1 //===--- Format.cpp - Format C++ code -------------------------------------===//
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 /// \brief This file implements functions declared in Format.h. This will be
12 /// split into separate files as we go.
13 ///
14 //===----------------------------------------------------------------------===//
15 
16 #define DEBUG_TYPE "format-formatter"
17 
18 #include "ContinuationIndenter.h"
19 #include "TokenAnnotator.h"
20 #include "UnwrappedLineParser.h"
21 #include "WhitespaceManager.h"
22 #include "clang/Basic/Diagnostic.h"
23 #include "clang/Basic/SourceManager.h"
24 #include "clang/Format/Format.h"
25 #include "clang/Lex/Lexer.h"
26 #include "llvm/ADT/STLExtras.h"
27 #include "llvm/Support/Allocator.h"
28 #include "llvm/Support/Debug.h"
29 #include "llvm/Support/YAMLTraits.h"
30 #include "llvm/Support/Path.h"
31 #include <queue>
32 #include <string>
33 
34 namespace llvm {
35 namespace yaml {
36 template <>
37 struct ScalarEnumerationTraits<clang::format::FormatStyle::LanguageStandard> {
38   static void enumeration(IO &IO,
39                           clang::format::FormatStyle::LanguageStandard &Value) {
40     IO.enumCase(Value, "Cpp03", clang::format::FormatStyle::LS_Cpp03);
41     IO.enumCase(Value, "C++03", clang::format::FormatStyle::LS_Cpp03);
42     IO.enumCase(Value, "Cpp11", clang::format::FormatStyle::LS_Cpp11);
43     IO.enumCase(Value, "C++11", clang::format::FormatStyle::LS_Cpp11);
44     IO.enumCase(Value, "Auto", clang::format::FormatStyle::LS_Auto);
45   }
46 };
47 
48 template <>
49 struct ScalarEnumerationTraits<clang::format::FormatStyle::UseTabStyle> {
50   static void enumeration(IO &IO,
51                           clang::format::FormatStyle::UseTabStyle &Value) {
52     IO.enumCase(Value, "Never", clang::format::FormatStyle::UT_Never);
53     IO.enumCase(Value, "false", clang::format::FormatStyle::UT_Never);
54     IO.enumCase(Value, "Always", clang::format::FormatStyle::UT_Always);
55     IO.enumCase(Value, "true", clang::format::FormatStyle::UT_Always);
56     IO.enumCase(Value, "ForIndentation",
57                 clang::format::FormatStyle::UT_ForIndentation);
58   }
59 };
60 
61 template <>
62 struct ScalarEnumerationTraits<clang::format::FormatStyle::BraceBreakingStyle> {
63   static void
64   enumeration(IO &IO, clang::format::FormatStyle::BraceBreakingStyle &Value) {
65     IO.enumCase(Value, "Attach", clang::format::FormatStyle::BS_Attach);
66     IO.enumCase(Value, "Linux", clang::format::FormatStyle::BS_Linux);
67     IO.enumCase(Value, "Stroustrup", clang::format::FormatStyle::BS_Stroustrup);
68     IO.enumCase(Value, "Allman", clang::format::FormatStyle::BS_Allman);
69   }
70 };
71 
72 template <>
73 struct ScalarEnumerationTraits<
74     clang::format::FormatStyle::NamespaceIndentationKind> {
75   static void
76   enumeration(IO &IO,
77               clang::format::FormatStyle::NamespaceIndentationKind &Value) {
78     IO.enumCase(Value, "None", clang::format::FormatStyle::NI_None);
79     IO.enumCase(Value, "Inner", clang::format::FormatStyle::NI_Inner);
80     IO.enumCase(Value, "All", clang::format::FormatStyle::NI_All);
81   }
82 };
83 
84 template <> struct MappingTraits<clang::format::FormatStyle> {
85   static void mapping(llvm::yaml::IO &IO, clang::format::FormatStyle &Style) {
86     if (IO.outputting()) {
87       StringRef StylesArray[] = { "LLVM",    "Google", "Chromium",
88                                   "Mozilla", "WebKit" };
89       ArrayRef<StringRef> Styles(StylesArray);
90       for (size_t i = 0, e = Styles.size(); i < e; ++i) {
91         StringRef StyleName(Styles[i]);
92         clang::format::FormatStyle PredefinedStyle;
93         if (clang::format::getPredefinedStyle(StyleName, &PredefinedStyle) &&
94             Style == PredefinedStyle) {
95           IO.mapOptional("# BasedOnStyle", StyleName);
96           break;
97         }
98       }
99     } else {
100       StringRef BasedOnStyle;
101       IO.mapOptional("BasedOnStyle", BasedOnStyle);
102       if (!BasedOnStyle.empty())
103         if (!clang::format::getPredefinedStyle(BasedOnStyle, &Style)) {
104           IO.setError(Twine("Unknown value for BasedOnStyle: ", BasedOnStyle));
105           return;
106         }
107     }
108 
109     IO.mapOptional("AccessModifierOffset", Style.AccessModifierOffset);
110     IO.mapOptional("ConstructorInitializerIndentWidth",
111                    Style.ConstructorInitializerIndentWidth);
112     IO.mapOptional("AlignEscapedNewlinesLeft", Style.AlignEscapedNewlinesLeft);
113     IO.mapOptional("AlignTrailingComments", Style.AlignTrailingComments);
114     IO.mapOptional("AllowAllParametersOfDeclarationOnNextLine",
115                    Style.AllowAllParametersOfDeclarationOnNextLine);
116     IO.mapOptional("AllowShortIfStatementsOnASingleLine",
117                    Style.AllowShortIfStatementsOnASingleLine);
118     IO.mapOptional("AllowShortLoopsOnASingleLine",
119                    Style.AllowShortLoopsOnASingleLine);
120     IO.mapOptional("AlwaysBreakTemplateDeclarations",
121                    Style.AlwaysBreakTemplateDeclarations);
122     IO.mapOptional("AlwaysBreakBeforeMultilineStrings",
123                    Style.AlwaysBreakBeforeMultilineStrings);
124     IO.mapOptional("BreakBeforeBinaryOperators",
125                    Style.BreakBeforeBinaryOperators);
126     IO.mapOptional("BreakConstructorInitializersBeforeComma",
127                    Style.BreakConstructorInitializersBeforeComma);
128     IO.mapOptional("BinPackParameters", Style.BinPackParameters);
129     IO.mapOptional("ColumnLimit", Style.ColumnLimit);
130     IO.mapOptional("ConstructorInitializerAllOnOneLineOrOnePerLine",
131                    Style.ConstructorInitializerAllOnOneLineOrOnePerLine);
132     IO.mapOptional("DerivePointerBinding", Style.DerivePointerBinding);
133     IO.mapOptional("ExperimentalAutoDetectBinPacking",
134                    Style.ExperimentalAutoDetectBinPacking);
135     IO.mapOptional("IndentCaseLabels", Style.IndentCaseLabels);
136     IO.mapOptional("MaxEmptyLinesToKeep", Style.MaxEmptyLinesToKeep);
137     IO.mapOptional("NamespaceIndentation", Style.NamespaceIndentation);
138     IO.mapOptional("ObjCSpaceBeforeProtocolList",
139                    Style.ObjCSpaceBeforeProtocolList);
140     IO.mapOptional("PenaltyBreakComment", Style.PenaltyBreakComment);
141     IO.mapOptional("PenaltyBreakString", Style.PenaltyBreakString);
142     IO.mapOptional("PenaltyBreakFirstLessLess",
143                    Style.PenaltyBreakFirstLessLess);
144     IO.mapOptional("PenaltyExcessCharacter", Style.PenaltyExcessCharacter);
145     IO.mapOptional("PenaltyReturnTypeOnItsOwnLine",
146                    Style.PenaltyReturnTypeOnItsOwnLine);
147     IO.mapOptional("PointerBindsToType", Style.PointerBindsToType);
148     IO.mapOptional("SpacesBeforeTrailingComments",
149                    Style.SpacesBeforeTrailingComments);
150     IO.mapOptional("Cpp11BracedListStyle", Style.Cpp11BracedListStyle);
151     IO.mapOptional("Standard", Style.Standard);
152     IO.mapOptional("IndentWidth", Style.IndentWidth);
153     IO.mapOptional("TabWidth", Style.TabWidth);
154     IO.mapOptional("UseTab", Style.UseTab);
155     IO.mapOptional("BreakBeforeBraces", Style.BreakBeforeBraces);
156     IO.mapOptional("IndentFunctionDeclarationAfterType",
