting is null, then an option from extraOption with Value == defaultValue is selected Create the necessary separator between two given tokens (for example. space between identifiers) If there is no necessity for separator, then return null If we indent only part of file, we need to calculate indent regions that start before this part but could be ended inside it. To do it, we take a suitable parent node (which starts at new line) and process all nodes (including children) till the start of our part. For example, to get the indent of class member, we take indent of the class and normally process all class members which precede ours. This could be very slow. To improve performance, this method could be used to provide a bound after which processing is not necessary. For example, to calculate indent regions for class, we only need to process nodes until "{", and we could skip all nodes between "{" and the start of our class member. Return parent if this parent node should not be processed at all until the relevant child node (for example, in case of file). Get the indentation of the line which contains given node Get the exact indentation of the given node Get the indentation of the line which contains given node Get the exact indentation of the given node Prerequisites: Context FirstNode and LastNode should be defined as: var firstNode = node; if (position != Position.After) { firstNode = node.FindFormattingRangeToLeft(IsWhitespaceToken, PrevFormattableSibling); if (firstNode == null) return null; } var lastNode = node; if (position != Position.Before) { lastNode = node.FindFormattingRangeToRight(IsWhitespaceToken, NextFormattableSibling); } Difference between IndentDeltaForAccu and IndentDelta is for alignment regions For them, IndentDelta contains full difference between current and previous region, while IndentDeltaForAccu only has delta in addition to align (align is calculated by accu) Signal that if there is no meaningful intersection with real document (zero-length ranges allowed), then no multilanguage accu-based formatting should take place, just plain old PSI modifications For use in Razor regenerator where we remove translated range and then do action Signal that if range is not mapped to a real document, then it shouldn't be formatted Please note non-standard operator precedence: Or() precedes And(), because all primitive conditions (like In, Satisfies, HasType) are already "and"-ed together, so rarely there is a need to explicitly "and" them. But sometimes there is a need to "And" several Or-s or Not-s. Only for cases where LPar and RPar are separated by zero-length node Doesn't work really Observed settings from indent region of this type will be added to parent indent region. Encapsulates alias optimization settings and allows to check if an alias with specified charasteristics should be removed. Creates an instance of . Calculates the set of used imports. Removes unnecessary imports and adds mandatory ones. Removes redundant qualifiers for the specified reference. Gets the kind of reference based on its location and target. Removes redundant qualifiers for static member references from the specified map. Checks if the specified reference can require an import. Adds import to resolve the specified extension member if one is required. Perform the optimization. Optimization kinds of references. Extensions have highest priority, then follows signature references, then others. The priority is taken into account for shortening and rebinding. Reference shouldn't be optimized. Reference shouldn't be optimized, but its resolve result should be saved. Reference shouldn't be shortened, but must but re-bound to its original resove result. Reference owner is located in a declaration signature. Reference owner is located in a code body. Reference refers to an extension member. Collects and sorts references which are affected by the optimization engine. This rule only describes wrap region, FormattingRule should be used to add positions of possible line breaks. Check if given declared elements are 'equal' For type Elements: This will check possible runtime binding, i.e. 2 type elements from assemblies equals if their FQN equals, and assembly names are the same For Type Members: Containing type equality + signature equality Check possible runtime binding, i.e. 2 type elements from assemblies equals if their FQN equals, and assembly names are the same Containing type equality + signature equality + extra checks Containing type equality + signature equality + extra checks http://msdn.microsoft.com/en-us/library/dd997297%28VS.100%29.aspx Any PSI change inside this scope: Affects resolve validity of inner references only Does not change 'global' model Returns null if this element actually is not scope. Intentionally global for all projects. For testing purposes only. Intentionally global for all projects. For testing purposes only. Temporarily overrides language level. For testing purposes only! Lifetime of the override New language level Project for which language level is set Temporarily overrides latest supported language level. For testing purposes only! Lifetime of the override New latest supported language level Project for which language level is set Push method's type parameter substitutions from origin to result find all interfaces of given type (except for ignored ones) which declared given member For creating operators with common type for parameters and return type Can hide implemented interface from API Compiler-generated backing field for field-like event Compiler-generated backing field for auto-implemented property Type parameters count was stripped from type name (i.e. no !1 or `1 at the end of type name) Number after ' or ! is not equal to actual number of generic parameters ! symbol used to specify number of generic parameters (i.e. !1) A collection that is optimized to store one Annotation without allocations. Works similarly to . per annotation file information cache processes xml annotation files and extracts structural information.. Serializable part of annotations for assembly attributes. per psi-assembly annotations bound to specific module. new instance of this class is created when psi-assembly is reused in tests. Serializable part of annotations for attribute instance. Serializable part of annotations for attribute argument. Serializable part of annotations for attribute argument. External providers may register attributes that should be added to [AttributeName -> TypeMemberId] index. Called when assembly loading is finished, so temporary data could be cleared Interned and cached for Imported from "C:\Program Files (x86)\Microsoft SDKs\Windows\v8.0A\bin\NETFX 4.0 Tools\WinMDExp.exe" class Microsoft.Tools.WinMDExp.ProjectedTypeInfo Updated with new entries from "C:\Program Files (x86)\Microsoft SDKs\Windows\v10.0A\bin\NETFX 4.6 Tools\WinMDExp.exe" Data in this level of cache - No value for key means "look upstream if there is upstream, otherwise recalculate" - Other value is calculated value Returns cached value for key, or FORGET if recalculation is required Resolve cache Reduce OutputAssemblies to containing project module during comparision. But Checks that the assembly is runtime (is not surface) for platforms that support System.Runtime.CompilerServices.ReferenceAssemblyAttribute If method returns true, then the assembly is exactly runtime, but if it returns false, then the assembly can be eiter runtime or surface (e.g. it always returns false for .Net Framework 2.0 runtime assemblies) Search for ":this()", ":base()" references.... Collects all names by which elements can be used in code Get all names of the possible inheritors and their count Get all names of the possible inheritors and their count Data class ... Represents an finder operation consists of number of states Operation can be interrupted and after starting (resuming) again Register collection of declared elements as "state" of operation This collection will be checked (IsValid invoken on all declared element) before each resume operation Start or Resume operation Encapsulate an applying FinderOperation with collection of Elements Schedule FinderOperation to run Declared Elements for operation FinderOperation factory Result processor Run FinderOperation sync Schedule FinderOperation to run Declared Elements for operation FinderOperation factory Result processor Run FinderOperation sync Structure for store FinderOperation and callback PsiModules with target framework id equals to current always comes first. To avoid races caused by results from source files with other TargetFrameworkId in separate thread we have to set up a barrier when first psimodule with other target framework wants to be processed By default, Finder operation can rely on validity of provided declared elements and expect no further calls if it was interrupted by WriteLock request. If you store for a period of time between write lock and use this operation after WriteLock was obtained for modifications, you must indicate your indent by using this method. Can be used anytime before first interrupt. Root, from which the search process will be started Represents order for search operations Narrow search domain to files which contains ANY of the given words Look for constructor usages in the immediate inheritors of the containing type Search for inheritors usages Search for related elements Represents a completed state of FinderOperation Calculates total number of standalone elements + standalone files + assemblies + files in each PsiModule Calculates total number of assemblies and files in each PsiModule True if further search is not needed True if further search is not needed True if further search is not needed Adds new generated pair TextRange in generated code Source range, if originalRange == ElementRange{T}.Invalid, the range is secondary Adds new generated pair TextRange in generated code Source range, if originalRange == ElementRange{T}.Invalid, the range is secondary Range type, should be some static-like instance Makes already added original range not greedy from right. Use only if you're sure map is ordered. Adds new generated pair if it doesn't intersects with other ranges, or merge it with existing ranges. Range properties must be static! TextRange in generated code Source range, if originalRange == ElementRange{T}.Invalid, the range is secondary Range type, should be some static-like instance You sure that map is ordered Used while code generation only Used during code decompilation (generated => original) If modifiedGeneratedRange.Length == 0 && it is on the bound of covering range, then return null. Otherwise, return GetCoveringRange. Adds new generated pair TextRange in generated code Source range, if originalRange == TreeTextRange{T}.Invalid, the range is secondary Adds new generated pair TextRange in generated code Source range, if originalRange == TreeTextRange{T}.Invalid, the range is secondary Range type, should be some static-like instance Adds new generated pair if it doesn't intersects with other ranges, or merge it with existing ranges. Range properties must be static! TextRange in generated code Source range, if originalRange == TreeTextRange{T}.Invalid, the range is secondary Range type, should be some static-like instance You sure that map is ordered Used while code generation only Used during code decompilation (generated => original) If modifiedGeneratedRange.Length == 0 && it is on the bound of covering range, then return null. Otherwise, return GetCoveringRange. Makes already added original range not greedy from right. Use only if you're sure map is ordered. Languages marked with this interface will invoke unchanged lexer for literals ID should match the pattern [a-zA-Z_#+:\-\.]