clang  8.0.0svn
TokenAnnotator.h
Go to the documentation of this file.
1 //===--- TokenAnnotator.h - Format C++ code ---------------------*- C++ -*-===//
2 //
3 // The LLVM Compiler Infrastructure
4 //
5 // This file is distributed under the University of Illinois Open Source
6 // License. See LICENSE.TXT for details.
7 //
8 //===----------------------------------------------------------------------===//
9 ///
10 /// \file
11 /// This file implements a token annotator, i.e. creates
12 /// \c AnnotatedTokens out of \c FormatTokens with required extra information.
13 ///
14 //===----------------------------------------------------------------------===//
15 
16 #ifndef LLVM_CLANG_LIB_FORMAT_TOKENANNOTATOR_H
17 #define LLVM_CLANG_LIB_FORMAT_TOKENANNOTATOR_H
18 
19 #include "UnwrappedLineParser.h"
20 #include "clang/Format/Format.h"
21 
22 namespace clang {
23 class SourceManager;
24 
25 namespace format {
26 
27 enum LineType {
30  LT_ObjCDecl, // An @interface, @implementation, or @protocol line.
32  LT_ObjCProperty, // An @property line.
36 };
37 
39 public:
41  : First(Line.Tokens.front().Tok), Level(Line.Level),
49  assert(!Line.Tokens.empty());
50 
51  // Calculate Next and Previous for all tokens. Note that we must overwrite
52  // Next and Previous for every token, as previous formatting runs might have
53  // left them in a different state.
54  First->Previous = nullptr;
55  FormatToken *Current = First;
56  for (std::list<UnwrappedLineNode>::const_iterator I = ++Line.Tokens.begin(),
57  E = Line.Tokens.end();
58  I != E; ++I) {
59  const UnwrappedLineNode &Node = *I;
60  Current->Next = I->Tok;
61  I->Tok->Previous = Current;
62  Current = Current->Next;
63  Current->Children.clear();
64  for (const auto &Child : Node.Children) {
65  Children.push_back(new AnnotatedLine(Child));
66  Current->Children.push_back(Children.back());
67  }
68  }
69  Last = Current;
70  Last->Next = nullptr;
71  }
72 
74  for (unsigned i = 0, e = Children.size(); i != e; ++i) {
75  delete Children[i];
76  }
77  FormatToken *Current = First;
78  while (Current) {
79  Current->Children.clear();
80  Current->Role.reset();
81  Current = Current->Next;
82  }
83  }
84 
85  /// \c true if this line starts with the given tokens in order, ignoring
86  /// comments.
87  template <typename... Ts> bool startsWith(Ts... Tokens) const {
88  return First && First->startsSequence(Tokens...);
89  }
90 
91  /// \c true if this line ends with the given tokens in reversed order,
92  /// ignoring comments.
93  /// For example, given tokens [T1, T2, T3, ...], the function returns true if
94  /// this line is like "... T3 T2 T1".
95  template <typename... Ts> bool endsWith(Ts... Tokens) const {
96  return Last && Last->endsSequence(Tokens...);
97  }
98 
99  /// \c true if this line looks like a function definition instead of a
100  /// function declaration. Asserts MightBeFunctionDecl.
102  assert(MightBeFunctionDecl);
103  // FIXME: Line.Last points to other characters than tok::semi
104  // and tok::lbrace.
105  return !Last->isOneOf(tok::semi, tok::comment);
106  }
107 
108  /// \c true if this line starts a namespace definition.
109  bool startsWithNamespace() const {
110  return startsWith(tok::kw_namespace) ||
111  startsWith(tok::kw_inline, tok::kw_namespace) ||
112  startsWith(tok::kw_export, tok::kw_namespace);
113  }
114 
117 
119 
121  unsigned Level;
128 
129  /// \c True if this line should be formatted, i.e. intersects directly or
130  /// indirectly with one of the input ranges.
131  bool Affected;
132 
133  /// \c True if the leading empty lines of this line intersect with one of the
134  /// input ranges.
136 
137  /// \c True if one of this line's children intersects with an input range.
139 
141 
142 private:
143  // Disallow copying.
144  AnnotatedLine(const AnnotatedLine &) = delete;
145  void operator=(const AnnotatedLine &) = delete;
146 };
147 
148 /// Determines extra information about the tokens comprising an
149 /// \c UnwrappedLine.
