1 //===--- NamespaceEndCommentsFixer.cpp --------------------------*- 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 /// \brief This file implements NamespaceEndCommentsFixer, a TokenAnalyzer that
12 /// fixes namespace end comments.
13 ///
14 //===----------------------------------------------------------------------===//
15 
16 #include "NamespaceEndCommentsFixer.h"
17 #include "llvm/Support/Debug.h"
18 #include "llvm/Support/Regex.h"
19 
20 #define DEBUG_TYPE "namespace-end-comments-fixer"
21 
22 namespace clang {
23 namespace format {
24 
25 namespace {
26 // The maximal number of unwrapped lines that a short namespace spans.
27 // Short namespaces don't need an end comment.
28 static const int kShortNamespaceMaxLines = 1;
29 
30 // Matches a valid namespace end comment.
31 // Valid namespace end comments don't need to be edited.
32 static llvm::Regex kNamespaceCommentPattern =
33     llvm::Regex("^/[/*] *(end (of )?)? *(anonymous|unnamed)? *"
34                 "namespace( +([a-zA-Z0-9:_]+))?\\.? *(\\*/)?$",
35                 llvm::Regex::IgnoreCase);
36 
37 // Computes the name of a namespace given the namespace token.
38 // Returns "" for anonymous namespace.
39 std::string computeName(const FormatToken *NamespaceTok) {
40   assert(NamespaceTok && NamespaceTok->is(tok::kw_namespace) &&
41          "expecting a namespace token");
42   std::string name = "";
43   // Collects all the non-comment tokens between 'namespace' and '{'.
44   const FormatToken *Tok = NamespaceTok->getNextNonComment();
45   while (Tok && !Tok->is(tok::l_brace)) {
46     name += Tok->TokenText;
47     Tok = Tok->getNextNonComment();
48   }
49   return name;
50 }
51 
52 std::string computeEndCommentText(StringRef NamespaceName, bool AddNewline) {
53   std::string text = "// namespace";
54   if (!NamespaceName.empty()) {
55     text += ' ';
56     text += NamespaceName;
57   }
58   if (AddNewline)
59     text += '\n';
60   return text;
61 }
62 
63 bool hasEndComment(const FormatToken *RBraceTok) {
64   return RBraceTok->Next && RBraceTok->Next->is(tok::comment);
65 }
66 
67 bool validEndComment(const FormatToken *RBraceTok, StringRef NamespaceName) {
68   assert(hasEndComment(RBraceTok));
69   const FormatToken *Comment = RBraceTok->Next;
70   SmallVector<StringRef, 7> Groups;
71   if (kNamespaceCommentPattern.match(Comment->TokenText, &Groups)) {
72     StringRef NamespaceNameInComment = Groups.size() > 5 ? Groups[5] : "";
73     // Anonymous namespace comments must not mention a namespace name.
74     if (NamespaceName.empty() && !NamespaceNameInComment.empty())
75       return false;
76     StringRef AnonymousInComment = Groups.size() > 3 ? Groups[3] : "";
77     // Named namespace comments must not mention anonymous namespace.