+, or update LiteralInjectorProviderUtil Represents presence of injectable literals inside of the node. Node definitely contains injectable literals. Node definitely doesn't contains injectable literals. Node may contain injectable literals and its descendants must be checked. Will always return , even if was used for this injection Determines if the given is the descendant of the Determines if the given is the descendant of the Allows restoring an array type. Allows restoring an array type. Returns type element retargeted in the current compilation context. Type pointer for a declared type Returns false if more attempts should be made Invoked on project model changes The current (and latest parser supported) version as string The current editorconfig version The configured name of the files holding editorconfig values, defaults to ".editorconfig" Raw data from catalog: the type of the settings key class, plus some metainfo on it from the attribute. Updated sync on changes from catalogset (under main thread guard). Update invalidates the 's flag, and on next access to the schema the keys set is rebuilt (newcomer keys are introspected for entries, etc). Locked with . Very slow, don't use in critical paths Make all entries obsolete Recalculate all expired pinned values Thread-safe dictionary-based cache where values are purged after some time unless pinned. Pinned values could be automatically updated (or re-calculated). !!! Purging and updating should be scheduled and run externally. Getting/calculating values for different keys should not block each other. Calculating value for the same key blocks. Parses a single glob pattern and tests strings against it. Checks whether a given string matches this pattern. Creates a new GlobMatcher instance, parsing the pattern into a regex. Expands all brace ranges in a pattern, returning a sequence containing every possible combination. Contains options that control how Minimatch matches strings. Suppresses the behavior of treating # at the start of a pattern as a comment. Suppresses the behavior of treating a leading ! character as negation. Do not expand {a,b} and {1.3} brace sets. Disable ** matching against multiple folder names. Ignores case differences when matching. Allow patterns to match filenames starting with a period, even if the pattern does not explicitly have a period in that spot. Note that by default, a/**/b will not match a/.d/b, unless dot is set. When a match is not found by Match(), return a list containing the pattern itself. If not set, an empty list is returned if there are no matches. Returns from negate expressions the same as if they were not negated. (ie, true on a hit, false on a miss). If set, then patterns without slashes will be matched against the basename of the path if it contains slashes. For example, a?b would match the path /xyz/123/acb, but not /xyz/acb/123. If true, backslahes in paths will be treated as forward slashes. If true, backslahes in patterns will be treated as forward slashes. This disables escape characters. Used to stack the s into a single one when working with s lifted from text config files. Collection of alias + description pairs Returns true if this member has interface fqn in its name (e.g. C# compiler generates name 'Namespace.IInterface.Method' for explicitly implemented interface method 'Method' from interface 'IInterface') Similar to . Override this property if you are also overriding Non-null values are used for grouping multiple languages into a single checkbox on the Language Injections page Used for marking implementations where an automatically injected language can be changed to any other from the same group. This is different from where the automatic injection is prioritized over any other and cannot be changed. Represents a language service which provides a diagnostic information. A language can have multiple providers. A dump of all available diagnostic information for dominant PSI files can be retrieved via the internal action Internal -> PSI -> Dump PSI diagnostics. Gets the name of the provider. Returns diagnostic information for a given file as a sequence of key-value pairs. The signature (name + parameters) of IParametersOwner. Signature is used to identify matching invocables, that is to say, invocables which have the same set of parameters and the identical name. Two InvocableSignature's are equal if they have the same name, parameter types and parameter kinds. Creates signature by an IParametersOwner. It's better to use . Invocable to create signature for Specify default signature comparer. If null is given, then use strict comparer (as it is in CLR) Returns substitution that should be applied to the types of this signature before comparing them for equality. Interface for unsafe code pointer type. Obtaining the underlying type, that is type of the element on which this pointer points. Example: for "int*" returns "int". Evaluates PsiLanguageType for a project file Evaluates PsiLanguageType for a source file Get the PSI properties (if any) for the specific project file Returns most likely PsiLanguageType for a project file of corresponing ProjectFileType. i.e. XAML files normally return XAML psi, but sometimes may return XML psi, but this method should return XAML Creates lexer, that corresponds to source file text (NOT generated!) IBuffer with the text Actual source file (if exists) for which the lexer is to be generated Get custom PSI properties for specific project file type Abstraction layer between PSI and project model. Incorporates source file, item of the Returns PSI source file's persistent id. May throw exceptions if file is not valid. For logging use PSI source file's persistent id PSI file properties Storage for PSI-specific stuff Timestamp if unsaved in-memory modified document. null if document is not modified in memory This property reflects modifications done in Content Model Forks Timestamp if document is externally modified. null if document is not modified externally UTC Timestamp of persistent document Implement the interface for IPSiSourceFiles that does not belong to solution, but exist on a HDD and must be navigateable. Path to an existing file on a HDD that contain PsiSourceFile's source code created in-memory When the file is open in the editor corresponding is created in Misc files project. In the case the needs to be rebound to the project file's document Corresponding project file from Misc files project Return to in-memory document on project file removal Removed project file PSI source file which corresponds to some Indicates if this file should be parsed Indicates this file is automatically generated Generally, modifying of generated files is useless, and no warnings/suggestions should be shown for it Indicates that file should be processed in PsiCaches subsystem. If the property is to true it does not mean the PSI is built or entries in symbol cache are created. Use and correspondingly for PSI and caches. Indicates that this file provides parts of code model. Usually, this value equals to , i.e. all parsible files provides code model. The rare exceptions are something like "external sources" feature Indicates that file is out of user scope. For example, external source or content files from NuGet. No analysis and modification could be done on this file Get pre-imported namespaces for specific project file Default namespace for specific project file Preprocessor directives for this file (joint with defined in module) Marker interface for all additional PSI source file properties Some of the bool properties of might have unexpected value for different reasons (i.e. file could be heuristically considered or ). Inherit you implementation from this interface to provide extra information. Explanation why are somewhat special Any explanation or null Provider of overridable The exact properties are calculated by Biggest priority value is applied last Returns overriden properties or Properties evaluated by previous providers Project for which IPsiModule containing the source file is created Project file that corresponds to the source file. May be null Source file in question Overriden properties or Per- storage for PSI specific data. PSI source file with some location. The location is used for paths resolution, not for navigation purposes. If navigation to the file is required use Sets to evaluated value to avoid its calculation Psi source file Cached file system data Cached file system data for location. Use to update it if you expect file changes. Psi source file Cached file system data Cached file system data for location. Use to update it if you expect file changes. Psi source file Cached file system data Represents object that traverses the tree and modifies elements. Indicates whether modification process is finished. Modification context. Determines whether descendants of the node could be modified. The current tree element. Modification context. true if descendants could be modified; otherwise, false. Modifies node before traversing to its descendants. The current tree element. Modification context.