151 public:
153  : Style(Style), Keywords(Keywords) {}
154 
155  /// Adapts the indent levels of comment lines to the indent of the
156  /// subsequent line.
157  // FIXME: Can/should this be done in the UnwrappedLineParser?
158  void setCommentLineLevels(SmallVectorImpl<AnnotatedLine *> &Lines);
159 
160  void annotate(AnnotatedLine &Line);
161  void calculateFormattingInformation(AnnotatedLine &Line);
162 
163 private:
164  /// Calculate the penalty for splitting before \c Tok.
165  unsigned splitPenalty(const AnnotatedLine &Line, const FormatToken &Tok,
166  bool InFunctionDecl);
167 
168  bool spaceRequiredBetween(const AnnotatedLine &Line, const FormatToken &Left,
169  const FormatToken &Right);
170 
171  bool spaceRequiredBefore(const AnnotatedLine &Line, const FormatToken &Right);
172 
173  bool mustBreakBefore(const AnnotatedLine &Line, const FormatToken &Right);
174 
175  bool canBreakBefore(const AnnotatedLine &Line, const FormatToken &Right);
176 
177  bool mustBreakForReturnType(const AnnotatedLine &Line) const;
178 
179  void printDebugInfo(const AnnotatedLine &Line);
180 
181  void calculateUnbreakableTailLengths(AnnotatedLine &Line);
182 
183  const FormatStyle &Style;
184 
185  const AdditionalKeywords &Keywords;
186 };
187 
188 } // end namespace format
189 } // end namespace clang
190 
191 #endif
bool endsSequence(A K1, Ts... Tokens) const
true if this token ends a sequence with the given tokens in order, following the Previous pointers...
Definition: FormatToken.h:343
SmallVector< UnwrappedLine, 0 > Children
bool endsWith(Ts... Tokens) const
true if this line ends with the given tokens in reversed order, ignoring comments.
Token Tok
The Token.
Definition: FormatToken.h:128
std::unique_ptr< TokenRole > Role
A token can have a special role that can carry extra information about the token&#39;s formatting...
Definition: FormatToken.h:204
static LLVM_ATTRIBUTE_UNUSED void printDebugInfo(const UnwrappedLine &Line, StringRef Prefix="")
bool LeadingEmptyLinesAffected
True if the leading empty lines of this line intersect with one of the input ranges.
FormatToken * Next
The next token in the unwrapped line.
Definition: FormatToken.h:293
FormatToken * Previous
The previous token in the unwrapped line.
Definition: FormatToken.h:290
const FormatToken & Tok
bool ChildrenAffected
True if one of this line&#39;s children intersects with an input range.
An unwrapped line is a sequence of Token, that we would like to put on a single line if there was no ...
bool mightBeFunctionDefinition() const
true if this line looks like a function definition instead of a function declaration.
bool isOneOf(A K1, B K2) const
Definition: FormatToken.h:316
Determines extra information about the tokens comprising an UnwrappedLine.
std::list< UnwrappedLineNode > Tokens
The Tokens comprising this UnwrappedLine.
const AnnotatedLine * Line
SmallVector< AnnotatedLine *, 0 > Children
A wrapper around a Token storing information about the whitespace characters preceding it...
Definition: FormatToken.h:124
TokenAnnotator(const FormatStyle &Style, const AdditionalKeywords &Keywords)
bool Affected
True if this line should be formatted, i.e.
#define false
Definition: stdbool.h:33
Various functions to configurably format source code.
Encapsulates keywords that are context sensitive or for languages not properly supported by Clang&#39;s l...
Definition: FormatToken.h:660
This file contains the declaration of the UnwrappedLineParser, which turns a stream of tokens into Un...
bool startsWith(Ts... Tokens) const
true if this line starts with the given tokens in order, ignoring comments.
The FormatStyle is used to configure the formatting to follow specific guidelines.
Definition: Format.h:50
ast_type_traits::DynTypedNode Node
Dataflow Directional Tag Classes.
AnnotatedLine(const UnwrappedLine &Line)
bool startsWithNamespace() const
true if this line starts a namespace definition.
SmallVector< AnnotatedLine *, 1 > Children
If this token starts a block, this contains all the unwrapped lines in it.
Definition: FormatToken.h:297
bool startsSequence(A K1, Ts... Tokens) const
true if this token starts a sequence with the given tokens in order, following the Next pointers...
Definition: FormatToken.h:336
const FormatStyle & Style