157                    Style.IndentFunctionDeclarationAfterType);
158     IO.mapOptional("SpacesInParentheses", Style.SpacesInParentheses);
159     IO.mapOptional("SpaceInEmptyParentheses", Style.SpaceInEmptyParentheses);
160     IO.mapOptional("SpacesInCStyleCastParentheses",
161                    Style.SpacesInCStyleCastParentheses);
162     IO.mapOptional("SpaceAfterControlStatementKeyword",
163                    Style.SpaceAfterControlStatementKeyword);
164     IO.mapOptional("SpaceBeforeAssignmentOperators",
165                    Style.SpaceBeforeAssignmentOperators);
166   }
167 };
168 }
169 }
170 
171 namespace clang {
172 namespace format {
173 
174 void setDefaultPenalties(FormatStyle &Style) {
175   Style.PenaltyBreakComment = 60;
176   Style.PenaltyBreakFirstLessLess = 120;
177   Style.PenaltyBreakString = 1000;
178   Style.PenaltyExcessCharacter = 1000000;
179 }
180 
181 FormatStyle getLLVMStyle() {
182   FormatStyle LLVMStyle;
183   LLVMStyle.AccessModifierOffset = -2;
184   LLVMStyle.AlignEscapedNewlinesLeft = false;
185   LLVMStyle.AlignTrailingComments = true;
186   LLVMStyle.AllowAllParametersOfDeclarationOnNextLine = true;
187   LLVMStyle.AllowShortIfStatementsOnASingleLine = false;
188   LLVMStyle.AllowShortLoopsOnASingleLine = false;
189   LLVMStyle.AlwaysBreakBeforeMultilineStrings = false;
190   LLVMStyle.AlwaysBreakTemplateDeclarations = false;
191   LLVMStyle.BinPackParameters = true;
192   LLVMStyle.BreakBeforeBinaryOperators = false;
193   LLVMStyle.BreakBeforeBraces = FormatStyle::BS_Attach;
194   LLVMStyle.BreakConstructorInitializersBeforeComma = false;
195   LLVMStyle.ColumnLimit = 80;
196   LLVMStyle.ConstructorInitializerAllOnOneLineOrOnePerLine = false;
197   LLVMStyle.ConstructorInitializerIndentWidth = 4;
198   LLVMStyle.Cpp11BracedListStyle = false;
199   LLVMStyle.DerivePointerBinding = false;
200   LLVMStyle.ExperimentalAutoDetectBinPacking = false;
201   LLVMStyle.IndentCaseLabels = false;
202   LLVMStyle.IndentFunctionDeclarationAfterType = false;
203   LLVMStyle.IndentWidth = 2;
204   LLVMStyle.TabWidth = 8;
205   LLVMStyle.MaxEmptyLinesToKeep = 1;
206   LLVMStyle.NamespaceIndentation = FormatStyle::NI_None;
207   LLVMStyle.ObjCSpaceBeforeProtocolList = true;
208   LLVMStyle.PointerBindsToType = false;
209   LLVMStyle.SpacesBeforeTrailingComments = 1;
210   LLVMStyle.Standard = FormatStyle::LS_Cpp03;
211   LLVMStyle.UseTab = FormatStyle::UT_Never;
212   LLVMStyle.SpacesInParentheses = false;
213   LLVMStyle.SpaceInEmptyParentheses = false;
214   LLVMStyle.SpacesInCStyleCastParentheses = false;
215   LLVMStyle.SpaceAfterControlStatementKeyword = true;
216   LLVMStyle.SpaceBeforeAssignmentOperators = true;
217 
218   setDefaultPenalties(LLVMStyle);
219   LLVMStyle.PenaltyReturnTypeOnItsOwnLine = 60;
220 
221   return LLVMStyle;
222 }
223 
224 FormatStyle getGoogleStyle() {
225   FormatStyle GoogleStyle;
226   GoogleStyle.AccessModifierOffset = -1;
227   GoogleStyle.AlignEscapedNewlinesLeft = true;
228   GoogleStyle.AlignTrailingComments = true;
229   GoogleStyle.AllowAllParametersOfDeclarationOnNextLine = true;
230   GoogleStyle.AllowShortIfStatementsOnASingleLine = true;
231   GoogleStyle.AllowShortLoopsOnASingleLine = true;
232   GoogleStyle.AlwaysBreakBeforeMultilineStrings = true;
233   GoogleStyle.AlwaysBreakTemplateDeclarations = true;
234   GoogleStyle.BinPackParameters = true;
235   GoogleStyle.BreakBeforeBinaryOperators = false;
236   GoogleStyle.BreakBeforeBraces = FormatStyle::BS_Attach;
237   GoogleStyle.BreakConstructorInitializersBeforeComma = false;
238   GoogleStyle.ColumnLimit = 80;
239   GoogleStyle.ConstructorInitializerAllOnOneLineOrOnePerLine = true;
240   GoogleStyle.ConstructorInitializerIndentWidth = 4;
241   GoogleStyle.Cpp11BracedListStyle = true;
242   GoogleStyle.DerivePointerBinding = true;
243   GoogleStyle.ExperimentalAutoDetectBinPacking = false;
244   GoogleStyle.IndentCaseLabels = true;
245   GoogleStyle.IndentFunctionDeclarationAfterType = true;
246   GoogleStyle.IndentWidth = 2;
247   GoogleStyle.TabWidth = 8;
248   GoogleStyle.MaxEmptyLinesToKeep = 1;
249   GoogleStyle.NamespaceIndentation = FormatStyle::NI_None;
250   GoogleStyle.ObjCSpaceBeforeProtocolList = false;
251   GoogleStyle.PointerBindsToType = true;
252   GoogleStyle.SpacesBeforeTrailingComments = 2;
253   GoogleStyle.Standard = FormatStyle::LS_Auto;
254   GoogleStyle.UseTab = FormatStyle::UT_Never;
255   GoogleStyle.SpacesInParentheses = false;
256   GoogleStyle.SpaceInEmptyParentheses = false;
257   GoogleStyle.SpacesInCStyleCastParentheses = false;
258   GoogleStyle.SpaceAfterControlStatementKeyword = true;
259   GoogleStyle.SpaceBeforeAssignmentOperators = true;
260 
261   setDefaultPenalties(GoogleStyle);
262   GoogleStyle.PenaltyReturnTypeOnItsOwnLine = 200;
263 
264   return GoogleStyle;
265 }
266 
267 FormatStyle getChromiumStyle() {
268   FormatStyle ChromiumStyle = getGoogleStyle();
269   ChromiumStyle.AllowAllParametersOfDeclarationOnNextLine = false;
270   ChromiumStyle.AllowShortIfStatementsOnASingleLine = false;
271   ChromiumStyle.AllowShortLoopsOnASingleLine = false;
272   ChromiumStyle.BinPackParameters = false;
273   ChromiumStyle.DerivePointerBinding = false;
274   ChromiumStyle.Standard = FormatStyle::LS_Cpp03;
275   return ChromiumStyle;
276 }
277 
278 FormatStyle getMozillaStyle() {
279   FormatStyle MozillaStyle = getLLVMStyle();
280   MozillaStyle.AllowAllParametersOfDeclarationOnNextLine = false;
281   MozillaStyle.ConstructorInitializerAllOnOneLineOrOnePerLine = true;
282   MozillaStyle.DerivePointerBinding = true;
283   MozillaStyle.IndentCaseLabels = true;
284   MozillaStyle.ObjCSpaceBeforeProtocolList = false;
285   MozillaStyle.PenaltyReturnTypeOnItsOwnLine = 200;
286   MozillaStyle.PointerBindsToType = true;
287   return MozillaStyle;
288 }
289 
290 FormatStyle getWebKitStyle() {
291   FormatStyle Style = getLLVMStyle();
292   Style.AccessModifierOffset = -4;
293   Style.AlignTrailingComments = false;
294   Style.BreakBeforeBinaryOperators = true;
295   Style.BreakBeforeBraces = FormatStyle::BS_Stroustrup;
296   Style.BreakConstructorInitializersBeforeComma = true;
297   Style.ColumnLimit = 0;
298   Style.IndentWidth = 4;
299   Style.NamespaceIndentation = FormatStyle::NI_Inner;
300   Style.PointerBindsToType = true;
301   return Style;
302 }
303 
304 bool getPredefinedStyle(StringRef Name, FormatStyle *Style) {
305   if (Name.equals_lower("llvm"))
306     *Style = getLLVMStyle();
307   else if (Name.equals_lower("chromium"))
308     *Style = getChromiumStyle();
309   else if (Name.equals_lower("mozilla"))
310     *Style = getMozillaStyle();
311   else if (Name.equals_lower("google"))
312     *Style = getGoogleStyle();
313   else if (Name.equals_lower("webkit"))
314     *Style = getWebKitStyle();
315   else
316     return false;
317 
318   return true;
319 }
320 
321 llvm::error_code parseConfiguration(StringRef Text, FormatStyle *Style) {
322   if (Text.trim().empty())
323     return llvm::make_error_code(llvm::errc::invalid_argument);
324   llvm::yaml::Input Input(Text);
325   Input >> *Style;
326   return Input.error();
327 }
328 
329 std::string configurationAsText(const FormatStyle &Style) {
330   std::string Text;
331   llvm::raw_string_ostream Stream(Text);
332   llvm::yaml::Output Output(Stream);
333   // We use the same mapping method for input and output, so we need a non-const
334   // reference here.