- if nothing is modified or further continuation is requested from the existing point;

- any other element of tree if transition is requested;

- null if modification cannot be continued.

 Modifies node after traversing its descendants. The current tree element. Modification context.

- if nothing is modified or further continuation is requested from the existing point;

- any other element of tree if transition is requested;

- null if modification cannot be continued.

 default implementation when not daemon is available in the context. Indicates that sandbox is in-place of the context Indicates that sandbox is a child of context The context for resolve for this sandbox. From sandbox POV, it is located *inplace* of the context node Collects project file paths that may be affected by the Project for which data is changed Old or new files data Paths that might be affected by the change Represents a CLI type. Instances of can be compared by and are equal if they represent the same type. Checks if this type can be implicitly converted to given type by the rules of the . Checks if this type can be explicitly converted to given type by the rules of the . Accepts a type visitor. Accepts a type visitor. Returns type of the scalar value of this type. T -> T T[] -> T T[][] -> T T* -> T Gets the symbol table for this type to be used in a given . According to C# specs 25.5.2 type parameter defines an open type An array type is an open type if and only if its element type is an open type A constructed type is an open type if and only if one or more of its type arguments is an open type. A constructed nested type is an open type if and only if one or more of its type arguments or the type arguments of its containing type(s) is an open type Gets the type kind, represented by the enumeration members. Checks whether values of all type parameters satisfies their constrains. Creates the pointer to the type. Determines whether the specified type is equal to the current type using a given comparer for constituent types. Gets the set of type decorations attached to this type. Recreates the type with the specified type decorations. Gets the nullability information. Recreates the type with the specified nullability. 'Unresolved' and 'Unknown' types, unconstrained type parameter types, pointer types This interface should be provided by corresponding language dependent tree node. Standard conversions are conversions which allowed to stack with user-defined conversion operators in order to make complex conversion. E.g. type T with user-defined conversion to byte is also convertible to ushort as byte is convertible to ushort via standard conversion. Following implicit conversions are considered standard: Identity, Numeric, Nullable, Reference, Boxing, Pointer, Tuple Note that underlying tuple elements' conversions could be any implicit conversion and not required to be standard. In case of tuple conversions this method might check multiple user-defined operators. Allows to restore instance without holding a strong reference to it. Gets module of original type. Finds original type. Specifies special language ID for services, which has semantic "for any language" Provide language-specific presentation of the declared elements. Returns null if default presenter should be used You should never call this method directly. Use DeclaredElementPresenter.Format> instead. Check if the given type member is visible in this language Please override this method only together with GetReferenceAccessType(IDeclaredElement target, IReference reference) Indicates where the container type of a type member should be shown. Enum member nameDescription or example NONEDo not show type (default value) AFTERfoo() in ClassName AFTER_WITH_KINDfoo() in class ClassName AFTER_IN_PARENTHESISfoo() (in ClassName) AFTER_IN_PARENTHESIS_WITH_KINDfoo() (in class ClassName) BEFORE_WITH_DOTClassName.foo() PSI module provider for regular projects Base class for delegating handlers. Primary used in filters AdditionalFilesModuleFactoryBase is a convenient base class for supporting files that are not directly included in a project model, but should be processed anyway, like files included from other files in JavaScript, editorconfig files, etc. It creates source files for them from FlushChanges method, you'll need to call it. It also implements IMiscFilesProjectPsiModuleProvider so that if that file is opened in Visual Studio, we would associate our source files with project files created for such files. Get provided modules Get the PSI module for given project file Process changes for specific project file and return list of corresponding source file changes Handler could fire any additional changes Module which reflects project output assembly Get all modules for the project Psi modules, that are referenced if a module (project or assembly) is referenced. Normally, the result contains primary module only but sometimes there might appear secondary modules either. I.e. global resources in web application. Get the PSI module for given project file Process changes of properties on underlying project, and return list of modifications Called when other modules are added or removed, so the reference resolve ot this could change Process changes for specific project file and return list of corresponding source file changes Provides PSI module interface for specific project Possibly aggregate existing handler. This interface should be implemented by a component Module name to display in UI For regular modules the same as Name Primary for web modules useful TargetFrameworkId corresponding to the source module. will always return "Primary" language of the PsiModule. For projects the property returns language obtained by For PsiModules corresponding to ASP files we return code-behind language. For Assembly PsiModules - UNKNOWN Default language for a project, UNKNOWN for assembly, code-behind language for ASP file PsiModules... The method calculates all referenced modules. The call is very expensive, use cached value IAssembly or IProject for which this PsiModule is created as either primary or secondary, null for empty project module Decorator for PSI module properties. Returned by Get current modules which aren't connected to a particular project Psi modules, that are referenced if a module (project or assembly) is referenced. Normally, the result contains primary module only but sometimes there might appear secondary modules, e.g. global resources in web application. All psi modules existing in project Returns true when compiled to CLR language of a project is supported Verifies passed against all known persistent ids and returns a unique one Persistent Id candidat Unique persistent id PSI module provider for misc files project Provides for files generated by msbuild The smaller priority - the sooner it is called. Overrides properties for the given file or returns Properties returned by previous provider Project to which the source file is added Generated source file Accumulates changed projects and fire the change via Must Be const!!! Can't change after rename Returns true when compiled to CLR language of a project is supported Takes a ReadLockCookie! Check that module can reference another specified module. Referencer may or may not provide additional data to context used for presentation or other means. Returns true if module is referenced Returns true if module with type is referenced Reference between PSI modules Referencer with lower priority executes first Checks if alias list got by denotes for global alias throws an assertion if the module is not source module and not Misc files project The IModuleReferencer with lower priority come first. Any value lower than the 's priority (0) PSI module provider for special solution projects Indicates if the containing type of a should be shown and where it should be shown. Enum member nameDescription or example NONEDo not show type (default value) AFTERNamespace in Namespace AFTER_IN_PARENTHESISNamespace (in Namespace) Indicates how the entity name should be presented Enum member nameDescription or example NONEDo not show name (default value) SHORTShow name in short form SHORT_RAWShow name in short non-verbatim form QUALIFIEDShow name in fully qualified form QUALIFIED_EXCEPT_MEMBERSShow name in fully qualified form if the element is not type member FULL_NESTED_TYPEFull nested type name without namespace FULL_NESTED_TYPE_EXCEPT_MEMBERSFull nested type name without namespace if the element is not type member 'null' for cross language elements Roslyn has a special "modifier" flag for naming rules. We have to have it also so we can order naming rules just like Roslyn (important for compatibility with editorconfig). Affects how name roots are sorted during naming suggestions. Roots are first sorted by their first and the amount of words/their length later. ElementOfPointer -> Pointer GetPointer -> Pointer ElementList -> Element Lambda parameters, Range variables Interface for components performing language-specific canonisation of symbol names. Canonises language-specific short name by returning the CLR-compilant equivalent Language-specific name (possibly escaped) Converts language-specific qualified name into CLR-compliant equivalent Populates inbound collection with language specific keywords abbreviations and lowers priority. Sample: Inbound: int (Essential) Outbound: i (Essential), int (Inane) Creates valid identifier and mangles name if necessary Applies to all languages Applies to all languages Clr languages specific For custom rule based providers Check if type is native integer or native unsigned integer without runtime support. Check if type is native integer or native