clang-tools-extra/clangd/index/SymbolCollector.h

   1 //===--- SymbolCollector.h ---------------------------------------*- C++-*-===//
   2 //
   3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
   4 // See https://llvm.org/LICENSE.txt for license information.
   5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
   6 //
   7 //===----------------------------------------------------------------------===//
   8 #ifndef LLVM_CLANG_TOOLS_EXTRA_CLANGD_INDEX_SYMBOLCOLLECTOR_H
   9 #define LLVM_CLANG_TOOLS_EXTRA_CLANGD_INDEX_SYMBOLCOLLECTOR_H
  10
  11 #include "CollectMacros.h"
  12 #include "clang-include-cleaner/Record.h"
  13 #include "clang-include-cleaner/Types.h"
  14 #include "index/Ref.h"
  15 #include "index/Relation.h"
  16 #include "index/Symbol.h"
  17 #include "index/SymbolID.h"
  18 #include "index/SymbolOrigin.h"
  19 #include "clang/AST/ASTContext.h"
  20 #include "clang/AST/Decl.h"
  21 #include "clang/Basic/SourceLocation.h"
  22 #include "clang/Basic/SourceManager.h"
  23 #include "clang/Index/IndexDataConsumer.h"
  24 #include "clang/Index/IndexSymbol.h"
  25 #include "clang/Sema/CodeCompleteConsumer.h"
  26 #include "llvm/ADT/DenseMap.h"
  27 #include <functional>
  28 #include <memory>
  29 #include <optional>
  30
  31 namespace clang {
  32 namespace clangd {
  33
  34 /// Collect declarations (symbols) from an AST.
  35 /// It collects most declarations except:
  36 /// - Implicit declarations
  37 /// - Anonymous declarations (anonymous enum/class/struct, etc)
  38 /// - Declarations in anonymous namespaces in headers
  39 /// - Local declarations (in function bodies, blocks, etc)
  40 /// - Template specializations
  41 /// - Library-specific private declarations (e.g. private declaration generated
  42 /// by protobuf compiler)
  43 ///
  44 /// References to main-file symbols are not collected.
  45 ///
  46 /// See also shouldCollectSymbol(...).
  47 ///
  48 /// Clients (e.g. clangd) can use SymbolCollector together with
  49 /// index::indexTopLevelDecls to retrieve all symbols when the source file is
  50 /// changed.
  51 class SymbolCollector : public index::IndexDataConsumer {
  52 public:
  53   struct Options {
  54     /// When symbol paths cannot be resolved to absolute paths (e.g. files in
  55     /// VFS that does not have absolute path), combine the fallback directory
  56     /// with symbols' paths to get absolute paths. This must be an absolute
  57     /// path.
  58     std::string FallbackDir;
  59     bool CollectIncludePath = false;
  60     /// If set, this is used to map symbol #include path to a potentially
  61     /// different #include path specified by IWYU pragmas.
  62     const include_cleaner::PragmaIncludes *PragmaIncludes = nullptr;
  63     // Populate the Symbol.References field.
  64     bool CountReferences = false;
  65     /// The symbol ref kinds that will be collected.
  66     /// If not set, SymbolCollector will not collect refs.
  67     /// Note that references of namespace decls are not collected, as they
  68     /// contribute large part of the index, and they are less useful compared
  69     /// with other decls.
  70     RefKind RefFilter = RefKind::Unknown;
  71     /// If set to true, SymbolCollector will collect all refs (from main file
  72     /// and included headers); otherwise, only refs from main file will be
  73     /// collected.
  74     /// This flag is only meaningful when RefFilter is set.
  75     bool RefsInHeaders = false;
  76     // Every symbol collected will be stamped with this origin.
  77     SymbolOrigin Origin = SymbolOrigin::Unknown;
  78     /// Collect macros.
  79     /// Note that SymbolCollector must be run with preprocessor in order to
  80     /// collect macros. For example, `indexTopLevelDecls` will not index any
  81     /// macro even if this is true.
  82     bool CollectMacro = false;
  83     /// Collect symbols local to main-files, such as static functions, symbols
  84     /// inside an anonymous namespace, function-local classes and its member
  85     /// functions.
  86     bool CollectMainFileSymbols = true;
  87     /// Collect references to main-file symbols.
  88     bool CollectMainFileRefs = false;
  89     /// Collect symbols with reserved names, like __Vector_base.
  90     /// This does not currently affect macros (many like _WIN32 are important!)
  91     /// This only affects system headers.
  92     bool CollectReserved = false;
  93     /// If set to true, SymbolCollector will collect doc for all symbols.
  94     /// Note that documents of symbols being indexed for completion will always
  95     /// be collected regardless of this option.
  96     bool StoreAllDocumentation = false;
  97     /// If this is set, only collect symbols/references from a file if
  98     /// `FileFilter(SM, FID)` is true. If not set, all files are indexed.
  99     std::function<bool(const SourceManager &, FileID)> FileFilter = nullptr;
 100   };
 101
 102   SymbolCollector(Options Opts);
 103   ~SymbolCollector();
 104
 105   /// Returns true is \p ND should be collected.
 106   static bool shouldCollectSymbol(const NamedDecl &ND, const ASTContext &ASTCtx,
 107                                   const Options &Opts, bool IsMainFileSymbol);
 108
 109   // Given a ref contained in enclosing decl `Enclosing`, return
 110   // the decl that should be used as that ref's Ref::Container. This is
 111   // usually `Enclosing` itself, but in cases where `Enclosing` is not
 112   // indexed, we walk further up because Ref::Container should always be
 113   // an indexed symbol.
