vendor/clang/clang-trunk-r300422

1 files changed, 248 insertions, 38 deletions

diff --git a/lib/Format/UnwrappedLineParser.cpp b/lib/Format/UnwrappedLineParser.cpp
index 8fc3b78aee010..5be68ad5c6b82 100644
--- a/lib/Format/UnwrappedLineParser.cpp
+++ b/lib/Format/UnwrappedLineParser.cpp
@@ -202,7 +202,8 @@ UnwrappedLineParser::UnwrappedLineParser(const FormatStyle &Style,
                                          ArrayRef<FormatToken *> Tokens,
                                          UnwrappedLineConsumer &Callback)
     : Line(new UnwrappedLine), MustBreakBeforeNextToken(false),
-      CurrentLines(&Lines), Style(Style), Keywords(Keywords), Tokens(nullptr),
+      CurrentLines(&Lines), Style(Style), Keywords(Keywords),
+      CommentPragmasRegex(Style.CommentPragmas), Tokens(nullptr),
       Callback(Callback), AllTokens(Tokens), PPBranchLevel(-1) {}
 
 void UnwrappedLineParser::reset() {
@@ -334,8 +335,11 @@ void UnwrappedLineParser::calculateBraceTypes(bool ExpectClassBody) {
     case tok::l_brace:
       if (Style.Language == FormatStyle::LK_JavaScript && PrevTok &&
           PrevTok->is(tok::colon))
-        // In TypeScript's TypeMemberLists, there can be semicolons between the
-        // individual members.
+        // A colon indicates this code is in a type, or a braced list following
+        // a label in an object literal ({a: {b: 1}}).
+        // The code below could be confused by semicolons between the individual
+        // members in a type member list, which would normally trigger BK_Block.
+        // In both cases, this must be parsed as an inline braced init.
         Tok->BlockKind = BK_BracedInit;
       else
         Tok->BlockKind = BK_Unknown;
@@ -424,6 +428,8 @@ void UnwrappedLineParser::parseBlock(bool MustBeDeclaration, bool AddLevel,
     parseParens();
 
   addUnwrappedLine();
+  size_t OpeningLineIndex =
+      Lines.empty() ? (UnwrappedLine::kInvalidIndex) : (Lines.size() - 1);
 
   ScopedDeclarationState DeclarationState(*Line, DeclarationScopeStack,
                                           MustBeDeclaration);
@@ -449,6 +455,7 @@ void UnwrappedLineParser::parseBlock(bool MustBeDeclaration, bool AddLevel,
   if (MunchSemi && FormatTok->Tok.is(tok::semi))
     nextToken();
   Line->Level = InitialLevel;
+  Line->MatchingOpeningBlockLineIndex = OpeningLineIndex;
 }
 
 static bool isGoogScope(const UnwrappedLine &Line) {
@@ -582,13 +589,14 @@ void UnwrappedLineParser::conditionalCompilationEnd() {
 }
 
 void UnwrappedLineParser::parsePPIf(bool IfDef) {
+  bool IfNDef = FormatTok->is(tok::pp_ifndef);
   nextToken();
-  bool IsLiteralFalse = (FormatTok->Tok.isLiteral() &&
-                         FormatTok->Tok.getLiteralData() != nullptr &&
-                         StringRef(FormatTok->Tok.getLiteralData(),
-                                   FormatTok->Tok.getLength()) == "0") ||
-                        FormatTok->Tok.is(tok::kw_false);
-  conditionalCompilationStart(!IfDef && IsLiteralFalse);
+  bool Unreachable = false;
+  if (!IfDef && (FormatTok->is(tok::kw_false) || FormatTok->TokenText == "0"))
+    Unreachable = true;
+  if (IfDef && !IfNDef && FormatTok->TokenText == "SWIG")
+    Unreachable = true;
+  conditionalCompilationStart(Unreachable);
   parsePPUnknown();
 }
 