unsigned integer Check if type is native integer Check if type is native unsigned integer Key to store containing NuGet package for . Wrapped to optional to give a way to store the fact that module does not belong to any NuGet package (null and absence of value are equal in current UserDataHolder implementation) T => true T where T : struct => false T where T : class => false T where T : class? => true T where T : IComparable => true T where T : IComparable? => true T where T : notnull => true C# 9 allows annotations regardless of constraints. Join nullable annotations from the set of lower bounds for fixing a type parameter. This uses the covariant merging rules. (Annotated wins over Oblivious which wins over NotAnnotated) Meet two nullable annotations for computing the nullable annotation of a type parameter from upper bounds. This uses the contravariant merging rules. (NotAnnotated wins over Oblivious which wins over Annotated) Return the nullable annotation to use when two annotations are expected to be "compatible", which means they could be the same. These are the "invariant" merging rules. (NotAnnotated wins over Annotated which wins over Oblivious) Hasn't any obsolete annotations [System.ObsoleteAttribute] or [Windows.Foundation.Metadata.DeprecatedAttribute], they are interchangeable [Windows.Foundation.Metadata.ExperimentalAttribute] [System.Diagnostics.CodeAnalysis.ExperimentalAttribute] Special [Obsolete("Types with embedded references are not supported in this version of your compiler.", true)] which compiler emits for all ref structs. It's ignored when target SDK supports ByRef-types Special [Obsolete("Constructors of types with required members are not supported in this version of your compiler.", true)] which compiler emits for constructors of types with required members. It's ignored when target SDK supports required members Indicates if the containing method should be shown and where it should be shown if the given element is a parameter. Enum member nameDescription or example NONEDo not show method (default value) AFTERparameter a of method foo AFTER_IN_PARENTHESISparameter a (of method foo) Accumulates incremental changes in the document Finds the token which covers the given position (i.e. offset in [start,end)) Token index. Less than 0 if not found Number of lexems that incremental re-lexing should step back to start relexing Abstract parser interface If a file contains non-trivial preprocessor or anything like that, RangeTranslator must be set befor file is returned Implement this language-specific interface if you need specific behaviour Scan next k tokens Type of the k'th token Scan next k tokens excepting node types of type "nodeTypeToSkip" Type of the k'th token Scan next k tokens Type of the k'th token Get the text of the current token of the lexer Get the text of the current token of the lexer Advance lexer while skip token is meet Position of first node after advancing (can be end of buffer) Advance lexer while skip tokens are meet Position of first node after advancing (can be end of buffer) This code relies that TokenStart and TokenEnd return -1 when it reaches end of file. Finds the token which covers the given position (i.e. offset in [start,end)) Token index. Less than 0 if not found Utility class for accessing Parsing services Finds the token which covers the given position (i.e. offset in [start,end)) Token index. Less than 0 if not found Rescans the buffer incrementally and returns new instance of token buffer For use in formatter, to tell if we need to insert space/new line between two token types Text to present token type to user, for example in error messages like "{0} expected" False negative detector, return true only for known file format (e.g. without false alarms) May return both false-negative and false-positive errors, but trying to be as fast and accurate as possible Simple binary file detector based on heuristic See: DefaultPerformanceThresholdSettings.xml with predefined settings for this component. Implements element pointer functionality Creates new smart element pointing to the specified declared element A smart element which resurrects from model changes by tracking its location Implements trivial element pointer functionality Implements trivial element pointer functionality A smart element which resurrects from model changes by tracking its location Implements trivial element pointer functionality Checks if given type equals to predefined sbyte, byte, short, ushort, int, uint, long, ulong, nint, nuint, decimal, double, float or char Checks if given type equals to predefined sbyte, byte, short, ushort, int, uint, long, ulong, nint, nuint or char Checks if given type equals to predefined sbyte, byte, short, ushort, int, uint, long, ulong, nint or nuint Checks if given type equals to predefined string, bool, sbyte, byte, short, ushort, int, uint, long, ulong, nint, nuint, decimal, double, float, char Checks if given type equals to predefined sbyte, byte, short, ushort, int, uint, long, ulong, nint, nuint, decimal, double, float, char Checks if given type equals to predefined sbyte, byte, short, ushort, int, uint, long, ulong, decimal, double, float, char, nint, nuint Checks if given type equals to predefined sbyte, byte, short, ushort, int, uint, long, ulong, nint, nuint or char Checks if given type equals to predefined sbyte, byte, short, ushort, int, uint, long, ulong, nint or nuint Checks if given type equals to predefined string, bool, sbyte, byte, short, ushort, int, uint, long, ulong, nint, nuint, decimal, double, float or char Enumerates possible node types Use this attribute if and only if your component's lifetime is PSI-shared (i.e. shell-level) Use other part attributes otherwise (this can be language or solution). See inheritors of IPsiIconManagerExtension Use this attribute if and only if your component's lifetime is PSI-shared (i.e. shell-level) Use other part attributes otherwise (this can be language or solution). See inheritors of IPsiIconManagerExtension Abstract interface for building model icon managers Get the empty image of the typical size Gets image for declared element Gets image for declared element Gets image by declared element type Gets image by declared element type Gets typical size of images produced by the manager Extends the specified node image (16x16) to typical size Attaches extension to given image Use the method for non-shell-lifetime extensions Use the method for non-shell-lifetime extensions Get the properties specific for language or technology For source generated files that have multiple documents returns an in-memory document used to build the corresponding IPsiSourceFile ISourceGeneratorOutputFile might have up to 2 documents - an in-memory document returned from a source generator - is used to build psi for this file and therefore always up to date - can be modified freely although the changes will be overwritten from a source generator - only exists in memory so it cannot provide document markup visible in IDE - a document visible in code editor - might be outdated and differ from its IPsiSourceFile due to having its text synced by Visual Studio - might be absent if the file is not currently opened in IDE - is usually readonly and modifying this document will result in error popup - provides markup model visible in IDE so any highlightings and other markup elements must be added to this document If you need an editor document for markup use ISourceGeneratorOutputFile.AssociatedEditorDocument. Subscribes to file system changes at location, updates its cached file system data and calls callback. Normally the callback sends corresponding to the Psi source file Lifetime of the subscription Callback for file system change. Free-threaded. Mode of modification during PSI tree traversal. Modify node only before its descendants. Modify node only after its descendants. Modify node both before and after its descendants. Modify node before its descendants and then after its descendants if one of descendants was changed. If descendants were modified reprocessing is being done up to a containing declaration. IAccessContext implementation based on element Returns form of extension member usage or value . Returns true if a classic extension method was invoked in extension form. Slow! Slow! Slow! Match file by project file type (exactly, without inheritance) Determines whether static members are accessible in this context. Staticness should be consistent with qualifier kind of access context. All qualifier Returns symbol table for code completion False means failed to resolve Returns the "owner" of the reference. Returns reference name. This name usually coincides with short name of corresponding DeclaredElement. (Only known exception is constructor initializer, its name is "this" or "base".) Returns true if reference could be resolved in declared elements with different names. E.g., attribute [XXX] could be resolved into class with name 'XXX' or 'XXXAttribute' Returns all reference names (i.e., all possible short names of corresponding DeclaredElement) Can be invoked on invalid reference! Returns full symbol table for the reference Returns text range of reference in the source file. Bind this reference to given Declared Element. May insert using directive. new "this" Bind this reference to given Declared Element and substitution. May insert using directive. new "this" Returns access context containing reference. It is useful to define, if context is static or to determine access rights of context. Late-bound reference could never be resolved to given declared element Late-bound reference could be resolved to given declared element Late-bound reference