78     if (!NamespaceName.empty() && !AnonymousInComment.empty())
79       return false;
80     return NamespaceNameInComment == NamespaceName;
81   }
82   return false;
83 }
84 
85 void addEndComment(const FormatToken *RBraceTok, StringRef EndCommentText,
86                    const SourceManager &SourceMgr,
87                    tooling::Replacements *Fixes) {
88   auto EndLoc = RBraceTok->Tok.getEndLoc();
89   auto Range = CharSourceRange::getCharRange(EndLoc, EndLoc);
90   auto Err = Fixes->add(tooling::Replacement(SourceMgr, Range, EndCommentText));
91   if (Err) {
92     llvm::errs() << "Error while adding namespace end comment: "
93                  << llvm::toString(std::move(Err)) << "\n";
94   }
95 }
96 
97 void updateEndComment(const FormatToken *RBraceTok, StringRef EndCommentText,
98                       const SourceManager &SourceMgr,
99                       tooling::Replacements *Fixes) {
100   assert(hasEndComment(RBraceTok));
101   const FormatToken *Comment = RBraceTok->Next;
102   auto Range = CharSourceRange::getCharRange(Comment->getStartOfNonWhitespace(),
103                                              Comment->Tok.getEndLoc());
104   auto Err = Fixes->add(tooling::Replacement(SourceMgr, Range, EndCommentText));
105   if (Err) {
106     llvm::errs() << "Error while updating namespace end comment: "
107                  << llvm::toString(std::move(Err)) << "\n";
108   }
109 }
110 } // namespace
111 
112 NamespaceEndCommentsFixer::NamespaceEndCommentsFixer(const Environment &Env,
113                                                      const FormatStyle &Style)
114     : TokenAnalyzer(Env, Style) {}
115 
116 tooling::Replacements NamespaceEndCommentsFixer::analyze(
117     TokenAnnotator &Annotator, SmallVectorImpl<AnnotatedLine *> &AnnotatedLines,
118     FormatTokenLexer &Tokens) {
119   const SourceManager &SourceMgr = Env.getSourceManager();
120   AffectedRangeMgr.computeAffectedLines(AnnotatedLines.begin(),
121                                         AnnotatedLines.end());
122   tooling::Replacements Fixes;
123   for (size_t I = 0, E = AnnotatedLines.size(); I != E; ++I) {
124     if (!AnnotatedLines[I]->Affected || AnnotatedLines[I]->InPPDirective ||
125         !AnnotatedLines[I]->startsWith(tok::r_brace))
126       continue;
127     const AnnotatedLine *EndLine = AnnotatedLines[I];
128     size_t StartLineIndex = EndLine->MatchingOpeningBlockLineIndex;
129     if (StartLineIndex == UnwrappedLine::kInvalidIndex)
130       continue;
131     assert(StartLineIndex < E);
132     const FormatToken *NamespaceTok = AnnotatedLines[StartLineIndex]->First;
133     // Detect "(inline)? namespace" in the beginning of a line.
134     if (NamespaceTok->is(tok::kw_inline))
135       NamespaceTok = NamespaceTok->getNextNonComment();
136     if (!NamespaceTok || NamespaceTok->isNot(tok::kw_namespace))
137       continue;
138     FormatToken *RBraceTok = EndLine->First;
139     if (RBraceTok->Finalized)
140       continue;
141     RBraceTok->Finalized = true;
142     const FormatToken *EndCommentPrevTok = RBraceTok;
143     // Namespaces often end with '};'. In that case, attach namespace end
144     // comments to the semicolon tokens.
145     if (RBraceTok->Next && RBraceTok->Next->is(tok::semi)) {
146       EndCommentPrevTok = RBraceTok->Next;
147     }
148     // The next token in the token stream after the place where the end comment
149     // token must be. This is either the next token on the current line or the
150     // first token on the next line.
151     const FormatToken *EndCommentNextTok = EndCommentPrevTok->Next;
152     if (EndCommentNextTok && EndCommentNextTok->is(tok::comment))
153       EndCommentNextTok = EndCommentNextTok->Next;
154     if (!EndCommentNextTok && I + 1 < E)
155       EndCommentNextTok = AnnotatedLines[I + 1]->First;
156     bool AddNewline = EndCommentNextTok &&
157                       EndCommentNextTok->NewlinesBefore == 0 &&
158                       EndCommentNextTok->isNot(tok::eof);
159     const std::string NamespaceName = computeName(NamespaceTok);
160     const std::string EndCommentText =
161         computeEndCommentText(NamespaceName, AddNewline);
162     if (!hasEndComment(EndCommentPrevTok)) {
163       bool isShort = I - StartLineIndex <= kShortNamespaceMaxLines + 1;
164       if (!isShort)
165         addEndComment(EndCommentPrevTok, EndCommentText, SourceMgr, &Fixes);
166       continue;
167     }
168     if (!validEndComment(EndCommentPrevTok, NamespaceName))
169       updateEndComment(EndCommentPrevTok, EndCommentText, SourceMgr, &Fixes);
170   }
171   return Fixes;
172 }
173 
174 } // namespace format
175 } // namespace clang
176