138 lines · cpp
1//===----------------------------------------------------------------------===//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-exception6//7//===----------------------------------------------------------------------===//8 9#include "MisleadingBidirectionalCheck.h"10 11#include "clang/Frontend/CompilerInstance.h"12#include "clang/Lex/Preprocessor.h"13#include "llvm/Support/ConvertUTF.h"14 15using namespace clang;16using namespace clang::tidy::misc;17 18static bool containsMisleadingBidi(StringRef Buffer,19 bool HonorLineBreaks = true) {20 const char *CurPtr = Buffer.begin();21 22 enum BidiChar {23 PS = 0x2029,24 RLO = 0x202E,25 RLE = 0x202B,26 LRO = 0x202D,27 LRE = 0x202A,28 PDF = 0x202C,29 RLI = 0x2067,30 LRI = 0x2066,31 FSI = 0x2068,32 PDI = 0x206933 };34 35 SmallVector<BidiChar> BidiContexts;36 37 // Scan each character while maintaining a stack of opened bidi context.38 // RLO/RLE/LRO/LRE all are closed by PDF while RLI LRI and FSI are closed by39 // PDI. New lines reset the context count. Extra PDF / PDI are ignored.40 //41 // Warn if we end up with an unclosed context.42 while (CurPtr < Buffer.end()) {43 const unsigned char C = *CurPtr;44 if (isASCII(C)) {45 ++CurPtr;46 const bool IsParagrapSep =47 (C == 0xA || C == 0xD || (0x1C <= C && C <= 0x1E) || C == 0x85);48 const bool IsSegmentSep = (C == 0x9 || C == 0xB || C == 0x1F);49 if (IsParagrapSep || IsSegmentSep)50 BidiContexts.clear();51 continue;52 }53 llvm::UTF32 CodePoint = 0;54 const llvm::ConversionResult Result = llvm::convertUTF8Sequence(55 (const llvm::UTF8 **)&CurPtr, (const llvm::UTF8 *)Buffer.end(),56 &CodePoint, llvm::strictConversion);57 58 // If conversion fails, utf-8 is designed so that we can just try next char.59 if (Result != llvm::conversionOK) {60 ++CurPtr;61 continue;62 }63 64 // Open a PDF context.65 if (CodePoint == RLO || CodePoint == RLE || CodePoint == LRO ||66 CodePoint == LRE)67 BidiContexts.push_back(PDF);68 // Close PDF Context.69 else if (CodePoint == PDF) {70 if (!BidiContexts.empty() && BidiContexts.back() == PDF)71 BidiContexts.pop_back();72 }73 // Open a PDI Context.74 else if (CodePoint == RLI || CodePoint == LRI || CodePoint == FSI)75 BidiContexts.push_back(PDI);76 // Close a PDI Context.77 else if (CodePoint == PDI) {78 auto R = llvm::find(llvm::reverse(BidiContexts), PDI);79 if (R != BidiContexts.rend())80 BidiContexts.resize(BidiContexts.rend() - R - 1);81 }82 // Line break or equivalent83 else if (CodePoint == PS)84 BidiContexts.clear();85 }86 return !BidiContexts.empty();87}88 89class MisleadingBidirectionalCheck::MisleadingBidirectionalHandler90 : public CommentHandler {91public:92 MisleadingBidirectionalHandler(MisleadingBidirectionalCheck &Check)93 : Check(Check) {}94 95 bool HandleComment(Preprocessor &PP, SourceRange Range) override {96 // FIXME: check that we are in a /* */ comment97 const StringRef Text =98 Lexer::getSourceText(CharSourceRange::getCharRange(Range),99 PP.getSourceManager(), PP.getLangOpts());100 101 if (containsMisleadingBidi(Text, true))102 Check.diag(103 Range.getBegin(),104 "comment contains misleading bidirectional Unicode characters");105 return false;106 }107 108private:109 MisleadingBidirectionalCheck &Check;110};111 112MisleadingBidirectionalCheck::MisleadingBidirectionalCheck(113 StringRef Name, ClangTidyContext *Context)114 : ClangTidyCheck(Name, Context),115 Handler(std::make_unique<MisleadingBidirectionalHandler>(*this)) {}116 117MisleadingBidirectionalCheck::~MisleadingBidirectionalCheck() = default;118 119void MisleadingBidirectionalCheck::registerPPCallbacks(120 const SourceManager &SM, Preprocessor *PP, Preprocessor *ModuleExpanderPP) {121 PP->addCommentHandler(Handler.get());122}123 124void MisleadingBidirectionalCheck::check(125 const ast_matchers::MatchFinder::MatchResult &Result) {126 if (const auto *SL = Result.Nodes.getNodeAs<StringLiteral>("strlit")) {127 const StringRef Literal = SL->getBytes();128 if (containsMisleadingBidi(Literal, false))129 diag(SL->getBeginLoc(), "string literal contains misleading "130 "bidirectional Unicode characters");131 }132}133 134void MisleadingBidirectionalCheck::registerMatchers(135 ast_matchers::MatchFinder *Finder) {136 Finder->addMatcher(ast_matchers::stringLiteral().bind("strlit"), this);137}138