actually is not late-bound (i.e. statically typed), and resolves to given element Expected build action. when check is not necessary If empty, no filtering by extension is done Custom reference provider References which depends on first class references. If is null, all references should be enumerated. If not null, then there might be an optimization to return only the references which might have given name Custom reference provider Incrementally update reference list. Old references. Can be invalid. Returns true if this reference provider may have reference on element with one of given names Attribute for custom reference provider factory Attribute for custom reference provider factory Attribute for custom reference provider factory Types of references which can be created by the reference provider factory. Creates custom reference provider for given file Creates implementation of for the and . the specified project file the specified PSI file word index to check for words before performing main logic implementation of or null .IsValid() is false for this method call! Fired when factory settings is changed, and for all active files reference providers should be re-evaluated Container with available element names Return true if value contains any name with any case in container in any position Return true if value contains any name with any case in container in any position Represents container as enumerable of names Implements a valuation of type parameters Controls if the filter will produce the set of candidates or not and how filter handles the set of candidates produced by the previous filters. Regular filter. If this filter fails (filters out all input symbols), it will produce the error of the corresponding with the candidates from all the original input symbols. If previous filter produced the error (filtered out all input symbols), this filter simply won't run and won't affect the set of candidates (if any). If this filter fails it will produce the error of the corresponding with the candidates from all the original input symbols (just like ). If previous filter produced the error with non-empty set of candidates, candidates will be filtered by this filter. If the candidates don't pass this filter - the previous error will be replaced with the from this filter and candidates set will remain intact. Otherwise, the set of candidates will remain the same or be shrinked by this filter. If this filter fails, it will produce the error of the corresponding without set of candidates provided (unlike REGULAR and MUST_RUN). If previous filter produced error with non-empty set of candidates, candidates will be filtered by this filter. If the candidates don't pass this filter - the previous error will be replaced with the ErrorType from this filter and empty candidates set. Tries to filter the symbols into the list, and returns true if the filtering result is actually stored in list. If filtering is a no-op (no symbols to filter out or replace), the implementation is free to use the list as a temporary list and return false to indicate that no actual filtering is actually performed. list must not be modified. list is guaranteed to be empty. "depth level" of declared symbol for imported symbol (class or namespace) level = -1 for namespaces and top-level classes without containing namespace level = 0 for top-level classes with containing namespace level = 1 for class members (including inner classes) level = level of containing class + 1 for method parameters and local variables level = level of containing class + 2 Symbol table deals with case-insensitive names Uncached, preserves the original items order. Used for symbol table builder. Determines if type members should be put into the table Used for symbol table builder. Determines if type members should be put into the table Check if the symbol table with can be used when this mode is requested Symbol table should contain only symbols, necessary for ReferenceName resolving. The vital that type members shouldn't be obtained Complete symbol table - all necessary infos should be there if can't handle uri, pass forward Depends on language specific implementation of the interface ITypeInferenceBoundsSolver Not scoped A reference scoped to the enclosing block or method A value scoped to the enclosing block or method 20220427081910_DropTranslationValueTable.Designer.cs TODO: this method is incorrect because base type substitition applied to a derived class has no sense. For original (i.e. first-class) result, this is null. For related search, this points to original result. Type element which was used as basis for quasi-implementation Transforms find result into user data. Can bind to PSI here. Method is unsynchronized. Returns null to skip result, will not be invoked. Accepts data from . Method is synchronized. Allows to provide file-specific information for searching object creations. Indicates whether the service is applicable to . Checks whether the can contain invocations of . Takes into account only explicit invocations ignoring implicit object creations (e.g. C# implicitly creates interpolated string handlers and collection builders). Attribute type name without 'Attribute' suffix Allows to provide language specific information for searching parameters. Gets files where parameter can be used. Regular parameters are only available within declarations of their containing and therefore is enough to determine the scope. However some languages have primary parameters which are available within declarations of their containing . Null for trivial domains. Expand whole solution domain with list of concrete modules Visit all IPsiModules contained in implicit trivial domain. return false to stop visiting Handy class for running search actions Inject or use The visitor does not traverse PSI tree, all methods are just different entry points called _once_ for domain element. If you need to process subnodes call recursive element processor manually in both VisitPsiSourceFile and VisitElement. Show 'Scanning file "Foo.cs" (2 / 567)' hint starting from this amount of search domain items. Leave intact or specify 0 to disable detailed progress reporting An element with same signature, but not overriding search target Compare like Execution Engine does. Case sensitive, ref/out parameters differs and type modifiers are significant The file can be overwritten or Rider Roslyn Worker restarted, so documentId+fileName pair changed. In this case we should remove the file first then add it as usual Controlled from Source Generator. {generatorAssembly}/{generatorFQN}/{name_provided_in_source_generator_code} After making psi source file invalid the ProjectModelPath is unavailable for source generated files To remove such files, we have to maintain bidirectional map with cached paths. We use these properties for Equals/GetHashCode in {nameof(myPendingAnalyzerReferences)} only")] We use these properties for Equals/GetHashCode in {nameof(myPendingAnalyzerReferences)} only")] Do PSI write bulk operation. Upon success, ensure all modified documents are writable Status of attempt to modify involved documents Add transaction action with no change in PSI. After creation of the custom modification cookie and until its disposal processing of changes from original to generated files (for example, from ASP to code behind C# filed and back) and vice versa are changed: 1. Changes to original file are not promoted to generated files, only range translation is corrected as needed. 2. Even if a changed range in generated file only partially intersects with translated range (but is not fully inside translated range) and change is not whitespace transaction is still processed and range translation is corrected for non-intersected ranges (without custom cookie such transaction would fail). Additional handlers of a transaction action. Similar to Applies transaction document changes to other files (i.e. IFile's of shared project files) Each of the document changes of the transaction action Replaced text Transaction action being handled Manages nested PSI transactions Manages nested PSI transactions Transaction succeeded Transaction failed due to some internal reasons Transaction failed due to OperationCanceledException is caught (usually, user press "cancel" button in progress window) The attempt to ensure writable failed for some files Some files were modified as a result of the attempt to make them writable Creates a new range from offset and length, rather than from start + end offsets, as the .ctor would do. Creates a new range from zero offset and length, rather than from start + end offsets, as the .ctor would do. Gets the or , whichever is smaller. Use for consistent processing of potentially non-normalized ranges. Gets the or , whichever is greater. Use for consistent processing of potentially non-normalized ranges. Gets whether this range is a subset of the . Gets whether this range is a proper subset of the . Determines whether the offset falls within the range, start and end offsets included. Checks whether the character at the given index falls within this range. Unlike , the right offset is not included, because the range ends before the character with such an index. Ranges [2,4) and [1,3) intersects. Ranges [2,4) and [1,2) contacts. See for strict intersection Calculates range of intersection with given range range to intersect with Range of intersection if current range intersects with given range. Empty range if current range contacts with given range (ie [1,2) and [2,3) ). Invalid range otherwise Checks that the range is not the . Asserts that this range is normalized, which means that its is nonnegative. Asserts that this range is