335   FormatStyle NonConstStyle = Style;
336   Output << NonConstStyle;
337   return Stream.str();
338 }
339 
340 namespace {
341 
342 class NoColumnLimitFormatter {
343 public:
344   NoColumnLimitFormatter(ContinuationIndenter *Indenter) : Indenter(Indenter) {}
345 
346   /// \brief Formats the line starting at \p State, simply keeping all of the
347   /// input's line breaking decisions.
348   void format(unsigned FirstIndent, const AnnotatedLine *Line) {
349     LineState State =
350         Indenter->getInitialState(FirstIndent, Line, /*DryRun=*/false);
351     while (State.NextToken != NULL) {
352       bool Newline =
353           Indenter->mustBreak(State) ||
354           (Indenter->canBreak(State) && State.NextToken->NewlinesBefore > 0);
355       Indenter->addTokenToState(State, Newline, /*DryRun=*/false);
356     }
357   }
358 
359 private:
360   ContinuationIndenter *Indenter;
361 };
362 
363 class UnwrappedLineFormatter {
364 public:
365   UnwrappedLineFormatter(ContinuationIndenter *Indenter,
366                          WhitespaceManager *Whitespaces,
367                          const FormatStyle &Style, const AnnotatedLine &Line)
368       : Indenter(Indenter), Whitespaces(Whitespaces), Style(Style), Line(Line),
369         Count(0) {}
370 
371   /// \brief Formats an \c UnwrappedLine and returns the penalty.
372   ///
373   /// If \p DryRun is \c false, directly applies the changes.
374   unsigned format(unsigned FirstIndent, bool DryRun = false) {
375     LineState State = Indenter->getInitialState(FirstIndent, &Line, DryRun);
376 
377     // If the ObjC method declaration does not fit on a line, we should format
378     // it with one arg per line.
379     if (Line.Type == LT_ObjCMethodDecl)
380       State.Stack.back().BreakBeforeParameter = true;
381 
382     // Find best solution in solution space.
383     return analyzeSolutionSpace(State, DryRun);
384   }
385 
386 private:
387   /// \brief An edge in the solution space from \c Previous->State to \c State,
388   /// inserting a newline dependent on the \c NewLine.
389   struct StateNode {
390     StateNode(const LineState &State, bool NewLine, StateNode *Previous)
391         : State(State), NewLine(NewLine), Previous(Previous) {}
392     LineState State;
393     bool NewLine;
394     StateNode *Previous;
395   };
396 
397   /// \brief A pair of <penalty, count> that is used to prioritize the BFS on.
398   ///
399   /// In case of equal penalties, we want to prefer states that were inserted
400   /// first. During state generation we make sure that we insert states first
401   /// that break the line as late as possible.
402   typedef std::pair<unsigned, unsigned> OrderedPenalty;
403 
404   /// \brief An item in the prioritized BFS search queue. The \c StateNode's
405   /// \c State has the given \c OrderedPenalty.
406   typedef std::pair<OrderedPenalty, StateNode *> QueueItem;
407 
408   /// \brief The BFS queue type.
409   typedef std::priority_queue<QueueItem, std::vector<QueueItem>,
410                               std::greater<QueueItem> > QueueType;
411 
412   /// \brief Analyze the entire solution space starting from \p InitialState.
413   ///
414   /// This implements a variant of Dijkstra's algorithm on the graph that spans
415   /// the solution space (\c LineStates are the nodes). The algorithm tries to
416   /// find the shortest path (the one with lowest penalty) from \p InitialState
417   /// to a state where all tokens are placed. Returns the penalty.
418   ///
419   /// If \p DryRun is \c false, directly applies the changes.
420   unsigned analyzeSolutionSpace(LineState &InitialState, bool DryRun = false) {
421     std::set<LineState> Seen;
422 
423     // Insert start element into queue.
424     StateNode *Node =
425         new (Allocator.Allocate()) StateNode(InitialState, false, NULL);
426     Queue.push(QueueItem(OrderedPenalty(0, Count), Node));
427     ++Count;
428 
429     unsigned Penalty = 0;
430 
431     // While not empty, take first element and follow edges.
432     while (!Queue.empty()) {
433       Penalty = Queue.top().first.first;
434       StateNode *Node = Queue.top().second;
435       if (Node->State.NextToken == NULL) {
436         DEBUG(llvm::dbgs() << "\n---\nPenalty for line: " << Penalty << "\n");
437         break;
438       }
439       Queue.pop();
440 
441       // Cut off the analysis of certain solutions if the analysis gets too
442       // complex. See description of IgnoreStackForComparison.
443       if (Count > 10000)
444         Node->State.IgnoreStackForComparison = true;
445 
446       if (!Seen.insert(Node->State).second)
447         // State already examined with lower penalty.
448         continue;
449 
450       FormatDecision LastFormat = Node->State.NextToken->Decision;
451       if (LastFormat == FD_Unformatted || LastFormat == FD_Continue)
452         addNextStateToQueue(Penalty, Node, /*NewLine=*/false);
453       if (LastFormat == FD_Unformatted || LastFormat == FD_Break)
454         addNextStateToQueue(Penalty, Node, /*NewLine=*/true);
455     }
456 
457     if (Queue.empty()) {
458       // We were unable to find a solution, do nothing.
459       // FIXME: Add diagnostic?
460       DEBUG(llvm::dbgs() << "Could not find a solution.\n");
461       return 0;
462     }
463 
464     // Reconstruct the solution.
465     if (!DryRun)
466       reconstructPath(InitialState, Queue.top().second);
467 
468     DEBUG(llvm::dbgs() << "Total number of analyzed states: " << Count << "\n");
469     DEBUG(llvm::dbgs() << "---\n");
470 
471     return Penalty;
472   }
473 
474   void reconstructPath(LineState &State, StateNode *Current) {
475     std::deque<StateNode *> Path;
476     // We do not need a break before the initial token.
477     while (Current->Previous) {
478       Path.push_front(Current);
479       Current = Current->Previous;
480     }
481     for (std::deque<StateNode *>::iterator I = Path.begin(), E = Path.end();
482          I != E; ++I) {
483       unsigned Penalty = 0;
484       formatChildren(State, (*I)->NewLine, /*DryRun=*/false, Penalty);
485       Penalty += Indenter->addTokenToState(State, (*I)->NewLine, false);
486 
487       DEBUG({
488         if ((*I)->NewLine) {
489           llvm::dbgs() << "Penalty for placing "
490                        << (*I)->Previous->State.NextToken->Tok.getName() << ": "
491                        << Penalty << "\n";
492         }
493       });
494     }
495   }
496 
497   /// \brief Add the following state to the analysis queue \c Queue.
498   ///
499   /// Assume the current state is \p PreviousNode and has been reached with a
500   /// penalty of \p Penalty. Insert a line break if \p NewLine is \c true.
501   void addNextStateToQueue(unsigned Penalty, StateNode *PreviousNode,
502                            bool NewLine) {
503     if (NewLine && !Indenter->canBreak(PreviousNode->State))
504       return;
505     if (!NewLine && Indenter->mustBreak(PreviousNode->State))
506       return;
507 
508     StateNode *Node = new (Allocator.Allocate())
509         StateNode(PreviousNode->State, NewLine, PreviousNode);
510     if (!formatChildren(Node->State, NewLine, /*DryRun=*/true, Penalty))
511       return;
512 
513     Penalty += Indenter->addTokenToState(Node->State, NewLine, true);
514 
515     Queue.push(QueueItem(OrderedPenalty(Penalty, Count), Node));
516     ++Count;
517   }
518 
519   /// \brief If the \p State's next token is an r_brace closing a nested block,
520   /// format the nested block before it.
521   ///
522   /// Returns \c true if all children could be placed successfully and adapts
523   /// \p Penalty as well as \p State. If \p DryRun is false, also directly
524   /// creates changes using \c Whitespaces.