 114   // Note: we don't use DeclContext as the container as in some cases
 115   // it's useful to use a Decl which is not a DeclContext. For example,
 116   // for a ref occurring in the initializer of a namespace-scope variable,
 117   // it's useful to use that variable as the container, as otherwise the
 118   // next enclosing DeclContext would be a NamespaceDecl or TranslationUnitDecl,
 119   // which are both not indexed and less granular than we'd like for use cases
 120   // like call hierarchy.
 121   static const Decl *getRefContainer(const Decl *Enclosing,
 122                                      const SymbolCollector::Options &Opts);
 123
 124   void initialize(ASTContext &Ctx) override;
 125
 126   void setPreprocessor(std::shared_ptr<Preprocessor> PP) override {
 127     this->PP = PP.get();
 128   }
 129   void setPreprocessor(Preprocessor &PP) { this->PP = &PP; }
 130
 131   bool
 132   handleDeclOccurrence(const Decl *D, index::SymbolRoleSet Roles,
 133                        ArrayRef<index::SymbolRelation> Relations,
 134                        SourceLocation Loc,
 135                        index::IndexDataConsumer::ASTNodeInfo ASTNode) override;
 136
 137   bool handleMacroOccurrence(const IdentifierInfo *Name, const MacroInfo *MI,
 138                              index::SymbolRoleSet Roles,
 139                              SourceLocation Loc) override;
 140
 141   void handleMacros(const MainFileMacros &MacroRefsToIndex);
 142
 143   SymbolSlab takeSymbols() { return std::move(Symbols).build(); }
 144   RefSlab takeRefs() { return std::move(Refs).build(); }
 145   RelationSlab takeRelations() { return std::move(Relations).build(); }
 146
 147   /// Returns true if we are interested in references and declarations from \p
 148   /// FID. If this function return false, bodies of functions inside those files
 149   /// will be skipped to decrease indexing time.
 150   bool shouldIndexFile(FileID FID);
 151
 152   void finish() override;
 153
 154 private:
 155   const Symbol *addDeclaration(const NamedDecl &, SymbolID,
 156                                bool IsMainFileSymbol);
 157   void addDefinition(const NamedDecl &, const Symbol &DeclSymbol);
 158   void processRelations(const NamedDecl &ND, const SymbolID &ID,
 159                         ArrayRef<index::SymbolRelation> Relations);
 160
 161   std::optional<SymbolLocation> getTokenLocation(SourceLocation TokLoc);
 162
 163   std::optional<std::string> getIncludeHeader(const Symbol &S, FileID);
 164
 165   SymbolID getSymbolIDCached(const Decl *D);
 166   SymbolID getSymbolIDCached(const llvm::StringRef MacroName,
 167                              const MacroInfo *MI, const SourceManager &SM);
 168
 169   // All Symbols collected from the AST.
 170   SymbolSlab::Builder Symbols;
 171   // File IDs used to determine if the code contains Obj-C constructs.
 172   // For Obj-C symbols, these File IDs are used to compute the include
 173   // headers.
 174   llvm::DenseMap<SymbolID, FileID> IncludeFiles;
 175   void setIncludeLocation(const Symbol &S, SourceLocation,
 176                           const include_cleaner::Symbol &Sym);
 177
 178   // Providers for Symbol.IncludeHeaders.
 179   // The final spelling is calculated in finish().
 180   llvm::DenseMap<SymbolID, std::optional<include_cleaner::Header>>
 181       SymbolProviders;
 182   // Files which contain ObjC symbols.
 183   // This is finalized and used in finish().
 184   llvm::DenseSet<FileID> FilesWithObjCConstructs;
 185
 186   // Indexed macros, to be erased if they turned out to be include guards.
 187   llvm::DenseSet<const IdentifierInfo *> IndexedMacros;
 188   // All refs collected from the AST. It includes:
 189   //   1) symbols declared in the preamble and referenced from the main file (
 190   //     which is not a header), or
 191   //   2) symbols declared and referenced from the main file (which is a header)
 192   RefSlab::Builder Refs;
 193   // All relations collected from the AST.
 194   RelationSlab::Builder Relations;
 195   ASTContext *ASTCtx;
 196   Preprocessor *PP = nullptr;
 197   std::shared_ptr<GlobalCodeCompletionAllocator> CompletionAllocator;
 198   std::unique_ptr<CodeCompletionTUInfo> CompletionTUInfo;
 199   Options Opts;
 200   struct SymbolRef {
 201     SourceLocation Loc;
 202     FileID FID;
 203     index::SymbolRoleSet Roles;
 204     const Decl *Container;
 205     bool Spelled;
 206   };
 207   void addRef(SymbolID ID, const SymbolRef &SR);
 208   // Symbols referenced from the current TU, flushed on finish().
 209   llvm::DenseSet<SymbolID> ReferencedSymbols;
 210   // Maps canonical declaration provided by clang to canonical declaration for
 211   // an index symbol, if clangd prefers a different declaration than that
 212   // provided by clang. For example, friend declaration might be considered
 213   // canonical by clang but should not be considered canonical in the index
 214   // unless it's a definition.
 215   llvm::DenseMap<const Decl *, const Decl *> CanonicalDecls;
 216   // Cache whether to index a file or not.
 217   llvm::DenseMap<FileID, bool> FilesToIndexCache;
 218   // Encapsulates calculations and caches around header paths, which headers
 219   // to insert for which symbol, etc.
 220   class HeaderFileURICache;
 221   std::unique_ptr<HeaderFileURICache> HeaderFileURIs;
 222   llvm::DenseMap<const Decl *, SymbolID> DeclToIDCache;
 223   llvm::DenseMap<const MacroInfo *, SymbolID> MacroToIDCache;
 224 };
 225
 226 } // namespace clangd
 227 } // namespace clang
 228
 229 #endif