@@ -746,8 +754,7 @@ void UnwrappedLineParser::readTokenWithJavaScriptASI() {
        Previous->isOneOf(tok::r_square, tok::r_paren, tok::plusplus,
                          tok::minusminus)))
     return addUnwrappedLine();
-  if ((PreviousMustBeValue || Previous->is(tok::r_brace)) &&
-      isJSDeclOrStmt(Keywords, Next))
+  if (PreviousMustBeValue && isJSDeclOrStmt(Keywords, Next))
     return addUnwrappedLine();
 }
 
@@ -909,7 +916,8 @@ void UnwrappedLineParser::parseStructuralElement() {
         return;
       }
     }
-    if (FormatTok->isOneOf(Keywords.kw_signals, Keywords.kw_qsignals,
+    if (Style.isCpp() &&
+        FormatTok->isOneOf(Keywords.kw_signals, Keywords.kw_qsignals,
                            Keywords.kw_slots, Keywords.kw_qslots)) {
       nextToken();
       if (FormatTok->is(tok::colon)) {
@@ -943,7 +951,7 @@ void UnwrappedLineParser::parseStructuralElement() {
       if (!parseEnum())
         break;
       // This only applies for C++.
-      if (Style.Language != FormatStyle::LK_Cpp) {
+      if (!Style.isCpp()) {
         addUnwrappedLine();
         return;
       }
@@ -1124,7 +1132,7 @@ void UnwrappedLineParser::parseStructuralElement() {
 }
 
 bool UnwrappedLineParser::tryToParseLambda() {
-  if (Style.Language != FormatStyle::LK_Cpp) {
+  if (!Style.isCpp()) {
     nextToken();
     return false;
   }
@@ -1298,6 +1306,12 @@ bool UnwrappedLineParser::parseBracedList(bool ContinueOnSemicolons) {
           continue;
         }
       }
+      if (FormatTok->is(tok::l_brace)) {
+        // Could be a method inside of a braced list `{a() { return 1; }}`.
+        if (tryToParseBracedList())
+          continue;
+        parseChildBlock();
+      }
     }
     switch (FormatTok->Tok.getKind()) {
     case tok::caret:
@@ -1309,12 +1323,6 @@ bool UnwrappedLineParser::parseBracedList(bool ContinueOnSemicolons) {
     case tok::l_square:
       tryToParseLambda();
       break;
-    case tok::l_brace:
-      // Assume there are no blocks inside a braced init list apart
-      // from the ones we explicitly parse out (like lambdas).
-      FormatTok->BlockKind = BK_BracedInit;
-      parseBracedList();
-      break;
     case tok::l_paren:
       parseParens();
       // JavaScript can just have free standing methods and getters/setters in
@@ -1325,6 +1333,12 @@ bool UnwrappedLineParser::parseBracedList(bool ContinueOnSemicolons) {
         break;
       }
       break;
+    case tok::l_brace:
+      // Assume there are no blocks inside a braced init list apart
+      // from the ones we explicitly parse out (like lambdas).
+      FormatTok->BlockKind = BK_BracedInit;
+      parseBracedList();
+      break;
     case tok::r_brace:
       nextToken();
       return !HasError;
@@ -1381,6 +1395,12 @@ void UnwrappedLineParser::parseParens() {
       if (FormatTok->Tok.is(tok::l_brace))
         parseBracedList();
       break;
+    case tok::kw_class:
+      if (Style.Language == FormatStyle::LK_JavaScript)
+        parseRecord(/*ParseAsExpr=*/true);
+      else
+        nextToken();
+      break;
     case tok::identifier:
       if (Style.Language == FormatStyle::LK_JavaScript &&
           (FormatTok->is(Keywords.kw_function) ||
@@ -1722,8 +1742,7 @@ bool UnwrappedLineParser::parseEnum() {
       nextToken();
       // If there are two identifiers in a row, this is likely an elaborate
       // return type. In Java, this can be "implements", etc.
-      if (Style.Language == FormatStyle::LK_Cpp &&
-          FormatTok->is(tok::identifier))
+      if (Style.isCpp() && FormatTok->is(tok::identifier))
         return false;
     }
   }
@@ -1819,7 +1838,7 @@ void UnwrappedLineParser::parseJavaEnumBody() {
   addUnwrappedLine();
 }
 