525   ///
526   /// The crucial idea here is that children always get formatted upon
527   /// encountering the closing brace right after the nested block. Now, if we
528   /// are currently trying to keep the "}" on the same line (i.e. \p NewLine is
529   /// \c false), the entire block has to be kept on the same line (which is only
530   /// possible if it fits on the line, only contains a single statement, etc.
531   ///
532   /// If \p NewLine is true, we format the nested block on separate lines, i.e.
533   /// break after the "{", format all lines with correct indentation and the put
534   /// the closing "}" on yet another new line.
535   ///
536   /// This enables us to keep the simple structure of the
537   /// \c UnwrappedLineFormatter, where we only have two options for each token:
538   /// break or don't break.
539   bool formatChildren(LineState &State, bool NewLine, bool DryRun,
540                       unsigned &Penalty) {
541     const FormatToken &LBrace = *State.NextToken->Previous;
542     if (LBrace.isNot(tok::l_brace) || LBrace.BlockKind != BK_Block ||
543         LBrace.Children.size() == 0)
544       // The previous token does not open a block. Nothing to do. We don't
545       // assert so that we can simply call this function for all tokens.
546       return true;
547 
548     if (NewLine) {
549       unsigned ParentIndent = State.Stack.back().Indent;
550       for (SmallVector<AnnotatedLine *, 1>::const_iterator
551                I = LBrace.Children.begin(),
552                E = LBrace.Children.end();
553            I != E; ++I) {
554         unsigned Indent =
555             ParentIndent + ((*I)->Level - Line.Level - 1) * Style.IndentWidth;
556         if (!DryRun) {
557           unsigned Newlines = std::min((*I)->First->NewlinesBefore,
558                                        Style.MaxEmptyLinesToKeep + 1);
559           Newlines = std::max(1u, Newlines);
560           Whitespaces->replaceWhitespace(
561               *(*I)->First, Newlines, (*I)->Level, /*Spaces=*/Indent,
562               /*StartOfTokenColumn=*/Indent, Line.InPPDirective);
563         }
564         UnwrappedLineFormatter Formatter(Indenter, Whitespaces, Style, **I);
565         Penalty += Formatter.format(Indent, DryRun);
566       }
567       return true;
568     }
569 
570     if (LBrace.Children.size() > 1)
571       return false; // Cannot merge multiple statements into a single line.
572 
573     // We can't put the closing "}" on a line with a trailing comment.
574     if (LBrace.Children[0]->Last->isTrailingComment())
575       return false;
576 
577     if (!DryRun) {
578       Whitespaces->replaceWhitespace(
579           *LBrace.Children[0]->First,
580           /*Newlines=*/0, /*IndentLevel=*/0, /*Spaces=*/1,
581           /*StartOfTokenColumn=*/State.Column, State.Line->InPPDirective);
582       UnwrappedLineFormatter Formatter(Indenter, Whitespaces, Style,
583                                        *LBrace.Children[0]);
584       Penalty += Formatter.format(State.Column + 1, DryRun);
585     }
586 
587     State.Column += 1 + LBrace.Children[0]->Last->TotalLength;
588     return true;
589   }
590 
591   ContinuationIndenter *Indenter;
592   WhitespaceManager *Whitespaces;
593   FormatStyle Style;
594   const AnnotatedLine &Line;
595 
596   llvm::SpecificBumpPtrAllocator<StateNode> Allocator;
597   QueueType Queue;
598   // Increasing count of \c StateNode items we have created. This is used
599   // to create a deterministic order independent of the container.
600   unsigned Count;
601 };
602 
603 class FormatTokenLexer {
604 public:
605   FormatTokenLexer(Lexer &Lex, SourceManager &SourceMgr, FormatStyle &Style,
606                    encoding::Encoding Encoding)
607       : FormatTok(NULL), GreaterStashed(false), Column(0),
608         TrailingWhitespace(0), Lex(Lex), SourceMgr(SourceMgr), Style(Style),
609         IdentTable(getFormattingLangOpts()), Encoding(Encoding) {
610     Lex.SetKeepWhitespaceMode(true);
611   }
612 
613   ArrayRef<FormatToken *> lex() {
614     assert(Tokens.empty());
615     do {
616       Tokens.push_back(getNextToken());
617       maybeJoinPreviousTokens();
618     } while (Tokens.back()->Tok.isNot(tok::eof));
619     return Tokens;
620   }
621 
622   IdentifierTable &getIdentTable() { return IdentTable; }
623 
624 private:
625   void maybeJoinPreviousTokens() {
626     if (Tokens.size() < 4)
627       return;
628     FormatToken *Last = Tokens.back();
629     if (!Last->is(tok::r_paren))
630       return;
631 
632     FormatToken *String = Tokens[Tokens.size() - 2];
633     if (!String->is(tok::string_literal) || String->IsMultiline)
634       return;
635 
636     if (!Tokens[Tokens.size() - 3]->is(tok::l_paren))
637       return;
638 
639     FormatToken *Macro = Tokens[Tokens.size() - 4];
640     if (Macro->TokenText != "_T")
641       return;
642 
643     const char *Start = Macro->TokenText.data();
644     const char *End = Last->TokenText.data() + Last->TokenText.size();
645     String->TokenText = StringRef(Start, End - Start);
646     String->IsFirst = Macro->IsFirst;
647     String->LastNewlineOffset = Macro->LastNewlineOffset;
648     String->WhitespaceRange = Macro->WhitespaceRange;
649     String->OriginalColumn = Macro->OriginalColumn;
650     String->ColumnWidth = encoding::columnWidthWithTabs(
651         String->TokenText, String->OriginalColumn, Style.TabWidth, Encoding);
652 
653     Tokens.pop_back();
654     Tokens.pop_back();
655     Tokens.pop_back();
656     Tokens.back() = String;
657   }
658 
659   FormatToken *getNextToken() {
660     if (GreaterStashed) {
661       // Create a synthesized second '>' token.
662       // FIXME: Increment Column and set OriginalColumn.
663       Token Greater = FormatTok->Tok;
664       FormatTok = new (Allocator.Allocate()) FormatToken;
665       FormatTok->Tok = Greater;
666       SourceLocation GreaterLocation =
667           FormatTok->Tok.getLocation().getLocWithOffset(1);
668       FormatTok->WhitespaceRange =
669           SourceRange(GreaterLocation, GreaterLocation);
670       FormatTok->TokenText = ">";
671       FormatTok->ColumnWidth = 1;
672       GreaterStashed = false;
673       return FormatTok;
674     }
675 
676     FormatTok = new (Allocator.Allocate()) FormatToken;
677     readRawToken(*FormatTok);
678     SourceLocation WhitespaceStart =
679         FormatTok->Tok.getLocation().getLocWithOffset(-TrailingWhitespace);
680     if (SourceMgr.getFileOffset(WhitespaceStart) == 0)
681       FormatTok->IsFirst = true;
682 
683     // Consume and record whitespace until we find a significant token.
684     unsigned WhitespaceLength = TrailingWhitespace;
685     while (FormatTok->Tok.is(tok::unknown)) {
686       for (int i = 0, e = FormatTok->TokenText.size(); i != e; ++i) {
687         switch (FormatTok->TokenText[i]) {
688         case '\n':
689           ++FormatTok->NewlinesBefore;
690           // FIXME: This is technically incorrect, as it could also
691           // be a literal backslash at the end of the line.
692           if (i == 0 || (FormatTok->TokenText[i - 1] != '\\' &&
693                          (FormatTok->TokenText[i - 1] != '\r' || i == 1 ||
694                           FormatTok->TokenText[i - 2] != '\\')))
695             FormatTok->HasUnescapedNewline = true;
696           FormatTok->LastNewlineOffset = WhitespaceLength + i + 1;
697           Column = 0;
698           break;
699         case '\r':
700         case '\f':
701         case '\v':
702           Column = 0;
703           break;
704         case ' ':
705           ++Column;
706           break;
707         case '\t':
708           Column += Style.TabWidth - Column % Style.TabWidth;
709           break;
710         case '\\':
711           ++Column;
712           if (i + 1 == e || (FormatTok->TokenText[i + 1] != '\r' &&
713                              FormatTok->TokenText[i + 1] != '\n'))
714             FormatTok->Type = TT_ImplicitStringLiteral;
715           break;
716         default:
717           FormatTok->Type = TT_ImplicitStringLiteral;
718           ++Column;
719           break;
720         }
721       }
722 
723       if (FormatTok->Type == TT_ImplicitStringLiteral)
724         break;
725       WhitespaceLength += FormatTok->Tok.getLength();
726 
727       readRawToken(*FormatTok);
728     }
729 
730     // In case the token starts with escaped newlines, we want to
731     // take them into account as whitespace - this pattern is quite frequent
732     // in macro definitions.