normalized, which means that its is nonnegative. Gets whether this range is normalized, which means that its is nonnegative. Returns a normalized version of the current text range (with a nonnegative length). Returns the distance between the and the nearest point that belongs to the range. Gets whether this range is normalized, which means that its is nonnegative. Creates a new range from offset and length, rather than from start + end offsets, as the .ctor would do. Creates a new range from zero offset and length, rather than from start + end offsets, as the .ctor would do. Gets the or , whichever is smaller. Use for consistent processing of potentially non-normalized ranges. Gets the or , whichever is greater. Use for consistent processing of potentially non-normalized ranges. Gets whether this range is a subset of the . Gets whether this range is a proper subset of the . Determines whether the offset falls within the range, start and end offsets included. Checks whether the character at the given index falls within this range. Unlike , the right offset is not included, because the range ends before the character with such an index. Checks that the range is not the . Asserts that this range is normalized, which means that its is nonnegative. Asserts that this range is normalized, which means that its is nonnegative. Returns a normalized version of the current text range (with a nonnegative length). Returns the distance between the and the nearest point that belongs to the range. Translate coordinates in any document into range the given psi tree. Translate the specified range in synchronized text into actual document range Get the union of intersecting document ranges Get all ranges in actual document which intersects with given range in synchronized text Should not be used except in special cases! Get all ranges in actual document which intersects with given range in synchronized text Also allows to include ranges that are normally not included (for example, HTML mapped to C# WriteLiteral in Razor) Get all ranges in the file which intersects with given range in actual document. Try to incrementally reparse the chameleon block lexer of the new text range in text which is modified Length of the inserted portion of the text new chameleon if successful. null if can't synchronize This interface is implemented by implementations when there is defaut namespace Get the XML node for this doc comment, using given as the header Doc comment XML node Get errors from XML parsing Type of expression. This function never returns null, if type cannot be determined it returns UnknownType. Obtains the caching lexer on file text Incrementally reparse file range in the old file text which is changed text which should replace given range the new file (may be the same) Starting from 0, each time the subtree is modified this counter increases Returns name denoted by this identifier without language-specific stuff like '@' in C# Returns qualified name of the namespace this declaration declares Get references which should be visited by resolver, and which are syntax-dependent (not semantic) Returns containing node of corresponding type. Whether to include the starting element * PLEASE, use static members of 'TTreeNodeNavigator'-like types to navigate to parent nodes in a type-safe manner if possible. * PLEASE, use language-specific extensions like C#'s "GetContainingTypeMemberDeclarationIgnoringClosures" to navigate common parent types. * This method traverses tree through nodes. Returns true if the subtree rooted at this element contains other element Checks if this element is located at physical file (and not in sandbox) Checks if this element is valid Checks if this element is filtered (whitespace, comment or error) Returns an offset in the text to move caret to when positioning to the element An offset in the text Implementations must always return parameter Get the text as a buffer Returns text representation of an element For performance reason, it's better to use method Finds the smallest element containing the specified text range Text range in this element Finds all elements containing the specified offset and such that none of the elements contains another element meeting this condition Offset to find element at Collection of found elements If an element returned by this method is not a leaf, it should be empty Returns the token containing the character corresponding to the specified offset User data bag for this element which is copied with the element Unifies cross-language reference names and reference expressions Attention!!! Synchronize changes with Attention!!! Synchronize changes with Helper methods to work with Optimized version of that skips leaf tree nodes (like tokens), and navigates in tree via non-interface calls (using direct field access). Use with care! Returns first token, that supplied node consists of. Returns last token, that supplied node consists of. Compare the position of two given nodes in the tree. By default compare nodes' StartOffset Compare the position of two given nodes in the tree. By default compare nodes' StartOffset Returns first source element satisfying the predicate and located before the current element Returns first tree node satisfying the predicate and located after the current node Returns next child which is meaningful in the language (not a comment and not a whitespace) Returns prev child which is meaningful in the language (not a comment and not a whitespace) Get the starting coordinate of this subtree in real document Get the starting coordinate of this subtree in real document Get the coordinates of this subtree in real document Get the coordinates of intersecting parts of this subtree in real document Get the union of intersecting document ranges Get the document range from the tree beginning till the end offset (if possible) Creates the copy of this element and puts it into dummy holder. The node to be cloned context for resolving, or null for none Creates the copy of this element and puts it into dummy holder. All references in the copy are fake-resolved. Searches for containing node inside of starting from . If achieved, but containing node not yet found then returns null. is eligible as a containing node. If isn't inside of then it will search through all containing nodes. Finds all references whose ranges in text contain the specified text range Get the starting coordinate of this subtree in real document Indicates if the containing type or namespace of a should be shown and where it should be shown. Enum member nameDescription or example NONEDo not show type (default value) AFTERClassName in Namespace AFTER_WITH_KINDClassName in namespace Namespace AFTER_IN_PARENTHESIS_WITH_KINDClassName (in namespace Namespace) Allows to customize a type by adding some additional data or by changing different aspects of its behavior. Gets called when a including this type decoration is being applied to the underlying type. Allows to create an updated type decoration with data computed for the underlying type (e.g. based on a new substitution). Returns the type which must be used instead of the specified underlying type. Returns a symbol table with additional symbols for a given decorated type with the specified symbol table of its underlying type. Transforms a symbol table after merging of original symbol table from an underlying type with additional symbols collected from all type decorations in a set with the method. Customizes the presentation of a type (). Customizes the string presentation of a type (). Creates a pointer allowing to restore the type decoration. Translated type decorator to another fork. Returns itself for trivial decorators. Compares two type decorations to determine if they are equal. Compares two type decorations to determine if they are not equal. Allows to restore a type decoration. Immutable set of type decorations. Contains uniquely-typed decorations (like map from to ). Adding decoration of the same type overwrites an existing value. An empty set. Applies the for every type decoration and forms a set from the results. Gets the number of type decorations in this set. Gets an enumerator. Although does not implement this method allows to 'foreach' over the set in allocation-less manner. Returns a type decoration of the specified type if it's present in a set; otherwise returns null. Determines whether the set contains a type decoration of the specified type. Gets a nullable annotation if the set contains ; otherwise returns . Gets a list of type modifiers if the set contains ; otherwise returns an empty list. Adds the specified type decoration to the set. In case the set contains a type decoration of the same type the old value gets overwritten by the value. Removed a type decoration of the specified type. Updates the nullability decoration in the set. Creates a pointer allowing to restore the set. Creates a set with the specified nullability. Creates a set with the specified type modifiers. Creates a set with the specified nullability and type modifiers. Creates a set from a single type decoration. Creates a set from two type decorations. Creates a set from three type decorations. Creates a builder allowing to construct a set. Produces a set that contains type decorations from this and specified set. Two decorations from different sets of the same type are merged by the . In case returns null none of these two is included into a resulting set. Produces a set that contains type decorations that exist in both this set and the specified set. Two decorations from different sets of the same type are merged by the . In case returns null none of these two is included into a resulting set. Dumps the contents of the set to a string. Allows to restore a type decoration set. Creates declared type from given type element with dynamic substitution, i.e. where any type parameter corresponds to dynamic type Creates a declared type for a given type element with given type parameter values. Type arguments should be specified for all type parameters of type element and its containing type elements. They should appear in array in outer-class-first, left-to-right order. modopt modreq Holds 'modopt' and 'modreq' modifiers. Types which has different type modifiers sets are still equals. For now, they are compared in signatures Allows to get a compact type presentation with some omissions. No compaction. Value indicating whether the type name compaction is enabled. Amount of depth levels of type arguments to include into presentation. Given the Dictionary{string, List{object}} type. 2: Dictionary{string, List{object}} 1: Dictionary{string, List{...