C minifier with Clang

C minifier with Clang
2022-10-9 15:0:0 Author: maskray.me(查看原文) 阅读量:19 收藏
I recently revamped Competitive programming in Nim. In short, I can create a C amalgamation from a Nim program and submit the C source code to various competitive programming websites.
clang-format safely removes whitespace and makes the program smaller.
Then I use a Clang based tool to shorten function, variables, and type names.
#include <clang/AST/ASTConsumer.h>
#include <clang/AST/Decl.h>
#include <clang/AST/RecursiveASTVisitor.h>
#include <clang/Basic/FileManager.h>
#include <clang/Basic/LangOptions.h>
#include <clang/Basic/SourceManager.h>
#include <clang/Basic/TargetInfo.h>
#include <clang/Driver/Action.h>
#include <clang/Driver/Compilation.h>
#include <clang/Driver/Driver.h>
#include <clang/Driver/Tool.h>
#include <clang/Frontend/CompilerInstance.h>
#include <clang/Frontend/FrontendAction.h>
#include <clang/Lex/Lexer.h>
#include <clang/Lex/PreprocessorOptions.h>
#include <clang/Tooling/Core/Replacement.h>
#include <llvm/ADT/CachedHashString.h>
#include <llvm/ADT/DenseSet.h>
#include <llvm/ADT/MapVector.h>
#include <llvm/ADT/STLExtras.h>
#include <llvm/Support/Host.h>
#include <llvm/Support/Path.h>
#include <llvm/Support/raw_ostream.h>

#include <memory>
#include <vector>

#include <assert.h>
#include <err.h>
#include <unistd.h>

using namespace clang;
using namespace llvm;

namespace {
std::unique_ptr<CompilerInvocation> buildCompilerInvocation(std::vector<const char *> args) {
  IntrusiveRefCntPtr<DiagnosticsEngine> diags(
      CompilerInstance::createDiagnostics(new DiagnosticOptions, new IgnoringDiagConsumer, true));

  driver::Driver d(args[0], llvm::sys::getDefaultTargetTriple(), *diags, "cminifier", llvm::vfs::getRealFileSystem());
  d.setCheckInputsExist(false);
  std::unique_ptr<driver::Compilation> comp(d.BuildCompilation(args));
  if (!comp)
    return nullptr;
  const driver::JobList &jobs = comp->getJobs();
  if (jobs.size() != 1 || !isa<driver::Command>(*jobs.begin()))
    return nullptr;

  const driver::Command &cmd = cast<driver::Command>(*jobs.begin());
  if (StringRef(cmd.getCreator().getName()) != "clang")
    return nullptr;
  const llvm::opt::ArgStringList &cc_args = cmd.getArguments();
  auto ci = std::make_unique<CompilerInvocation>();
  if (!CompilerInvocation::CreateFromArgs(*ci, cc_args, *diags))
    return nullptr;

  ci->getDiagnosticOpts().IgnoreWarnings = true;
  ci->getFrontendOpts().DisableFree = false;
  return ci;
}

SmallVector<StringRef, 0> ignores;
MapVector<Decl *, std::string> d2name;
DenseSet<CachedHashStringRef> used;
std::string newCode;

struct Collector : RecursiveASTVisitor<Collector> {
  SourceManager &sm;

  Collector(ASTContext &ctx) : sm(ctx.getSourceManager()) {}
  bool VisitFunctionDecl(FunctionDecl *fd) {
    if (fd->isOverloadedOperator() || !fd->getIdentifier())
      return true;
    used.insert(CachedHashStringRef(fd->getName()));
    if (!fd->isDefined())
      return true;
    std::string name = fd->getNameAsString();
    if (sm.isWrittenInMainFile(fd->getLocation())) {
      if (!is_contained(ignores, name))
        d2name[fd->getCanonicalDecl()] = "_f";
      for (ParmVarDecl *param : fd->parameters())
        VisitVarDecl(param);
    }
    return true;
  }
  bool VisitVarDecl(VarDecl *vd) {
    if (!vd->getIdentifier())
      return true;
    used.insert(CachedHashStringRef(vd->getName()));
    auto kind = vd->isThisDeclarationADefinition();
    if (kind != VarDecl::Definition || !sm.isWrittenInMainFile(vd->getLocation()))
      return true;
    d2name[vd->getCanonicalDecl()] = "_v";
    return true;
  }

