1 //===--- SymbolID.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
9 #ifndef LLVM_CLANG_TOOLS_EXTRA_CLANGD_INDEX_SYMBOLID_H
10 #define LLVM_CLANG_TOOLS_EXTRA_CLANGD_INDEX_SYMBOLID_H
11
12 #include "llvm/ADT/Hashing.h"
13 #include "llvm/ADT/StringRef.h"
14 #include "llvm/Support/Error.h"
15 #include "llvm/Support/raw_ostream.h"
16 #include <array>
17 #include <cstddef>
18 #include <cstdint>
19 #include <string>
20
21 namespace clang {
22 namespace clangd {
23
24 // The class identifies a particular C++ symbol (class, function, method, etc).
25 //
26 // As USRs (Unified Symbol Resolution) could be large, especially for functions
27 // with long type arguments, SymbolID is using truncated SHA1(USR) values to
28 // guarantee the uniqueness of symbols while using a relatively small amount of
29 // memory (vs storing USRs directly).
30 //
31 // SymbolID can be used as key in the symbol indexes to lookup the symbol.
32 class SymbolID {
33 public:
34 SymbolID() = default;
35 explicit SymbolID(llvm::StringRef USR);
36
37 bool operator==(const SymbolID &Sym) const {
38 return HashValue == Sym.HashValue;
39 }
40 bool operator!=(const SymbolID &Sym) const { return !(*this == Sym); }
41 bool operator<(const SymbolID &Sym) const {
42 // Avoid lexicographic compare which requires swapping bytes or even memcmp!
43 return llvm::bit_cast<IntTy>(HashValue) <
44 llvm::bit_cast<IntTy>(Sym.HashValue);
45 }
46
47 // The stored hash is truncated to RawSize bytes.
48 // This trades off memory against the number of symbols we can handle.
49 constexpr static size_t RawSize = 8;
50 llvm::StringRef raw() const;
51 static SymbolID fromRaw(llvm::StringRef);
52
53 // Returns a hex encoded string.
54 std::string str() const;
55 static llvm::Expected<SymbolID> fromStr(llvm::StringRef);
56
isNull()57 bool isNull() const { return *this == SymbolID(); }
58 explicit operator bool() const { return !isNull(); }
59
60 private:
61 using IntTy = uint64_t;
62 static_assert(sizeof(IntTy) == RawSize);
63 std::array<uint8_t, RawSize> HashValue{};
64 };
65
hash_value(const SymbolID & ID)66 inline llvm::hash_code hash_value(const SymbolID &ID) {
67 // We already have a good hash, just return the first bytes.
68 static_assert(sizeof(size_t) <= SymbolID::RawSize,
69 "size_t longer than SHA1!");
70 size_t Result;
71 memcpy(&Result, ID.raw().data(), sizeof(size_t));
72 return llvm::hash_code(Result);
73 }
74
75 // Write SymbolID into the given stream. SymbolID is encoded as ID.str().
76 llvm::raw_ostream &operator<<(llvm::raw_ostream &OS, const SymbolID &ID);
77
78 } // namespace clangd
79 } // namespace clang
80
81 namespace llvm {
82 // Support SymbolIDs as DenseMap keys.
83 template <> struct DenseMapInfo<clang::clangd::SymbolID> {
84 static inline clang::clangd::SymbolID getEmptyKey() {
85 static clang::clangd::SymbolID EmptyKey("EMPTYKEY");
86 return EmptyKey;
87 }
88 static inline clang::clangd::SymbolID getTombstoneKey() {
89 static clang::clangd::SymbolID TombstoneKey("TOMBSTONEKEY");
90 return TombstoneKey;
91 }
92 static unsigned getHashValue(const clang::clangd::SymbolID &Sym) {
93 return hash_value(Sym);
94 }
95 static bool isEqual(const clang::clangd::SymbolID &LHS,
96 const clang::clangd::SymbolID &RHS) {
97 return LHS == RHS;
98 }
99 };
100 } // namespace llvm
101
102 #endif // LLVM_CLANG_TOOLS_EXTRA_CLANGD_INDEX_SYMBOLID_H
103