}} 0: Dictionary{...} The style of the type parameters presenting. Enum member nameDescription or example NONEDo not show type parameters CLRCLR-style, i.e. the number of type parameters after ` INDICATEShows syntax indicating presence of type parameters, but doesn't list them FULLShow type parameters with their names(default value) FULL_WITH_VARIANCEShow type parameters with their names(default value) and in/ outs Use keywords for predefined types. For example, "int" instead of "System.Int32" in C# or "Integer" instead of "System.Integer" in Visual Basic. Append '?' to nullable reference types. Append '!' to non-nullable reference types. Use special syntax for presenting tuple types instead of presenting as regular generic type. Use keywords for native integer types if runtime supports them "nint", "nuint" instead of "System.IntPtr" and "System.UintPtr". Use name of the type alias symbol when type is aliased. For example, if there is type alias 'using A = System.String' then "A" will be used for presentation instead of "System.String". Qualify with extern alias. For example, "global::" in C# Use special syntax for Nullable{T} type. For example, "T?" in C# Value indicating whether type should be presented with full qualification. Type presentation options. Text styles for rich markup. Type name compaction. Indicates how return type of a function or type of a variable are presented Enum member nameDescription or example NONEDo not show type (default value) BEFORELanguage default (for example: Type foo) AFTERfoo:Type Check that accessibility if is not lower that accessibility of Returns false in the following cases: - cache directory does not exist - cache format version did not match - one or more record's CRC did not match - total cache CRC did not match - IO problem occurred In all these cases it automatically resets the cache on the disk to an empty one. The key feature - files with unchanged persistent id should keep old index A type T{A1, …, An} is variance-convertible to a type T{B1, …, Bn} if T is either an interface or a delegate type declared with the variant type parameters T{X1, …, Xn}, and for each variant type parameter Xi one of the following holds: • Xi is covariant and an implicit reference or identity conversion exists from Ai to Bi • Xi is contravariant and an implicit reference or identity conversion exists from Bi to Ai • Xi is invariant and an identity conversion exists from Ai to Bi Select the single type in the collection, which is implicitly convertible to all others Checks if two different types are indistinguishable in runtime even when no conversion exists e.g. int[] and uint[] byteEnum[] and sbyte[] long[] and IntPtr[] (32-bit runtimes) Returns element type if collectionType is a _collection_type_ (described in specification, 15.8.4) otherwise, returns null Get the topmost type which contains corresponding 'GetEnumerator' by given type of collection Check if expression calls only to pure methods, and has no side effects. Custom check predicate allows to filter-out write to locals. Check if expression calls only to pure methods, and has no side effects. Custom check predicate allows to filter-out write to locals. Check if expression calls only to pure methods, and has no side effects. Custom check predicate allows to filter-out write to locals. Returns all declarations of a type where type member can be created. Preferable declarations are first. Methods acts as follow: 1. Get all declarations of a type. 2. Ask language specific services for extra declarations (e.g. base type of ASP page...) 3. Filter declarations in other languages... 4. Try filter generated declarations... Like windows forms designer code... 5. Filter declarations written in languages other then required... 6. Try use preferred declaration... Returns all declarations of a type where type member can be created. Preferable declarations are first. Methods acts as follow: 1. Get all declarations of a type. 2. Ask language specific services for extra declarations (e.g. base type of ASP page...) 3. Filter declarations in other languages... 4. Try filter generated declarations... Like windows forms designer code... 5. Filter declarations written in languages other then required... 6. Try use preferred declaration... Checks if there is any type which is both and Checks if there is any type which is both and Language service for accessing collection type info. The class provides static methods which call to language specific implementation of this interface. By default: yield break By default: yield return declaration Works on auto-generated code-behind classed. By default - false By default - false Works on auto-generated type members. By default - true Default - true Default - true Default - true Default - true Default - false Default - typeElement.GetContainingElement<INamespaceDeclaration>(false); Default - projectFile.Location.Directory Default - element.DeclaredName Default - false Default - InvalidRange Default - false Default - true Get the base constructor call from given constructor Return the candidates to navigation actions null if you want to use default reference resolve result Empty list will exclude this reference from navigation Otherwise return candidates list to navigate Extend search element Wraps implementation so that all operations go through interning instance. Wraps implementation so that all operations go through interning instance. not a format method Formatting operation with arguments accessed by indices and passed immediately, e.g. System.String.Format(string, params object[]) Formatting operation with arguments accessed by indices, but not yet provided, e.g. System.Text.CompositeFormat.Parse(string) Formatting operation with positional named arguments, e.g. Microsoft.Extensions.Logging.ILogger.Log(string, params object[]) Formatting operation with positional named arguments, but not yet provided, e.g. Microsoft.Extensions.Logging.LoggerMessage.Define(string) Formatting annotation with case-insensitive named arguments, that are bound to parameters, e.g. Microsoft.Extensions.Logging.LoggerMessageAttribute(string) Assumes that the is an attribute on some s If is an inline params collection, returns its elements; otherwise returns argument value. Utility for saving comments after transformation of tree. State for MoveCommentsUtil State for MoveCommentsUtil Creates snapshot of tree for moving comments Moves comments from old tree to new tree Retrieves the obsolete information for a declared element in the corresponding context. Note: reference is not considered obsolete if it's referenced from permitted context (e.g. from another obsolete method etc.) Retrieves the obsolete information for a declared element in the corresponding context. Note: reference is not considered obsolete if it's referenced from permitted context (e.g. from another obsolete method etc.) Returns true is the entity is obsolete in all versions of compiler. Note: this method ignores special compiler-emitted [Obsolete] for new language features (required members, byref-types etc.) Note: this method is not aware of context, so element could be considered obsolete even if it's referenced from another obsolete method. Returns true is the entity is obsolete in all versions of compiler. It ignores special compiler-emitted [Obsolete] for new language features (required members, byref-types etc.) Note: this method is not aware of context, so element could be considered obsolete even if it's referenced from another obsolete method. Returns true is the entity is experimental or obsolete in all versions of compiler. Note: this method ignores special compiler-emitted [Obsolete] for new language features (required members, byref-types etc.) Note: this method is not aware of context, so element could be considered obsolete even if it's referenced from another obsolete method. Find overloaded method with same signature but required return type Find optional parameter in current method, or parameter in overloaded method with same signature Find overloaded method with similar signature but required return type and additional parameter Find optional parameter in current method, or overloaded method with same signature without parameter Performs project model data cleanup in the sake of memory consumption. Should not happen in tests or command-line tools. A set of utilities for converting from a decimal floating-point literal string to its IEEE float or double representation, which coniders all digits signficant and correctly rounds according to the IEEE round-to-nearest-ties-to-even mode. This code does not support a leading sign character, as that is not part of the C# or VB floating-point literal lexical syntax. Hey, dear %RANDOM_DUDE%, periodically