-void UnwrappedLineParser::parseRecord() {
+void UnwrappedLineParser::parseRecord(bool ParseAsExpr) {
   const FormatToken &InitialToken = *FormatTok;
   nextToken();
 
@@ -1863,11 +1882,15 @@ void UnwrappedLineParser::parseRecord() {
     }
   }
   if (FormatTok->Tok.is(tok::l_brace)) {
-    if (ShouldBreakBeforeBrace(Style, InitialToken))
-      addUnwrappedLine();
+    if (ParseAsExpr) {
+      parseChildBlock();
+    } else {
+      if (ShouldBreakBeforeBrace(Style, InitialToken))
+        addUnwrappedLine();
 
-    parseBlock(/*MustBeDeclaration=*/true, /*AddLevel=*/true,
-               /*MunchSemi=*/false);
+      parseBlock(/*MustBeDeclaration=*/true, /*AddLevel=*/true,
+                 /*MunchSemi=*/false);
+    }
   }
   // There is no addUnwrappedLine() here so that we fall through to parsing a
   // structural element afterwards. Thus, in "class A {} n, m;",
@@ -1999,7 +2022,9 @@ LLVM_ATTRIBUTE_UNUSED static void printDebugInfo(const UnwrappedLine &Line,
   for (std::list<UnwrappedLineNode>::const_iterator I = Line.Tokens.begin(),
                                                     E = Line.Tokens.end();
        I != E; ++I) {
-    llvm::dbgs() << I->Tok->Tok.getName() << "[" << I->Tok->Type << "] ";
+    llvm::dbgs() << I->Tok->Tok.getName() << "["
+                 << "T=" << I->Tok->Type
+                 << ", OC=" << I->Tok->OriginalColumn << "] ";
   }
   for (std::list<UnwrappedLineNode>::const_iterator I = Line.Tokens.begin(),
                                                     E = Line.Tokens.end();
@@ -2024,6 +2049,7 @@ void UnwrappedLineParser::addUnwrappedLine() {
   });
   CurrentLines->push_back(std::move(*Line));
   Line->Tokens.clear();
+  Line->MatchingOpeningBlockLineIndex = UnwrappedLine::kInvalidIndex;
   if (CurrentLines == &Lines && !PreprocessorDirectives.empty()) {
     CurrentLines->append(
         std::make_move_iterator(PreprocessorDirectives.begin()),
@@ -2039,13 +2065,139 @@ bool UnwrappedLineParser::isOnNewLine(const FormatToken &FormatTok) {
          FormatTok.NewlinesBefore > 0;
 }
 