733     // FIXME: Add a more explicit test.
734     while (FormatTok->TokenText.size() > 1 && FormatTok->TokenText[0] == '\\' &&
735            FormatTok->TokenText[1] == '\n') {
736       // FIXME: ++FormatTok->NewlinesBefore is missing...
737       WhitespaceLength += 2;
738       Column = 0;
739       FormatTok->TokenText = FormatTok->TokenText.substr(2);
740     }
741 
742     FormatTok->WhitespaceRange = SourceRange(
743         WhitespaceStart, WhitespaceStart.getLocWithOffset(WhitespaceLength));
744 
745     FormatTok->OriginalColumn = Column;
746 
747     TrailingWhitespace = 0;
748     if (FormatTok->Tok.is(tok::comment)) {
749       // FIXME: Add the trimmed whitespace to Column.
750       StringRef UntrimmedText = FormatTok->TokenText;
751       FormatTok->TokenText = FormatTok->TokenText.rtrim(" \t\v\f");
752       TrailingWhitespace = UntrimmedText.size() - FormatTok->TokenText.size();
753     } else if (FormatTok->Tok.is(tok::raw_identifier)) {
754       IdentifierInfo &Info = IdentTable.get(FormatTok->TokenText);
755       FormatTok->Tok.setIdentifierInfo(&Info);
756       FormatTok->Tok.setKind(Info.getTokenID());
757     } else if (FormatTok->Tok.is(tok::greatergreater)) {
758       FormatTok->Tok.setKind(tok::greater);
759       FormatTok->TokenText = FormatTok->TokenText.substr(0, 1);
760       GreaterStashed = true;
761     }
762 
763     // Now FormatTok is the next non-whitespace token.
764 
765     StringRef Text = FormatTok->TokenText;
766     size_t FirstNewlinePos = Text.find('\n');
767     if (FirstNewlinePos == StringRef::npos) {
768       // FIXME: ColumnWidth actually depends on the start column, we need to
769       // take this into account when the token is moved.
770       FormatTok->ColumnWidth =
771           encoding::columnWidthWithTabs(Text, Column, Style.TabWidth, Encoding);
772       Column += FormatTok->ColumnWidth;
773     } else {
774       FormatTok->IsMultiline = true;
775       // FIXME: ColumnWidth actually depends on the start column, we need to
776       // take this into account when the token is moved.
777       FormatTok->ColumnWidth = encoding::columnWidthWithTabs(
778           Text.substr(0, FirstNewlinePos), Column, Style.TabWidth, Encoding);
779 
780       // The last line of the token always starts in column 0.
781       // Thus, the length can be precomputed even in the presence of tabs.
782       FormatTok->LastLineColumnWidth = encoding::columnWidthWithTabs(
783           Text.substr(Text.find_last_of('\n') + 1), 0, Style.TabWidth,
784           Encoding);
785       Column = FormatTok->LastLineColumnWidth;
786     }
787 
788     return FormatTok;
789   }
790 
791   FormatToken *FormatTok;
792   bool GreaterStashed;
793   unsigned Column;
794   unsigned TrailingWhitespace;
795   Lexer &Lex;
796   SourceManager &SourceMgr;
797   FormatStyle &Style;
798   IdentifierTable IdentTable;
799   encoding::Encoding Encoding;
800   llvm::SpecificBumpPtrAllocator<FormatToken> Allocator;
801   SmallVector<FormatToken *, 16> Tokens;
802 
803   void readRawToken(FormatToken &Tok) {
804     Lex.LexFromRawLexer(Tok.Tok);
805     Tok.TokenText = StringRef(SourceMgr.getCharacterData(Tok.Tok.getLocation()),
806                               Tok.Tok.getLength());
807     // For formatting, treat unterminated string literals like normal string
808     // literals.
809     if (Tok.is(tok::unknown) && !Tok.TokenText.empty() &&
810         Tok.TokenText[0] == '"') {
811       Tok.Tok.setKind(tok::string_literal);
812       Tok.IsUnterminatedLiteral = true;
813     }
814   }
815 };
816 
817 class Formatter : public UnwrappedLineConsumer {
818 public:
819   Formatter(const FormatStyle &Style, Lexer &Lex, SourceManager &SourceMgr,
820             const std::vector<CharSourceRange> &Ranges)
821       : Style(Style), Lex(Lex), SourceMgr(SourceMgr),
822         Whitespaces(SourceMgr, Style, inputUsesCRLF(Lex.getBuffer())),
823         Ranges(Ranges), UnwrappedLines(1),
824         Encoding(encoding::detectEncoding(Lex.getBuffer())) {
825     DEBUG(llvm::dbgs() << "File encoding: "
826                        << (Encoding == encoding::Encoding_UTF8 ? "UTF8"
827                                                                : "unknown")
828                        << "\n");
829   }
830 
831   tooling::Replacements format() {
832     tooling::Replacements Result;
833     FormatTokenLexer Tokens(Lex, SourceMgr, Style, Encoding);
834 
835     UnwrappedLineParser Parser(Style, Tokens.lex(), *this);
836     bool StructuralError = Parser.parse();
837     assert(UnwrappedLines.rbegin()->empty());
838     for (unsigned Run = 0, RunE = UnwrappedLines.size(); Run + 1 != RunE;
839          ++Run) {
840       DEBUG(llvm::dbgs() << "Run " << Run << "...\n");
841       SmallVector<AnnotatedLine *, 16> AnnotatedLines;
842       for (unsigned i = 0, e = UnwrappedLines[Run].size(); i != e; ++i) {
843         AnnotatedLines.push_back(new AnnotatedLine(UnwrappedLines[Run][i]));
844       }
845       tooling::Replacements RunResult =
846           format(AnnotatedLines, StructuralError, Tokens);
847       DEBUG({
848         llvm::dbgs() << "Replacements for run " << Run << ":\n";
849         for (tooling::Replacements::iterator I = RunResult.begin(),
850                                              E = RunResult.end();
851              I != E; ++I) {
852           llvm::dbgs() << I->toString() << "\n";
853         }
854       });
855       for (unsigned i = 0, e = AnnotatedLines.size(); i != e; ++i) {
856         delete AnnotatedLines[i];
857       }
858       Result.insert(RunResult.begin(), RunResult.end());
859       Whitespaces.reset();
860     }
861     return Result;
862   }
863 
864   tooling::Replacements format(SmallVectorImpl<AnnotatedLine *> &AnnotatedLines,
865                                bool StructuralError, FormatTokenLexer &Tokens) {
866     TokenAnnotator Annotator(Style, Tokens.getIdentTable().get("in"));
867     for (unsigned i = 0, e = AnnotatedLines.size(); i != e; ++i) {
868       Annotator.annotate(*AnnotatedLines[i]);
869     }
870     deriveLocalStyle(AnnotatedLines);
871     for (unsigned i = 0, e = AnnotatedLines.size(); i != e; ++i) {
872       Annotator.calculateFormattingInformation(*AnnotatedLines[i]);
873     }
874 
875     Annotator.setCommentLineLevels(AnnotatedLines);
876 
877     std::vector<int> IndentForLevel;
878     bool PreviousLineWasTouched = false;
879     const AnnotatedLine *PreviousLine = NULL;
880     bool FormatPPDirective = false;
881     for (SmallVectorImpl<AnnotatedLine *>::iterator I = AnnotatedLines.begin(),
882                                                     E = AnnotatedLines.end();
883          I != E; ++I) {
884       const AnnotatedLine &TheLine = **I;
885       const FormatToken *FirstTok = TheLine.First;
886       int Offset = getIndentOffset(*TheLine.First);
887 
888       // Check whether this line is part of a formatted preprocessor directive.
889       if (FirstTok->HasUnescapedNewline)
890         FormatPPDirective = false;
891       if (!FormatPPDirective && TheLine.InPPDirective &&
892           (touchesLine(TheLine) || touchesPPDirective(I + 1, E)))
893         FormatPPDirective = true;
894 
895       // Determine indent and try to merge multiple unwrapped lines.
