Implement clang tool that converts std::string("") to std::string().

tools/clang/empty_string/EmptyStringConverter.cpp

Index: tools/clang/empty_string/EmptyStringConverter.cpp
diff --git a/tools/clang/empty_string/EmptyStringConverter.cpp b/tools/clang/empty_string/EmptyStringConverter.cpp
new file mode 100644
index 0000000000000000000000000000000000000000..526b20af3eda14efef7780669fe1d9bd28bbd0cd
--- /dev/null
+++ b/tools/clang/empty_string/EmptyStringConverter.cpp
@@ -0,0 +1,162 @@
+// Copyright (c) 2013 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#include "clang/ASTMatchers/ASTMatchers.h"
+#include "clang/ASTMatchers/ASTMatchFinder.h"
+#include "clang/Basic/SourceManager.h"
+#include "clang/Frontend/FrontendActions.h"
+#include "clang/Tooling/CommonOptionsParser.h"
+#include "clang/Tooling/Refactoring.h"
+#include "clang/Tooling/Tooling.h"
+#include "llvm/Support/CommandLine.h"
+
+using clang::ast_matchers::MatchFinder;
+using clang::ast_matchers::argumentCountIs;
+using clang::ast_matchers::bindTemporaryExpr;
+using clang::ast_matchers::constructorDecl;
+using clang::ast_matchers::constructExpr;
+using clang::ast_matchers::defaultArgExpr;
+using clang::ast_matchers::expr;
+using clang::ast_matchers::forEach;
+using clang::ast_matchers::has;
+using clang::ast_matchers::hasArgument;
+using clang::ast_matchers::hasDeclaration;
+using clang::ast_matchers::hasName;
+using clang::ast_matchers::id;
+using clang::ast_matchers::methodDecl;
+using clang::ast_matchers::newExpr;
+using clang::ast_matchers::ofClass;
+using clang::ast_matchers::stringLiteral;
+using clang::ast_matchers::varDecl;
+using clang::tooling::CommonOptionsParser;
+using clang::tooling::Replacement;
+using clang::tooling::Replacements;
+
+namespace {
+
+enum ReplacementMode {
+  kConstructorMode,
+  kInitializerMode,
+  kTemporaryMode,
+};
+
+class EmptyStringConverterCallback : public MatchFinder::MatchCallback {
+ public:
+  EmptyStringConverterCallback(ReplacementMode mode, Replacements* replacements)
+      : mode_(mode), replacements_(replacements) {}
+
+  virtual void run(const MatchFinder::MatchResult& result);
+
+ private:
+  const ReplacementMode mode_;
+  Replacements* const replacements_;
+};
+
+class EmptyStringConverter {
+ public:
+  explicit EmptyStringConverter(Replacements* replacements)
+      : constructor_callback_(kConstructorMode, replacements),
+        initializer_callback_(kInitializerMode, replacements),
+        temporary_callback_(kTemporaryMode, replacements) {}
+
+  void SetupMatchers(MatchFinder* match_finder);
+
+ private:
+  EmptyStringConverterCallback constructor_callback_;
+  EmptyStringConverterCallback initializer_callback_;
+  EmptyStringConverterCallback temporary_callback_;
+};
+
+void EmptyStringConverter::SetupMatchers(MatchFinder* match_finder) {
+  const auto& constructor_call =
+      id("call",
+         constructExpr(
+             hasDeclaration(methodDecl(ofClass(hasName("std::basic_string")))),
+             argumentCountIs(2),
+             hasArgument(0, id("literal", stringLiteral())),
+             hasArgument(1, defaultArgExpr())));
+
+  // Note that expr(has()) in the matcher is significant; the Clang AST wraps
+  // calls to the std::string constructor with exprWithCleanups nodes. Without
+  // the expr(has()) matcher, the first and last rules would not match anything!
+  match_finder->addMatcher(varDecl(forEach(expr(has(constructor_call)))),
+                           &constructor_callback_);
+  match_finder->addMatcher(newExpr(has(constructor_call)),
+                           &constructor_callback_);
+  match_finder->addMatcher(bindTemporaryExpr(has(constructor_call)),
+                           &temporary_callback_);
+  match_finder->addMatcher(
+      constructorDecl(forEach(expr(has(constructor_call)))),
+      &initializer_callback_);
+}
+
+void EmptyStringConverterCallback::run(const MatchFinder::MatchResult& result) {
+  const clang::StringLiteral* literal =
+      result.Nodes.getNodeAs<clang::StringLiteral>("literal");
+  if (literal->getLength() > 0)
+    return;
+
+  const clang::CXXConstructExpr* call =
+      result.Nodes.getNodeAs<clang::CXXConstructExpr>("call");
+
+  switch (mode_) {
+    case kConstructorMode: {
+      clang::CharSourceRange range =
+          clang::CharSourceRange::getTokenRange(call->getParenRange());
+      replacements_->insert(Replacement(*result.SourceManager, range, ""));
+      break;
+    }
+    case kInitializerMode: {
+      replacements_->insert(Replacement(*result.SourceManager, call, ""));
+      break;
+    }
+    case kTemporaryMode: {
+      // Differentiate between explicit and implicit calls to std::string's
+      // constructor. An implicitly generated constructor won't have a valid
+      // source range for the parenthesis.
+      clang::SourceRange range = call->getParenRange();
+      if (range.isValid()) {
+        replacements_->insert(Replacement(*result.SourceManager, literal, ""));
+      } else {
+        replacements_->insert(
+            Replacement(*result.SourceManager, call, "std::string()"));
+      }
+      break;
+    }
+  }
+}
+
+}  // namespace
+
+static llvm::cl::extrahelp common_help(CommonOptionsParser::HelpMessage);
+
+int main(int argc, const char* argv[]) {
+  CommonOptionsParser options(argc, argv);
+
+  clang::tooling::ClangTool tool(options.getCompilations(),
+                                 options.getSourcePathList());
+
+  Replacements replacements;
+  EmptyStringConverter converter(&replacements);
+  MatchFinder match_finder;
+  converter.SetupMatchers(&match_finder);
+
+  int result =
+      tool.run(clang::tooling::newFrontendActionFactory(&match_finder));
+  if (result != 0)
+    return result;
+
+  // Each replacement line should have the following format:
+  // r:<file path>:<offset>:<length>:<replacement text>
+  // Only the <replacement text> field can contain embedded ":" characters.
+  // TODO(dcheng): Use a more clever serialization.
+  llvm::outs() << "==== BEGIN EDITS ====\n";
+  for (const Replacement& r : replacements) {
+    llvm::outs() << "r:" << r.getFilePath() << ":" << r.getOffset() << ":"
+                 << r.getLength() << ":" << r.getReplacementText() << "\n";
+  }
+  llvm::outs() << "==== END EDITS ====\n";
+
+  return 0;
+}