+static bool isLineComment(const FormatToken &FormatTok) {
+  return FormatTok.is(tok::comment) &&
+         FormatTok.TokenText.startswith("//");
+}
+
+// Checks if \p FormatTok is a line comment that continues the line comment
+// section on \p Line.
+static bool continuesLineComment(const FormatToken &FormatTok,
+                                 const UnwrappedLine &Line,
+                                 llvm::Regex &CommentPragmasRegex) {
+  if (Line.Tokens.empty())
+    return false;
+
+  StringRef IndentContent = FormatTok.TokenText;
+  if (FormatTok.TokenText.startswith("//") ||
+      FormatTok.TokenText.startswith("/*"))
+    IndentContent = FormatTok.TokenText.substr(2);
+  if (CommentPragmasRegex.match(IndentContent))
+    return false;
+
+  // If Line starts with a line comment, then FormatTok continues the comment
+  // section if its original column is greater or equal to the original start
+  // column of the line.
+  //
+  // Define the min column token of a line as follows: if a line ends in '{' or
+  // contains a '{' followed by a line comment, then the min column token is
+  // that '{'. Otherwise, the min column token of the line is the first token of
+  // the line.
+  //
+  // If Line starts with a token other than a line comment, then FormatTok
+  // continues the comment section if its original column is greater than the
+  // original start column of the min column token of the line.
+  //
+  // For example, the second line comment continues the first in these cases:
+  //
+  // // first line
+  // // second line
+  //
+  // and:
+  //
+  // // first line
+  //  // second line
+  //
+  // and:
+  //
+  // int i; // first line
+  //  // second line
+  //
+  // and:
+  //
+  // do { // first line
+  //      // second line
+  //   int i;
+  // } while (true);
+  //
+  // and:
+  //
+  // enum {
+  //   a, // first line
+  //    // second line
+  //   b
+  // };
+  //
+  // The second line comment doesn't continue the first in these cases:
+  //
+  //   // first line
+  //  // second line
+  //
+  // and:
+  //
+  // int i; // first line
+  // // second line
+  //
+  // and:
+  //
+  // do { // first line
+  //   // second line
+  //   int i;
+  // } while (true);
+  //
+  // and:
+  //
+  // enum {
+  //   a, // first line
+  //   // second line
+  // };
+  const FormatToken *MinColumnToken = Line.Tokens.front().Tok;
+
+  // Scan for '{//'. If found, use the column of '{' as a min column for line
+  // comment section continuation.
+  const FormatToken *PreviousToken = nullptr;
+  for (const UnwrappedLineNode &Node : Line.Tokens) {
+    if (PreviousToken && PreviousToken->is(tok::l_brace) &&
+        isLineComment(*Node.Tok)) {
+      MinColumnToken = PreviousToken;
+      break;
+    }
+    PreviousToken = Node.Tok;
+
+    // Grab the last newline preceding a token in this unwrapped line.
+    if (Node.Tok->NewlinesBefore > 0) {
+      MinColumnToken = Node.Tok;
+    }
+  }
+  if (PreviousToken && PreviousToken->is(tok::l_brace)) {
+    MinColumnToken = PreviousToken;
+  }
+
+  unsigned MinContinueColumn =
+      MinColumnToken->OriginalColumn +
+      (isLineComment(*MinColumnToken) ? 0 : 1);
+  return isLineComment(FormatTok) && FormatTok.NewlinesBefore == 1 &&
+         isLineComment(*(Line.Tokens.back().Tok)) &&
+         FormatTok.OriginalColumn >= MinContinueColumn;
+}
+
 void UnwrappedLineParser::flushComments(bool NewlineBeforeNext) {
   bool JustComments = Line->Tokens.empty();
   for (SmallVectorImpl<FormatToken *>::const_iterator
            I = CommentsBeforeNextToken.begin(),
            E = CommentsBeforeNextToken.end();
        I != E; ++I) {
-    if (isOnNewLine(**I) && JustComments)
+    // Line comments that belong to the same line comment section are put on the
+    // same line since later we might want to reflow content between them.
+    // Additional fine-grained breaking of line comment sections is controlled
+    // by the class BreakableLineCommentSection in case it is desirable to keep
+    // several line comment sections in the same unwrapped line.
+    //
+    // FIXME: Consider putting separate line comment sections as children to the
+    // unwrapped line instead.
+    (*I)->ContinuesLineCommentSection =
+        continuesLineComment(**I, *Line, CommentPragmasRegex);
+    if (isOnNewLine(**I) && JustComments && !(*I)->ContinuesLineCommentSection)
       addUnwrappedLine();
     pushToken(*I);
   }
@@ -2073,13 +2225,71 @@ const FormatToken *UnwrappedLineParser::getPreviousToken() {
   return Line->Tokens.back().Tok;
 }
 