896       while (IndentForLevel.size() <= TheLine.Level)
897         IndentForLevel.push_back(-1);
898       IndentForLevel.resize(TheLine.Level + 1);
899       unsigned Indent = getIndent(IndentForLevel, TheLine.Level);
900       if (static_cast<int>(Indent) + Offset >= 0)
901         Indent += Offset;
902       tryFitMultipleLinesInOne(Indent, I, E);
903 
904       bool WasMoved = PreviousLineWasTouched && FirstTok->NewlinesBefore == 0;
905       if (TheLine.First->is(tok::eof)) {
906         if (PreviousLineWasTouched) {
907           unsigned Newlines = std::min(FirstTok->NewlinesBefore, 1u);
908           Whitespaces.replaceWhitespace(*TheLine.First, Newlines,
909                                         /*IndentLevel=*/0, /*Spaces=*/0,
910                                         /*TargetColumn=*/0);
911         }
912       } else if (TheLine.Type != LT_Invalid &&
913                  (WasMoved || FormatPPDirective || touchesLine(TheLine))) {
914         unsigned LevelIndent = getIndent(IndentForLevel, TheLine.Level);
915         if (FirstTok->WhitespaceRange.isValid() &&
916             // Insert a break even if there is a structural error in case where
917             // we break apart a line consisting of multiple unwrapped lines.
918             (FirstTok->NewlinesBefore == 0 || !StructuralError)) {
919           formatFirstToken(*TheLine.First, PreviousLine, TheLine.Level, Indent,
920                            TheLine.InPPDirective);
921         } else {
922           Indent = LevelIndent = FirstTok->OriginalColumn;
923         }
924         ContinuationIndenter Indenter(Style, SourceMgr, Whitespaces, Encoding,
925                                       BinPackInconclusiveFunctions);
926 
927         // If everything fits on a single line, just put it there.
928         unsigned ColumnLimit = Style.ColumnLimit;
929         AnnotatedLine *NextLine = *(I + 1);
930         if ((I + 1) != E && NextLine->InPPDirective &&
931             !NextLine->First->HasUnescapedNewline)
932           ColumnLimit = getColumnLimit(TheLine.InPPDirective);
933 
934         if (TheLine.Last->TotalLength + Indent <= ColumnLimit) {
935           LineState State =
936               Indenter.getInitialState(Indent, &TheLine, /*DryRun=*/false);
937           while (State.NextToken != NULL)
938             Indenter.addTokenToState(State, false, false);
939         } else if (Style.ColumnLimit == 0) {
940           NoColumnLimitFormatter Formatter(&Indenter);
941           Formatter.format(Indent, &TheLine);
942         } else {
943           UnwrappedLineFormatter Formatter(&Indenter, &Whitespaces, Style,
944                                            TheLine);
945           Formatter.format(Indent);
946         }
947 
948         IndentForLevel[TheLine.Level] = LevelIndent;
949         PreviousLineWasTouched = true;
950       } else {
951         // Format the first token if necessary, and notify the WhitespaceManager
952         // about the unchanged whitespace.
953         for (FormatToken *Tok = TheLine.First; Tok != NULL; Tok = Tok->Next) {
954           if (Tok == TheLine.First &&
955               (Tok->NewlinesBefore > 0 || Tok->IsFirst)) {
956             unsigned LevelIndent = Tok->OriginalColumn;
957             // Remove trailing whitespace of the previous line if it was
958             // touched.
959             if (PreviousLineWasTouched || touchesEmptyLineBefore(TheLine)) {
960               formatFirstToken(*Tok, PreviousLine, TheLine.Level, LevelIndent,
961                                TheLine.InPPDirective);
962             } else {
963               Whitespaces.addUntouchableToken(*Tok, TheLine.InPPDirective);
964             }
965 
966             if (static_cast<int>(LevelIndent) - Offset >= 0)
967               LevelIndent -= Offset;
968             if (Tok->isNot(tok::comment))
969               IndentForLevel[TheLine.Level] = LevelIndent;
970           } else {
971             Whitespaces.addUntouchableToken(*Tok, TheLine.InPPDirective);
972           }
973         }
974         // If we did not reformat this unwrapped line, the column at the end of
975         // the last token is unchanged - thus, we can calculate the end of the
976         // last token.
977         PreviousLineWasTouched = false;
978       }
979       for (FormatToken *Tok = TheLine.First; Tok != NULL; Tok = Tok->Next) {
980         Tok->Finalized = true;
981       }
982       PreviousLine = *I;
983     }
984     return Whitespaces.generateReplacements();
985   }
986 
987 private:
988   static bool inputUsesCRLF(StringRef Text) {
989     return Text.count('\r') * 2 > Text.count('\n');
990   }
991 
992   void
993   deriveLocalStyle(const SmallVectorImpl<AnnotatedLine *> &AnnotatedLines) {
994     unsigned CountBoundToVariable = 0;
995     unsigned CountBoundToType = 0;
996     bool HasCpp03IncompatibleFormat = false;
997     bool HasBinPackedFunction = false;
998     bool HasOnePerLineFunction = false;
999     for (unsigned i = 0, e = AnnotatedLines.size(); i != e; ++i) {
1000       if (!AnnotatedLines[i]->First->Next)
1001         continue;
1002       FormatToken *Tok = AnnotatedLines[i]->First->Next;
1003       while (Tok->Next) {
1004         if (Tok->Type == TT_PointerOrReference) {
1005           bool SpacesBefore =
1006               Tok->WhitespaceRange.getBegin() != Tok->WhitespaceRange.getEnd();
1007           bool SpacesAfter = Tok->Next->WhitespaceRange.getBegin() !=
1008                              Tok->Next->WhitespaceRange.getEnd();
1009           if (SpacesBefore && !SpacesAfter)
1010             ++CountBoundToVariable;
1011           else if (!SpacesBefore && SpacesAfter)
1012             ++CountBoundToType;
1013         }
1014 
1015         if (Tok->WhitespaceRange.getBegin() == Tok->WhitespaceRange.getEnd()) {
1016           if (Tok->is(tok::coloncolon) &&
1017               Tok->Previous->Type == TT_TemplateOpener)
1018             HasCpp03IncompatibleFormat = true;
1019           if (Tok->Type == TT_TemplateCloser &&
1020               Tok->Previous->Type == TT_TemplateCloser)
1021             HasCpp03IncompatibleFormat = true;
1022         }
1023 
1024         if (Tok->PackingKind == PPK_BinPacked)
1025           HasBinPackedFunction = true;
1026         if (Tok->PackingKind == PPK_OnePerLine)
1027           HasOnePerLineFunction = true;
1028 
1029         Tok = Tok->Next;
1030       }
1031     }
1032     if (Style.DerivePointerBinding) {
1033       if (CountBoundToType > CountBoundToVariable)
1034         Style.PointerBindsToType = true;
1035       else if (CountBoundToType < CountBoundToVariable)
1036         Style.PointerBindsToType = false;
1037     }
1038     if (Style.Standard == FormatStyle::LS_Auto) {
1039       Style.Standard = HasCpp03IncompatibleFormat ? FormatStyle::LS_Cpp11
1040                                                   : FormatStyle::LS_Cpp03;
1041     }
1042     BinPackInconclusiveFunctions =
1043         HasBinPackedFunction || !HasOnePerLineFunction;
1044   }
1045 
1046   /// \brief Get the indent of \p Level from \p IndentForLevel.
1047   ///
1048   /// \p IndentForLevel must contain the indent for the level \c l
1049   /// at \p IndentForLevel[l], or a value < 0 if the indent for
1050   /// that level is unknown.
1051   unsigned getIndent(const std::vector<int> IndentForLevel, unsigned Level) {
1052     if (IndentForLevel[Level] != -1)
1053       return IndentForLevel[Level];
1054     if (Level == 0)
1055       return 0;
1056     return getIndent(IndentForLevel, Level - 1) + Style.IndentWidth;
1057   }
1058 
1059   /// \brief Get the offset of the line relatively to the level.
1060   ///
1061   /// For example, 'public:' labels in classes are offset by 1 or 2
1062   /// characters to the left from their level.
1063   int getIndentOffset(const FormatToken &RootToken) {
1064     if (RootToken.isAccessSpecifier(false) || RootToken.isObjCAccessSpecifier())
1065       return Style.AccessModifierOffset;
1066     return 0;
1067   }
1068 
1069   /// \brief Tries to merge lines into one.
1070   ///
1071   /// This will change \c Line and \c AnnotatedLine to contain the merged line,
1072   /// if possible; note that \c I will be incremented when lines are merged.
1073   void tryFitMultipleLinesInOne(unsigned Indent,
1074                                 SmallVectorImpl<AnnotatedLine *>::iterator &I,
1075                                 SmallVectorImpl<AnnotatedLine *>::iterator E) {
1076     // We can never merge stuff if there are trailing line comments.
1077     AnnotatedLine *TheLine = *I;
1078     if (TheLine->Last->Type == TT_LineComment)
1079       return;
1080 
1081     if (Indent > Style.ColumnLimit)
1082       return;
1083 
1084     unsigned Limit = Style.ColumnLimit - Indent;
1085     // If we already exceed the column limit, we set 'Limit' to 0. The different
1086     // tryMerge..() functions can then decide whether to still do merging.
