1 //===--- UnicodeCharRanges.h - Types and functions for character ranges ---===// 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_SUPPORT_UNICODECHARRANGES_H 9 #define LLVM_SUPPORT_UNICODECHARRANGES_H 10 11 #include "llvm/ADT/ArrayRef.h" 12 #include "llvm/ADT/SmallPtrSet.h" 13 #include "llvm/Support/Compiler.h" 14 #include "llvm/Support/Debug.h" 15 #include "llvm/Support/Mutex.h" 16 #include "llvm/Support/MutexGuard.h" 17 #include "llvm/Support/raw_ostream.h" 18 #include <algorithm> 19 20 #define DEBUG_TYPE "unicode" 21 22 namespace llvm { 23 namespace sys { 24 25 /// Represents a closed range of Unicode code points [Lower, Upper]. 26 struct UnicodeCharRange { 27 uint32_t Lower; 28 uint32_t Upper; 29 }; 30 31 inline bool operator<(uint32_t Value, UnicodeCharRange Range) { 32 return Value < Range.Lower; 33 } 34 inline bool operator<(UnicodeCharRange Range, uint32_t Value) { 35 return Range.Upper < Value; 36 } 37 38 /// Holds a reference to an ordered array of UnicodeCharRange and allows 39 /// to quickly check if a code point is contained in the set represented by this 40 /// array. 41 class UnicodeCharSet { 42 public: 43 typedef ArrayRef<UnicodeCharRange> CharRanges; 44 45 /// Constructs a UnicodeCharSet instance from an array of 46 /// UnicodeCharRanges. 47 /// 48 /// Array pointed by \p Ranges should have the lifetime at least as long as 49 /// the UnicodeCharSet instance, and should not change. Array is validated by 50 /// the constructor, so it makes sense to create as few UnicodeCharSet 51 /// instances per each array of ranges, as possible. 52 #ifdef NDEBUG 53 54 // FIXME: This could use constexpr + static_assert. This way we 55 // may get rid of NDEBUG in this header. Unfortunately there are some 56 // problems to get this working with MSVC 2013. Change this when 57 // the support for MSVC 2013 is dropped. UnicodeCharSet(CharRanges Ranges)58 constexpr UnicodeCharSet(CharRanges Ranges) : Ranges(Ranges) {} 59 #else UnicodeCharSet(CharRanges Ranges)60 UnicodeCharSet(CharRanges Ranges) : Ranges(Ranges) { 61 assert(rangesAreValid()); 62 } 63 #endif 64 65 /// Returns true if the character set contains the Unicode code point 66 /// \p C. contains(uint32_t C)67 bool contains(uint32_t C) const { 68 return std::binary_search(Ranges.begin(), Ranges.end(), C); 69 } 70 71 private: 72 /// Returns true if each of the ranges is a proper closed range 73 /// [min, max], and if the ranges themselves are ordered and non-overlapping. rangesAreValid()74 bool rangesAreValid() const { 75 uint32_t Prev = 0; 76 for (CharRanges::const_iterator I = Ranges.begin(), E = Ranges.end(); 77 I != E; ++I) { 78 if (I != Ranges.begin() && Prev >= I->Lower) { 79 LLVM_DEBUG(dbgs() << "Upper bound 0x"); 80 LLVM_DEBUG(dbgs().write_hex(Prev)); 81 LLVM_DEBUG(dbgs() << " should be less than succeeding lower bound 0x"); 82 LLVM_DEBUG(dbgs().write_hex(I->Lower) << "\n"); 83 return false; 84 } 85 if (I->Upper < I->Lower) { 86 LLVM_DEBUG(dbgs() << "Upper bound 0x"); 87 LLVM_DEBUG(dbgs().write_hex(I->Lower)); 88 LLVM_DEBUG(dbgs() << " should not be less than lower bound 0x"); 89 LLVM_DEBUG(dbgs().write_hex(I->Upper) << "\n"); 90 return false; 91 } 92 Prev = I->Upper; 93 } 94 95 return true; 96 } 97 98 const CharRanges Ranges; 99 }; 100 101 } // namespace sys 102 } // namespace llvm 103 104 #undef DEBUG_TYPE // "unicode" 105 106 #endif // LLVM_SUPPORT_UNICODECHARRANGES_H 107