| 1 | //===--- FasterStrsplitDelimiterCheck.cpp - clang-tidy---------------------===// |
| 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 | #include "FasterStrsplitDelimiterCheck.h" |
| 10 | #include "clang/AST/ASTContext.h" |
| 11 | #include "clang/ASTMatchers/ASTMatchFinder.h" |
| 12 | #include "clang/Tooling/FixIt.h" |
| 13 | #include <optional> |
| 14 | |
| 15 | using namespace clang::ast_matchers; |
| 16 | |
| 17 | namespace clang::tidy::abseil { |
| 18 | |
| 19 | namespace { |
| 20 | |
| 21 | AST_MATCHER(StringLiteral, lengthIsOne) { return Node.getLength() == 1; } |
| 22 | |
| 23 | std::optional<std::string> makeCharacterLiteral(const StringLiteral *Literal, |
| 24 | const ASTContext &Context) { |
| 25 | assert(Literal->getLength() == 1 && |
| 26 | "Only single character string should be matched" ); |
| 27 | assert(Literal->getCharByteWidth() == 1 && |
| 28 | "StrSplit doesn't support wide char" ); |
| 29 | std::string Result = clang::tooling::fixit::getText(Node: *Literal, Context).str(); |
| 30 | bool IsRawStringLiteral = StringRef(Result).starts_with(Prefix: R"(R")" ); |
| 31 | // Since raw string literal might contain unescaped non-printable characters, |
| 32 | // we normalize them using `StringLiteral::outputString`. |
| 33 | if (IsRawStringLiteral) { |
| 34 | Result.clear(); |
| 35 | llvm::raw_string_ostream Stream(Result); |
| 36 | Literal->outputString(OS&: Stream); |
| 37 | } |
| 38 | // Special case: If the string contains a single quote, we just need to return |
| 39 | // a character of the single quote. This is a special case because we need to |
| 40 | // escape it in the character literal. |
| 41 | if (Result == R"("'")" ) |
| 42 | return std::string(R"('\'')" ); |
| 43 | |
| 44 | // Now replace the " with '. |
| 45 | std::string::size_type Pos = Result.find_first_of(c: '"'); |
| 46 | if (Pos == std::string::npos) |
| 47 | return std::nullopt; |
| 48 | Result[Pos] = '\''; |
| 49 | Pos = Result.find_last_of(c: '"'); |
| 50 | if (Pos == std::string::npos) |
| 51 | return std::nullopt; |
| 52 | Result[Pos] = '\''; |
| 53 | return Result; |
| 54 | } |
| 55 | |
| 56 | } // anonymous namespace |
| 57 | |
| 58 | void FasterStrsplitDelimiterCheck::registerMatchers(MatchFinder *Finder) { |
| 59 | // Binds to one character string literals. |
| 60 | const auto SingleChar = |
| 61 | expr(ignoringParenCasts(InnerMatcher: stringLiteral(lengthIsOne()).bind(ID: "Literal" ))); |
| 62 | |
| 63 | // Binds to a string_view (either absl or std) that was passed by value and |
| 64 | // constructed from string literal. |
| 65 | auto StringViewArg = ignoringElidableConstructorCall(InnerMatcher: ignoringImpCasts( |
| 66 | InnerMatcher: cxxConstructExpr(hasType(InnerMatcher: recordDecl(hasName(Name: "::absl::string_view" ))), |
| 67 | hasArgument(N: 0, InnerMatcher: ignoringParenImpCasts(InnerMatcher: SingleChar))))); |
| 68 | |
| 69 | // Need to ignore the elidable constructor as otherwise there is no match for |
| 70 | // c++14 and earlier. |
| 71 | auto ByAnyCharArg = |
| 72 | expr(has(ignoringElidableConstructorCall( |
| 73 | InnerMatcher: ignoringParenCasts(InnerMatcher: cxxBindTemporaryExpr(has(cxxConstructExpr( |
| 74 | hasType(InnerMatcher: recordDecl(hasName(Name: "::absl::ByAnyChar" ))), |
| 75 | hasArgument(N: 0, InnerMatcher: StringViewArg)))))))) |
| 76 | .bind(ID: "ByAnyChar" ); |
| 77 | |
| 78 | // Find uses of absl::StrSplit(..., "x") and absl::StrSplit(..., |
| 79 | // absl::ByAnyChar("x")) to transform them into absl::StrSplit(..., 'x'). |
| 80 | Finder->addMatcher( |
| 81 | NodeMatch: traverse(TK: TK_AsIs, |
| 82 | InnerMatcher: callExpr(callee(InnerMatcher: functionDecl(hasName(Name: "::absl::StrSplit" ))), |
| 83 | hasArgument(N: 1, InnerMatcher: anyOf(ByAnyCharArg, SingleChar)), |
| 84 | unless(isInTemplateInstantiation())) |
| 85 | .bind(ID: "StrSplit" )), |
| 86 | Action: this); |
| 87 | |
| 88 | // Find uses of absl::MaxSplits("x", N) and |
| 89 | // absl::MaxSplits(absl::ByAnyChar("x"), N) to transform them into |
| 90 | // absl::MaxSplits('x', N). |
| 91 | Finder->addMatcher( |
| 92 | NodeMatch: traverse(TK: TK_AsIs, |
| 93 | InnerMatcher: callExpr(callee(InnerMatcher: functionDecl(hasName(Name: "::absl::MaxSplits" ))), |
| 94 | hasArgument(N: 0, InnerMatcher: anyOf(ByAnyCharArg, |
| 95 | ignoringParenCasts(InnerMatcher: SingleChar))), |
| 96 | unless(isInTemplateInstantiation()))), |
| 97 | Action: this); |
| 98 | } |
| 99 | |
| 100 | void FasterStrsplitDelimiterCheck::check( |
| 101 | const MatchFinder::MatchResult &Result) { |
| 102 | const auto *Literal = Result.Nodes.getNodeAs<StringLiteral>(ID: "Literal" ); |
| 103 | |
| 104 | if (Literal->getBeginLoc().isMacroID() || Literal->getEndLoc().isMacroID()) |
| 105 | return; |
| 106 | |
| 107 | std::optional<std::string> Replacement = |
| 108 | makeCharacterLiteral(Literal, Context: *Result.Context); |
| 109 | if (!Replacement) |
| 110 | return; |
| 111 | SourceRange Range = Literal->getSourceRange(); |
| 112 | |
| 113 | if (const auto *ByAnyChar = Result.Nodes.getNodeAs<Expr>(ID: "ByAnyChar" )) |
| 114 | Range = ByAnyChar->getSourceRange(); |
| 115 | |
| 116 | diag( |
| 117 | Loc: Literal->getBeginLoc(), |
| 118 | Description: "%select{absl::StrSplit()|absl::MaxSplits()}0 called with a string " |
| 119 | "literal " |
| 120 | "consisting of a single character; consider using the character overload" ) |
| 121 | << (Result.Nodes.getNodeAs<CallExpr>(ID: "StrSplit" ) ? 0 : 1) |
| 122 | << FixItHint::CreateReplacement(RemoveRange: CharSourceRange::getTokenRange(R: Range), |
| 123 | Code: *Replacement); |
| 124 | } |
| 125 | |
| 126 | } // namespace clang::tidy::abseil |
| 127 | |