path: root/lib/Format/Format.cpp

//===--- Format.cpp - Format C++ code -------------------------------------===//
//
//                     The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
///
/// \file
/// \brief This file implements functions declared in Format.h. This will be
/// split into separate files as we go.
///
/// This is EXPERIMENTAL code under heavy development. It is not in a state yet,
/// where it can be used to format real code.
///
//===----------------------------------------------------------------------===//

#include "clang/Format/Format.h"
#include "UnwrappedLineParser.h"
#include "clang/Basic/SourceManager.h"
#include "clang/Basic/OperatorPrecedence.h"
#include "clang/Lex/Lexer.h"

#include <string>

namespace clang {
namespace format {

// FIXME: Move somewhere sane.
struct TokenAnnotation {
  enum TokenType {
    TT_Unknown,
    TT_TemplateOpener,
    TT_TemplateCloser,
    TT_BinaryOperator,
    TT_UnaryOperator,
    TT_TrailingUnaryOperator,
    TT_OverloadedOperator,
    TT_PointerOrReference,
    TT_ConditionalExpr,
    TT_CtorInitializerColon,
    TT_LineComment,
    TT_BlockComment,
    TT_DirectorySeparator,
    TT_ObjCMethodSpecifier
  };

  TokenType Type;

  bool SpaceRequiredBefore;
  bool CanBreakBefore;
  bool MustBreakBefore;
};

using llvm::MutableArrayRef;

FormatStyle getLLVMStyle() {
  FormatStyle LLVMStyle;
  LLVMStyle.ColumnLimit = 80;
  LLVMStyle.MaxEmptyLinesToKeep = 1;
  LLVMStyle.PointerAndReferenceBindToType = false;
  LLVMStyle.AccessModifierOffset = -2;
  LLVMStyle.SplitTemplateClosingGreater = true;
  LLVMStyle.IndentCaseLabels = false;
  return LLVMStyle;
}

FormatStyle getGoogleStyle() {
  FormatStyle GoogleStyle;
  GoogleStyle.ColumnLimit = 80;
  GoogleStyle.MaxEmptyLinesToKeep = 1;
  GoogleStyle.PointerAndReferenceBindToType = true;
  GoogleStyle.AccessModifierOffset = -1;
  GoogleStyle.SplitTemplateClosingGreater = false;
  GoogleStyle.IndentCaseLabels = true;
  return GoogleStyle;
}

struct OptimizationParameters {
  unsigned PenaltyIndentLevel;
};

class UnwrappedLineFormatter {
public:
  UnwrappedLineFormatter(const FormatStyle &Style, SourceManager &SourceMgr,
                         const UnwrappedLine &Line,
                         const std::vector<TokenAnnotation> &Annotations,
                         tooling::Replacements &Replaces, bool StructuralError)
      : Style(Style), SourceMgr(SourceMgr), Line(Line),
        Annotations(Annotations), Replaces(Replaces),
        StructuralError(StructuralError) {
    Parameters.PenaltyIndentLevel = 5;
  }

  void format() {
    // Format first token and initialize indent.
    unsigned Indent = formatFirstToken();

    // Initialize state dependent on indent.
    IndentState State;
    State.Column = Indent;
    State.ConsumedTokens = 0;
    State.Indent.push_back(Indent + 4);
    State.LastSpace.push_back(Indent);
    State.FirstLessLess.push_back(0);
    State.ForLoopVariablePos = 0;
    State.LineContainsContinuedForLoopSection = false;

    // The first token has already been indented and thus consumed.
    moveStateToNextToken(State);

    // Check whether the UnwrappedLine can be put onto a single line. If so,
    // this is bound to be the optimal solution (by definition) and we don't
    // need to analyze the entire solution space. 
    unsigned Columns = State.Column;
    bool FitsOnALine = true;
    for (unsigned i = 1, n = Line.Tokens.size(); i != n; ++i) {
      Columns += (Annotations[i].SpaceRequiredBefore ? 1 : 0) +
          Line.Tokens[i].Tok.getLength();
      // A special case for the colon of a constructor initializer as this only
      // needs to be put on a new line if the line needs to be split.
      if (Columns > Style.ColumnLimit ||
          (Annotations[i].MustBreakBefore &&
           Annotations[i].Type != TokenAnnotation::TT_CtorInitializerColon)) {
        FitsOnALine = false;
        break;
      }
    }

    // Start iterating at 1 as we have correctly formatted of Token #0 above.
    for (unsigned i = 1, n = Line.Tokens.size(); i != n; ++i) {
      if (FitsOnALine) {
        addTokenToState(false, false, State);
      } else {
        unsigned NoBreak = calcPenalty(State, false, UINT_MAX);
        unsigned Break = calcPenalty(State, true, NoBreak);
        addTokenToState(Break < NoBreak, false, State);
      }
    }
  }

private:
  /// \brief The current state when indenting a unwrapped line.
  ///
  /// As the indenting tries different combinations this is copied by value.
  struct IndentState {
    /// \brief The number of used columns in the current line.
    unsigned Column;

    /// \brief The number of tokens already consumed.
    unsigned ConsumedTokens;

    /// \brief The position to which a specific parenthesis level needs to be
    /// indented.
    std::vector<unsigned&g