  bool VisitTagDecl(TagDecl *td) {
    used.insert(CachedHashStringRef(td->getName()));
    if (!td->isThisDeclarationADefinition() || !sm.isWrittenInMainFile(td->getLocation()))
      return true;
    d2name[td->getCanonicalDecl()] = "_t";
    return true;
  }
  bool VisitTypedefNameDecl(TypedefNameDecl *d) {
    if (d->isTransparentTag() || !sm.isWrittenInMainFile(d->getLocation()))
      return true;
    d2name[d->getCanonicalDecl()] = "_t";
    return true;
  }
};

struct Renamer : RecursiveASTVisitor<Renamer> {
  SourceManager &sm;
  tooling::Replacements &reps;

  Renamer(ASTContext &ctx, tooling::Replacements &reps) : sm(ctx.getSourceManager()), reps(reps) {}
  void replace(CharSourceRange csr, StringRef newText) { cantFail(reps.add(tooling::Replacement(sm, csr, newText))); }

  bool VisitFunctionDecl(FunctionDecl *fd) {
    auto *canon = fd->getCanonicalDecl();
    auto it = d2name.find(canon);
    if (it != d2name.end())
      replace(CharSourceRange::getTokenRange(fd->getLocation()), it->second);
    return true;
  }
  bool VisitVarDecl(VarDecl *vd) {
    auto *canon = vd->getCanonicalDecl();
    auto it = d2name.find(canon);
    if (it != d2name.end())
      replace(CharSourceRange::getTokenRange(vd->getLocation()), it->second);
    return true;
  }
  bool VisitDeclRefExpr(DeclRefExpr *dre) {
    Decl *d = dre->getDecl();
    if (!(isa<FunctionDecl>(d) || isa<VarDecl>(d)))
      return true;
    auto it = d2name.find(d->getCanonicalDecl());
    if (it != d2name.end())
      replace(CharSourceRange::getTokenRange(SourceRange(dre->getBeginLoc(), dre->getEndLoc())), it->second);
    return true;
  }

  bool VisitTagDecl(TagDecl *d) {
    auto *canon = d->getCanonicalDecl();
    if (auto it = d2name.find(canon); it != d2name.end())
      replace(CharSourceRange::getTokenRange(d->getLocation()), it->second);
    return true;
  }
  bool VisitTagTypeLoc(TagTypeLoc tl) {
    TagDecl *td = tl.getDecl()->getCanonicalDecl();
    if (auto it = d2name.find(td); it != d2name.end())
      replace(CharSourceRange::getTokenRange(tl.getNameLoc()), it->second);
    return true;
  }
  bool VisitTypedefNameDecl(TypedefNameDecl *d) {
    if (auto it = d2name.find(d->getCanonicalDecl()); it != d2name.end())
      replace(CharSourceRange::getTokenRange(d->getLocation()), it->second);
    return true;
  }
  bool VisitTypedefTypeLoc(TypedefTypeLoc tl) {
    TypedefNameDecl *td = tl.getTypedefNameDecl();
    if (auto it = d2name.find(td); it != d2name.end())
      replace(CharSourceRange::getTokenRange(tl.getNameLoc()), it->second);
    return true;
  }
};

struct MiniASTConsumer : ASTConsumer {
  ASTContext *ctx;
  int n_fn = 0, n_var = 0, n_type = 0;