check Roslyn's 'Microsoft.CodeAnalysis/RealParser.cs' file for changes and merge them here! Try parsing a correctly-formatted double floating-point literal into the nearest representable double using IEEE round-to-nearest-ties-to-even rounding mode. Behavior is not defined for inputs that are not valid C# floating-point literals. The decimal floating-point constant's string True if the input was converted; false if there was an overflow Try parsing a correctly-formatted float floating-point literal into the nearest representable float using IEEE round-to-nearest-ties-to-even rounding mode. Behavior is not defined for inputs that are not valid C# floating-point literals. The float floating-point constant's string The nearest float value, if conversion succeeds True if the input was converted; false if there was an overflow Properties of an IEEE floating-point representation. Converts the floating point value 0.mantissa * 2^exponent into the correct form for the FloatingPointType and stores the bits of the resulting value into the result object. The caller must ensure that the mantissa and exponent are correctly computed such that either [1] the most significant bit of the mantissa is in the correct position for the FloatingType, or [2] the exponent has been correctly adjusted to account for the shift of the mantissa that will be required. This function correctly handles range errors and stores a zero or infinity in the result object on underflow and overflow errors, respectively. This function correctly forms denormal numbers when required. If the provided mantissa has more bits of precision than can be stored in the result object, the mantissa is rounded to the available precision. Thus, if possible, the caller should provide a mantissa with at least one more bit of precision than is required, to ensure that the mantissa is correctly rounded. (The caller should not round the mantissa before calling this function.) The bits of the mantissa The exponent Whether there are any nonzero bits past the supplied mantissa Where the bits of the floating-point number are stored A status indicating whether the conversion succeeded and why Properties of a C# float. Properties of a C# double. This type is used to hold a partially-parsed string representation of a floating point number. The number is stored in the following form: 
                0.Mantissa * 10^Exponent
The Mantissa buffer stores the mantissa digits as characters in a string. The MantissaCount gives the number of digits present in the Mantissa buffer. There shall be neither leading nor trailing zero digits in the Mantissa. Note that this represents only nonnegative floating-point literals; the negative sign in C# and VB is actually a separate unary negation operator. Create a DecimalFloatingPointString from a string representing a floating-point literal. The text of the floating-point literal Convert a DecimalFloatingPointString to the bits of the given floating-point type. This function is part of the fast track for integer floating point strings. It takes an integer stored as an array of bytes (lsb first) and converts the value into its FloatingType representation, storing the bits into "result". If the value is not representable, +/-infinity is stored and overflow is reported (since this function only deals with integers, underflow is impossible). the bits of the integer, least significant bits first the number of bits of precision in integerValueAsBytes whether there are nonzero digits after the decimal the kind of real number to build the result An indicator of the kind of result Parse a sequence of digits into a BigInteger. The DecimalFloatingPointString containing the digits in its Mantissa The index of the first digit to convert The index just past the last digit to convert The BigInteger result Return the number of significant bits set. Return the number of significant bits set. Return the number of significant bits set. Return the number of significant bits set. Computes value / 2^shift, then rounds the result according to the current rounding mode. By the time we call this function, we will already have discarded most digits. The caller must pass true for has_zero_tail if all discarded bits were zeroes. The value to shift The amount of shift Whether there are any less significant nonzero bits in the value Determines whether a mantissa should be rounded up in the round-to-nearest-ties-to-even mode given [1] the value of the least significant bit of the mantissa, [2] the value of the next bit after the least significant bit (the "round" bit) and [3] whether any trailing bits after the round bit are set. The mantissa is treated as an unsigned integer magnitude. For this function, "round up" is defined as "increase the magnitude" of the mantissa. the least-significant bit of the representable value the bit following the least-significant bit true if there are any (less significant) bits set following roundBit Multiply a BigInteger by the given power of two. The BigInteger to multiply by a power of two and replace with the product The power of two to multiply it by Multiply a BigInteger by the given power of ten. The BigInteger to multiply by a power of ten and replace with the product The power of ten to multiply it by Convert a float value to the bits of its representation Convert the bits of its representation to a float A union used to convert between a float and the bits of its representation Summary description for StringLiteralConverter. Creates a new instance of CharEscape SurvivableTreeRangePointer doesn't guarantee that it returns the same range. Most likely in case of PSI node's death it would return tokens (first and last in case of trackExternals, last and first otherwise). But it is far more likely to return anything at all that simple TreeRangePointer, because it is based on one large range marker for the entire range instead of two small range markers from TreeRangePointer which could die oh-so-easily in mixed language PSI modification. Returns target declaration Returns target declaration file Sometimes we have messed up PSI structure like this: #if SOME_DEFINE // this belongs to file // Here starts UsingList using Aaa; #endif // this still belongs to UsingList #if SOME_DEFINE using Bbb; // Here ends UsingList #endif // this belongs to file, not to UsingList This is because preprocessor directives and comments are considered not meaningful, and UsingList must start with meaningful child. In this case, when we want to add a new using above first "#if SOME_DEFINE", we would need to move "non-meaningful" lexems after it inside UsingList. This method gets a range of meaningful or "non-meaningful" nodes, before which we want to add a new node, and a parent node where this new node should go according to PSI logic, and checks if we need to move anything inside parent. Node where a new node belongs according to PSI logic (like UsingList) A range of nodes (meaningful or not), which we want to add our node before Anchor, before which we should add our node Works only with range lists where ranges cannot strictly intersect with each other and that are normalized ECMA-335 part II (Metadata) paragraph 9.2 "Generics and recursive inheritance graphs" Redirects given type to that one which is 'actual' in the context of given source module Rules of conversion: convert embedded COM-interop type to original convert framework classes to which are used in given source module (AKA binding redirect, AKA framework unification) Ecma-335, Partition II, 13 Semantics of value types return false if adding member causes problem in flattening graph Check if type is value non-nullable type Check that type is either Nullable or a generic type constrained to Nullable Select the single type in the collection, which is implicitly convertible to all others Select the single type in the collection, to which all others are implicitly convertible Redirects given type to that one which is 'actual' in the context of given source module Rules of conversion: convert embedded COM-interop type to original convert framework classes to which are used in given source module (AKA binding redirect) Relative virtual path relative to website root or VirtualFileSystemPath.GetEmptyPathFor(GetInteractionContext.SolutionContext) if fails Relative virtual path relative to website root or VirtualFileSystemPath.GetEmptyPathFor(GetInteractionContext.SolutionContext) if fails Returns immediate base elements. Modifies passed string builder typeMember.Name != short name for explicitly implemented members (e.g. Namespace.IInterface<T>.MemberName and MemberName) index of first type argument (in arguments array) for nested types The accessibility domain of some type member The accessibility domain of some type member Unrestricted access to member Access is restricted to assembly, in which member is defined Access is restricted to and it derivatives Access is restricted to , all it derivatives, and the assembly where member is defined Access is restricted to and it derivatives in assembly where member is declared Access is restricted to only Access is restricted to file only Error occurred while computing accessibility domain Determines whether the specified is string literal. element for test true - if success; otherwise - false Determines whether the specified is constant literal. element for test true - if success; otherwise - false Builds index for compiled XML docs assembly to index doc for indicates whether indexing should be synchronous, or may execute in another thread Returns true if index was updated Xml file lexer with byte offset in file for each token Container for getting ALL (not only leaf) applicable components