1087     Limit = TheLine->Last->TotalLength > Limit
1088                 ? 0
1089                 : Limit - TheLine->Last->TotalLength;
1090 
1091     if (I + 1 == E || (*(I + 1))->Type == LT_Invalid)
1092       return;
1093 
1094     if (TheLine->Last->is(tok::l_brace)) {
1095       tryMergeSimpleBlock(I, E, Limit);
1096     } else if (Style.AllowShortIfStatementsOnASingleLine &&
1097                TheLine->First->is(tok::kw_if)) {
1098       tryMergeSimpleControlStatement(I, E, Limit);
1099     } else if (Style.AllowShortLoopsOnASingleLine &&
1100                TheLine->First->isOneOf(tok::kw_for, tok::kw_while)) {
1101       tryMergeSimpleControlStatement(I, E, Limit);
1102     } else if (TheLine->InPPDirective && (TheLine->First->HasUnescapedNewline ||
1103                                           TheLine->First->IsFirst)) {
1104       tryMergeSimplePPDirective(I, E, Limit);
1105     }
1106   }
1107 
1108   void tryMergeSimplePPDirective(SmallVectorImpl<AnnotatedLine *>::iterator &I,
1109                                  SmallVectorImpl<AnnotatedLine *>::iterator E,
1110                                  unsigned Limit) {
1111     if (Limit == 0)
1112       return;
1113     AnnotatedLine &Line = **I;
1114     if (!(*(I + 1))->InPPDirective || (*(I + 1))->First->HasUnescapedNewline)
1115       return;
1116     if (I + 2 != E && (*(I + 2))->InPPDirective &&
1117         !(*(I + 2))->First->HasUnescapedNewline)
1118       return;
1119     if (1 + (*(I + 1))->Last->TotalLength > Limit)
1120       return;
1121     join(Line, **(++I));
1122   }
1123 
1124   void
1125   tryMergeSimpleControlStatement(SmallVectorImpl<AnnotatedLine *>::iterator &I,
1126                                  SmallVectorImpl<AnnotatedLine *>::iterator E,
1127                                  unsigned Limit) {
1128     if (Limit == 0)
1129       return;
1130     if (Style.BreakBeforeBraces == FormatStyle::BS_Allman &&
1131         (*(I + 1))->First->is(tok::l_brace))
1132       return;
1133     if ((*(I + 1))->InPPDirective != (*I)->InPPDirective ||
1134         ((*(I + 1))->InPPDirective && (*(I + 1))->First->HasUnescapedNewline))
1135       return;
1136     AnnotatedLine &Line = **I;
1137     if (Line.Last->isNot(tok::r_paren))
1138       return;
1139     if (1 + (*(I + 1))->Last->TotalLength > Limit)
1140       return;
1141     if ((*(I + 1))->First->isOneOf(tok::semi, tok::kw_if, tok::kw_for,
1142                                    tok::kw_while) ||
1143         (*(I + 1))->First->Type == TT_LineComment)
1144       return;
1145     // Only inline simple if's (no nested if or else).
1146     if (I + 2 != E && Line.First->is(tok::kw_if) &&
1147         (*(I + 2))->First->is(tok::kw_else))
1148       return;
1149     join(Line, **(++I));
1150   }
1151 
1152   void tryMergeSimpleBlock(SmallVectorImpl<AnnotatedLine *>::iterator &I,
1153                            SmallVectorImpl<AnnotatedLine *>::iterator E,
1154                            unsigned Limit) {
1155     // No merging if the brace already is on the next line.
1156     if (Style.BreakBeforeBraces != FormatStyle::BS_Attach)
1157       return;
1158 
1159     // First, check that the current line allows merging. This is the case if
1160     // we're not in a control flow statement and the last token is an opening
1161     // brace.
1162     AnnotatedLine &Line = **I;
1163     if (Line.First->isOneOf(tok::kw_if, tok::kw_while, tok::kw_do, tok::r_brace,
1164                             tok::kw_else, tok::kw_try, tok::kw_catch,
1165                             tok::kw_for,
1166                             // This gets rid of all ObjC @ keywords and methods.
1167                             tok::at, tok::minus, tok::plus))
1168       return;
1169 
1170     FormatToken *Tok = (*(I + 1))->First;
1171     if (Tok->is(tok::r_brace) && !Tok->MustBreakBefore &&
1172         (Tok->getNextNonComment() == NULL ||
1173          Tok->getNextNonComment()->is(tok::semi))) {
1174       // We merge empty blocks even if the line exceeds the column limit.
1175       Tok->SpacesRequiredBefore = 0;
1176       Tok->CanBreakBefore = true;
1177       join(Line, **(I + 1));
1178       I += 1;
1179     } else if (Limit != 0 && Line.First->isNot(tok::kw_namespace)) {
1180       // Check that we still have three lines and they fit into the limit.
1181       if (I + 2 == E || (*(I + 2))->Type == LT_Invalid ||
1182           !nextTwoLinesFitInto(I, Limit))
1183         return;
1184 
1185       // Second, check that the next line does not contain any braces - if it
1186       // does, readability declines when putting it into a single line.
1187       if ((*(I + 1))->Last->Type == TT_LineComment || Tok->MustBreakBefore)
1188         return;
1189       do {
1190         if (Tok->isOneOf(tok::l_brace, tok::r_brace))
1191           return;
1192         Tok = Tok->Next;
1193       } while (Tok != NULL);
1194 
1195       // Last, check that the third line contains a single closing brace.
1196       Tok = (*(I + 2))->First;
1197       if (Tok->getNextNonComment() != NULL || Tok->isNot(tok::r_brace) ||
1198           Tok->MustBreakBefore)
1199         return;
1200 
1201       join(Line, **(I + 1));
1202       join(Line, **(I + 2));
1203       I += 2;
1204     }
1205   }
1206 
1207   bool nextTwoLinesFitInto(SmallVectorImpl<AnnotatedLine *>::iterator I,
1208                            unsigned Limit) {
1209     return 1 + (*(I + 1))->Last->TotalLength + 1 +
1210                (*(I + 2))->Last->TotalLength <=
1211            Limit;
1212   }
1213 
1214   void join(AnnotatedLine &A, const AnnotatedLine &B) {
1215     assert(!A.Last->Next);
1216     assert(!B.First->Previous);
1217     A.Last->Next = B.First;
1218     B.First->Previous = A.Last;
1219     unsigned LengthA = A.Last->TotalLength + B.First->SpacesRequiredBefore;
1220     for (FormatToken *Tok = B.First; Tok; Tok = Tok->Next) {
1221       Tok->TotalLength += LengthA;
1222       A.Last = Tok;
1223     }
1224   }
1225 
1226   bool touchesRanges(const CharSourceRange &Range) {
1227     for (unsigned i = 0, e = Ranges.size(); i != e; ++i) {
1228       if (!SourceMgr.isBeforeInTranslationUnit(Range.getEnd(),
1229                                                Ranges[i].getBegin()) &&
1230           !SourceMgr.isBeforeInTranslationUnit(Ranges[i].getEnd(),
1231                                                Range.getBegin()))
1232         return true;
1233     }
1234     return false;
1235   }
1236 
1237   bool touchesLine(const AnnotatedLine &TheLine) {
1238     const FormatToken *First = TheLine.First;
1239     const FormatToken *Last = TheLine.Last;
1240     CharSourceRange LineRange = CharSourceRange::getCharRange(
1241         First->WhitespaceRange.getBegin().getLocWithOffset(
1242             First->LastNewlineOffset),
1243         Last->getStartOfNonWhitespace().getLocWithOffset(
1244             Last->TokenText.size() - 1));
1245     return touchesRanges(LineRange);
1246   }
1247 
1248   bool touchesPPDirective(SmallVectorImpl<AnnotatedLine *>::iterator I,
1249                           SmallVectorImpl<AnnotatedLine *>::iterator E) {
1250     for (; I != E; ++I) {
1251       if ((*I)->First->HasUnescapedNewline)
1252         return false;
1253       if (touchesLine(**I))
1254         return true;
1255     }
1256     return false;
1257   }
1258 
1259   bool touchesEmptyLineBefore(const AnnotatedLine &TheLine) {
1260     const FormatToken *First = TheLine.First;
1261     CharSourceRange LineRange = CharSourceRange::getCharRange(
1262         First->WhitespaceRange.getBegin(),
1263         First->WhitespaceRange.getBegin().getLocWithOffset(
1264             First->LastNewlineOffset));
1265     return touchesRanges(LineRange);
1266   }
1267 
1268   virtual void consumeUnwrappedLine(const UnwrappedLine &TheLine) {
1269     assert(!UnwrappedLines.empty());
1270     UnwrappedLines.back().push_back(TheLine);
1271   }
1272 
1273   virtual void finishRun() {
1274     UnwrappedLines.push_back(SmallVector<UnwrappedLine, 16>());
1275   }
1276 
1277   /// \brief Add a new line and the required indent before the first Token
1278   /// of the \c UnwrappedLine if there was no structural parsing error.