+void UnwrappedLineParser::distributeComments(
+    const SmallVectorImpl<FormatToken *> &Comments,
+    const FormatToken *NextTok) {
+  // Whether or not a line comment token continues a line is controlled by
+  // the method continuesLineComment, with the following caveat:
+  //
+  // Define a trail of Comments to be a nonempty proper postfix of Comments such
+  // that each comment line from the trail is aligned with the next token, if
+  // the next token exists. If a trail exists, the beginning of the maximal
+  // trail is marked as a start of a new comment section.
+  //
+  // For example in this code:
+  //
+  // int a; // line about a
+  //   // line 1 about b
+  //   // line 2 about b
+  //   int b;
+  //
+  // the two lines about b form a maximal trail, so there are two sections, the
+  // first one consisting of the single comment "// line about a" and the
+  // second one consisting of the next two comments.
+  if (Comments.empty())
+    return;
+  bool ShouldPushCommentsInCurrentLine = true;
+  bool HasTrailAlignedWithNextToken = false;
+  unsigned StartOfTrailAlignedWithNextToken = 0;
+  if (NextTok) {
+    // We are skipping the first element intentionally.
+    for (unsigned i = Comments.size() - 1; i > 0; --i) {
+      if (Comments[i]->OriginalColumn == NextTok->OriginalColumn) {
+        HasTrailAlignedWithNextToken = true;
+        StartOfTrailAlignedWithNextToken = i;
+      }
+    }
+  }
+  for (unsigned i = 0, e = Comments.size(); i < e; ++i) {
+    FormatToken *FormatTok = Comments[i];
+    if (HasTrailAlignedWithNextToken &&
+        i == StartOfTrailAlignedWithNextToken) {
+      FormatTok->ContinuesLineCommentSection = false;
+    } else {
+      FormatTok->ContinuesLineCommentSection =
+          continuesLineComment(*FormatTok, *Line, CommentPragmasRegex);
+    }
+    if (!FormatTok->ContinuesLineCommentSection &&
+        (isOnNewLine(*FormatTok) || FormatTok->IsFirst)) {
+      ShouldPushCommentsInCurrentLine = false;
+    }
+    if (ShouldPushCommentsInCurrentLine) {
+      pushToken(FormatTok);
+    } else {
+      CommentsBeforeNextToken.push_back(FormatTok);
+    }
+  }
+}
+
 void UnwrappedLineParser::readToken() {
-  bool CommentsInCurrentLine = true;
+  SmallVector<FormatToken *, 1> Comments;
   do {
     FormatTok = Tokens->getNextToken();
     assert(FormatTok);
     while (!Line->InPPDirective && FormatTok->Tok.is(tok::hash) &&
            (FormatTok->HasUnescapedNewline || FormatTok->IsFirst)) {
+      distributeComments(Comments, FormatTok);
+      Comments.clear();
       // If there is an unfinished unwrapped line, we flush the preprocessor
       // directives only after that unwrapped line was finished later.
       bool SwitchToPreprocessorLines = !Line->Tokens.empty();
@@ -2109,17 +2319,17 @@ void UnwrappedLineParser::readToken() {
       continue;
     }
 
-    if (!FormatTok->Tok.is(tok::comment))
+    if (!FormatTok->Tok.is(tok::comment)) {
+      distributeComments(Comments, FormatTok);
+      Comments.clear();
       return;
-    if (isOnNewLine(*FormatTok) || FormatTok->IsFirst) {
-      CommentsInCurrentLine = false;
-    }
-    if (CommentsInCurrentLine) {
-      pushToken(FormatTok);
-    } else {
-      CommentsBeforeNextToken.push_back(FormatTok);
     }
+
+    Comments.push_back(FormatTok);
   } while (!eof());
+
+  distributeComments(Comments, nullptr);
+  Comments.clear();
 }
 
 void UnwrappedLineParser::pushToken(FormatToken *Tok) {