  void Initialize(ASTContext &ctx) override { this->ctx = &ctx; }
  static std::string getName(StringRef prefix, int &id) {
    static const char digits[] = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
    std::string newName;
    for (;;) {
      newName = std::string(1, prefix[id % prefix.size()]);
      if (int i = id / prefix.size())
        while (newName += digits[i % 62], i /= 62);
      id++;
      if (!used.contains(CachedHashStringRef(newName))) break;
    }
    return newName;
  }
  bool HandleTopLevelDecl(DeclGroupRef dgr) override {
    Collector c(*ctx);
    for (Decl *d : dgr)
      c.TraverseDecl(d);
    for (auto &[d, name] : d2name) {
      if (name == "_f")
        name = getName("abcdefghij", n_fn);
      else if (name == "_v") {
        int old_n_var = n_var;
        auto newName = getName("klmnopqrstuvwxyz", n_var);
        if (newName.size() < static_cast<VarDecl *>(d)->getName().size())
          name = newName;
        else {
          name = static_cast<VarDecl *>(d)->getName();
          n_var = old_n_var;
        }
      } else if (name == "_t")
        name = getName("ABCDEFGHIJKLMNOPQRSTUVWXYZ", n_type);
    }
    return true;
  }
  void HandleTranslationUnit(ASTContext &ctx) override {
    tooling::Replacements reps;
    Renamer c(ctx, reps);
    c.TraverseDecl(ctx.getTranslationUnitDecl());

    auto &sm = ctx.getSourceManager();
    StringRef code = sm.getBufferData(sm.getMainFileID());
    auto res = tooling::applyAllReplacements(code, reps);
    if (!res)
      errx(2, "failed to apply replacements: %s", toString(res.takeError()).c_str());
    newCode = *res;
  }
};

struct MiniAction : ASTFrontendAction {
  std::unique_ptr<ASTConsumer> CreateASTConsumer(CompilerInstance &ci,
                                                 StringRef inFile) override {
    return std::make_unique<MiniASTConsumer>();
  }
};
}

int main(int argc, char *argv[]) {
  int opt = 1;
  bool inplace = false;
  for (; opt < argc; opt++) {
    if (strcmp(argv[opt], "-h") == 0) {
      printf("Usage: %s [-i] [-f fun]... a.c\n\nOptions:\n  -i      edit a.c in place\n", argv[0]);
      return 0;
    } else if (strcmp(argv[opt], "-i") == 0)
      inplace = true;
    else if (strcmp(argv[opt], "-f") == 0 && opt + 1 < argc)
      ignores.push_back(argv[++opt]);
    else
      break;
  }
  ignores.push_back("main");

  std::vector<const char *> args{argv[0], "-fsyntax-only"};
  for (int i = opt; i != argc; i++)
    args.push_back(argv[i]);
  auto ci = buildCompilerInvocation(args);
  if (!ci)
    errx(1, "failed to build CompilerInvocation");

  auto inst = std::make_unique<CompilerInstance>(std::make_shared<PCHContainerOperations>());
  IgnoringDiagConsumer dc;
  inst->setInvocation(std::move(ci));
  inst->createDiagnostics(&dc, false);
  inst->getDiagnostics().setIgnoreAllWarnings(true);
  inst->setTarget(TargetInfo::CreateTargetInfo(inst->getDiagnostics(), inst->getInvocation().TargetOpts));
  if (!inst->hasTarget())
    errx(1, "hasTarget returns false");
  inst->createFileManager(llvm::vfs::getRealFileSystem());
  inst->setSourceManager(new SourceManager(inst->getDiagnostics(), inst->getFileManager(), true));

  MiniAction action;
  if (!action.BeginSourceFile(*inst, inst->getFrontendOpts().Inputs[0]))
    errx(2, "failed to parse");
  if (Error e = action.Execute())
    errx(2, "failed to execute");
  action.EndSourceFile();
  if (inplace) {
    std::error_code ec;
    raw_fd_ostream os(inst->getFrontendOpts().Inputs[0].getFile(), ec, sys::fs::OF_None);
    os << newCode;
  } else {
    outs() << newCode;
  }
}
文章来源: https://maskray.me/blog/2022-10-09-c-minifier-with-clang
如有侵权请联系:admin#unsafe.sh