1279   void formatFirstToken(FormatToken &RootToken,
1280                         const AnnotatedLine *PreviousLine, unsigned IndentLevel,
1281                         unsigned Indent, bool InPPDirective) {
1282     unsigned Newlines =
1283         std::min(RootToken.NewlinesBefore, Style.MaxEmptyLinesToKeep + 1);
1284     // Remove empty lines before "}" where applicable.
1285     if (RootToken.is(tok::r_brace) &&
1286         (!RootToken.Next ||
1287          (RootToken.Next->is(tok::semi) && !RootToken.Next->Next)))
1288       Newlines = std::min(Newlines, 1u);
1289     if (Newlines == 0 && !RootToken.IsFirst)
1290       Newlines = 1;
1291 
1292     // Insert extra new line before access specifiers.
1293     if (PreviousLine && PreviousLine->Last->isOneOf(tok::semi, tok::r_brace) &&
1294         RootToken.isAccessSpecifier() && RootToken.NewlinesBefore == 1)
1295       ++Newlines;
1296 
1297     // Remove empty lines after access specifiers.
1298     if (PreviousLine && PreviousLine->First->isAccessSpecifier())
1299       Newlines = std::min(1u, Newlines);
1300 
1301     Whitespaces.replaceWhitespace(
1302         RootToken, Newlines, IndentLevel, Indent, Indent,
1303         InPPDirective && !RootToken.HasUnescapedNewline);
1304   }
1305 
1306   unsigned getColumnLimit(bool InPPDirective) const {
1307     // In preprocessor directives reserve two chars for trailing " \"
1308     return Style.ColumnLimit - (InPPDirective ? 2 : 0);
1309   }
1310 
1311   FormatStyle Style;
1312   Lexer &Lex;
1313   SourceManager &SourceMgr;
1314   WhitespaceManager Whitespaces;
1315   std::vector<CharSourceRange> Ranges;
1316   SmallVector<SmallVector<UnwrappedLine, 16>, 2> UnwrappedLines;
1317 
1318   encoding::Encoding Encoding;
1319   bool BinPackInconclusiveFunctions;
1320 };
1321 
1322 } // end anonymous namespace
1323 
1324 tooling::Replacements reformat(const FormatStyle &Style, Lexer &Lex,
1325                                SourceManager &SourceMgr,
1326                                std::vector<CharSourceRange> Ranges) {
1327   Formatter formatter(Style, Lex, SourceMgr, Ranges);
1328   return formatter.format();
1329 }
1330 
1331 tooling::Replacements reformat(const FormatStyle &Style, StringRef Code,
1332                                std::vector<tooling::Range> Ranges,
1333                                StringRef FileName) {
1334   FileManager Files((FileSystemOptions()));
1335   DiagnosticsEngine Diagnostics(
1336       IntrusiveRefCntPtr<DiagnosticIDs>(new DiagnosticIDs),
1337       new DiagnosticOptions);
1338   SourceManager SourceMgr(Diagnostics, Files);
1339   llvm::MemoryBuffer *Buf = llvm::MemoryBuffer::getMemBuffer(Code, FileName);
1340   const clang::FileEntry *Entry =
1341       Files.getVirtualFile(FileName, Buf->getBufferSize(), 0);
1342   SourceMgr.overrideFileContents(Entry, Buf);
1343   FileID ID =
1344       SourceMgr.createFileID(Entry, SourceLocation(), clang::SrcMgr::C_User);
1345   Lexer Lex(ID, SourceMgr.getBuffer(ID), SourceMgr,
1346             getFormattingLangOpts(Style.Standard));
1347   SourceLocation StartOfFile = SourceMgr.getLocForStartOfFile(ID);
1348   std::vector<CharSourceRange> CharRanges;
1349   for (unsigned i = 0, e = Ranges.size(); i != e; ++i) {
1350     SourceLocation Start = StartOfFile.getLocWithOffset(Ranges[i].getOffset());
1351     SourceLocation End = Start.getLocWithOffset(Ranges[i].getLength());
1352     CharRanges.push_back(CharSourceRange::getCharRange(Start, End));
1353   }
1354   return reformat(Style, Lex, SourceMgr, CharRanges);
1355 }
1356 
1357 LangOptions getFormattingLangOpts(FormatStyle::LanguageStandard Standard) {
1358   LangOptions LangOpts;
1359   LangOpts.CPlusPlus = 1;
1360   LangOpts.CPlusPlus11 = Standard == FormatStyle::LS_Cpp03 ? 0 : 1;
1361   LangOpts.LineComment = 1;
1362   LangOpts.Bool = 1;
1363   LangOpts.ObjC1 = 1;
1364   LangOpts.ObjC2 = 1;
1365   return LangOpts;
1366 }
1367 
1368 const char *StyleOptionHelpDescription =
1369     "Coding style, currently supports:\n"
1370     "  LLVM, Google, Chromium, Mozilla, WebKit.\n"
1371     "Use -style=file to load style configuration from\n"
1372     ".clang-format file located in one of the parent\n"
1373     "directories of the source file (or current\n"
1374     "directory for stdin).\n"
1375     "Use -style=\"{key: value, ...}\" to set specific\n"
1376     "parameters, e.g.:\n"
1377     "  -style=\"{BasedOnStyle: llvm, IndentWidth: 8}\"";
1378 
1379 FormatStyle getStyle(StringRef StyleName, StringRef FileName) {
1380   // Fallback style in case the rest of this function can't determine a style.
1381   StringRef FallbackStyle = "LLVM";
1382   FormatStyle Style;
1383   getPredefinedStyle(FallbackStyle, &Style);
1384 
1385   if (StyleName.startswith("{")) {
1386     // Parse YAML/JSON style from the command line.
1387     if (llvm::error_code ec = parseConfiguration(StyleName, &Style)) {
1388       llvm::errs() << "Error parsing -style: " << ec.message() << ", using "
1389                    << FallbackStyle << " style\n";
1390     }
1391     return Style;
1392   }
1393 
1394   if (!StyleName.equals_lower("file")) {
1395     if (!getPredefinedStyle(StyleName, &Style))
1396       llvm::errs() << "Invalid value for -style, using " << FallbackStyle
1397                    << " style\n";
1398     return Style;
1399   }
1400 
1401   SmallString<128> Path(FileName);
1402   llvm::sys::fs::make_absolute(Path);
1403   for (StringRef Directory = Path; !Directory.empty();
1404        Directory = llvm::sys::path::parent_path(Directory)) {
1405     if (!llvm::sys::fs::is_directory(Directory))
1406       continue;
1407     SmallString<128> ConfigFile(Directory);
1408 
1409     llvm::sys::path::append(ConfigFile, ".clang-format");
1410     DEBUG(llvm::dbgs() << "Trying " << ConfigFile << "...\n");
1411     bool IsFile = false;
1412     // Ignore errors from is_regular_file: we only need to know if we can read
1413     // the file or not.
1414     llvm::sys::fs::is_regular_file(Twine(ConfigFile), IsFile);
1415 
1416     if (!IsFile) {
1417       // Try _clang-format too, since dotfiles are not commonly used on Windows.
1418       ConfigFile = Directory;
1419       llvm::sys::path::append(ConfigFile, "_clang-format");
1420       DEBUG(llvm::dbgs() << "Trying " << ConfigFile << "...\n");
1421       llvm::sys::fs::is_regular_file(Twine(ConfigFile), IsFile);
1422     }
1423 
1424     if (IsFile) {
1425       OwningPtr<llvm::MemoryBuffer> Text;
1426       if (llvm::error_code ec = llvm::MemoryBuffer::getFile(ConfigFile, Text)) {
1427         llvm::errs() << ec.message() << "\n";
1428         continue;
1429       }
1430       if (llvm::error_code ec = parseConfiguration(Text->getBuffer(), &Style)) {
1431         llvm::errs() << "Error reading " << ConfigFile << ": " << ec.message()
1432                      << "\n";
1433         continue;
1434       }
1435       DEBUG(llvm::dbgs() << "Using configuration file " << ConfigFile << "\n");
1436       return Style;
1437     }
1438   }
1439   llvm::errs() << "Can't find usable .clang-format, using " << FallbackStyle
1440                << " style\n";
1441   return Style;
1442 }
1443 
1444 } // namespace format
1445 } // namespace clang
1446