Unified token breaking logic: support for line comments.
Summary:
Added BreakableLineComment, moved common code from
BreakableBlockComment to newly added BreakableComment. As a side-effect of the
rewrite, found another problem with escaped newlines and had to change
code which removes trailing whitespace from line comments not to break after
this patch.
Reviewers: klimek, djasper
Reviewed By: klimek
CC: cfe-commits
Differential Revision: http://llvm-reviews.chandlerc.com/D682
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@179693 91177308-0d34-0410-b5e6-96231b3b80d8
diff --git a/lib/Format/Format.cpp b/lib/Format/Format.cpp
index a847d4e..5c064a6 100644
--- a/lib/Format/Format.cpp
+++ b/lib/Format/Format.cpp
@@ -617,54 +617,47 @@
unsigned StartColumn = State.Column - Current.FormatTok.TokenLength;
if (Current.is(tok::string_literal)) {
// Only break up default narrow strings.
- const char *LiteralData = Current.FormatTok.Tok.getLiteralData();
+ const char *LiteralData = SourceMgr.getCharacterData(
+ Current.FormatTok.getStartOfNonWhitespace());
if (!LiteralData || *LiteralData != '"')
return 0;
- Token.reset(new BreakableStringLiteral(Current.FormatTok, StartColumn));
+ Token.reset(new BreakableStringLiteral(SourceMgr, Current.FormatTok,
+ StartColumn));
} else if (Current.Type == TT_BlockComment) {
BreakableBlockComment *BBC =
new BreakableBlockComment(SourceMgr, Current, StartColumn);
if (!DryRun)
BBC->alignLines(Whitespaces);
Token.reset(BBC);
+ } else if (Current.Type == TT_LineComment) {
+ Token.reset(new BreakableLineComment(SourceMgr, Current, StartColumn));
} else {
return 0;
}
- if (Token->getPrefixLength() + Token->getSuffixLength(0) >
- getColumnLimit()) {
- return 0;
- }
-
bool BreakInserted = false;
unsigned Penalty = 0;
for (unsigned LineIndex = 0; LineIndex < Token->getLineCount();
++LineIndex) {
- unsigned TokenLineSize = Token->getLineSize(LineIndex);
unsigned TailOffset = 0;
unsigned RemainingLength =
Token->getLineLengthAfterSplit(LineIndex, TailOffset);
while (RemainingLength > getColumnLimit()) {
- unsigned DecorationLength =
- RemainingLength - (TokenLineSize - TailOffset);
- if (DecorationLength + 1 > getColumnLimit()) {
- // Can't reduce line length by splitting here.
- break;
- }
BreakableToken::Split Split =
Token->getSplit(LineIndex, TailOffset, getColumnLimit());
if (Split.first == StringRef::npos)
break;
assert(Split.first != 0);
+ unsigned NewRemainingLength = Token->getLineLengthAfterSplit(
+ LineIndex, TailOffset + Split.first + Split.second);
+ if (NewRemainingLength >= RemainingLength)
+ break;
if (!DryRun) {
Token->insertBreak(LineIndex, TailOffset, Split, Line.InPPDirective,
Whitespaces);
}
TailOffset += Split.first + Split.second;
- unsigned NewRemainingLength =
- Token->getLineLengthAfterSplit(LineIndex, TailOffset);
- assert(NewRemainingLength < RemainingLength);
RemainingLength = NewRemainingLength;
Penalty += Style.PenaltyExcessCharacter;
BreakInserted = true;
@@ -925,6 +918,11 @@
// Now FormatTok is the next non-whitespace token.
FormatTok.TokenLength = Text.size();
+ if (FormatTok.Tok.is(tok::comment)) {
+ FormatTok.TrailingWhiteSpaceLength = Text.size() - Text.rtrim().size();
+ FormatTok.TokenLength -= FormatTok.TrailingWhiteSpaceLength;
+ }
+
// In case the token starts with escaped newlines, we want to
// take them into account as whitespace - this pattern is quite frequent
// in macro definitions.
@@ -951,11 +949,6 @@
GreaterStashed = true;
}
- // If we reformat comments, we remove trailing whitespace. Update the length
- // accordingly.
- if (FormatTok.Tok.is(tok::comment))
- FormatTok.TokenLength = Text.rtrim().size();
-
return FormatTok;
}