[PATCH] D79388: [clang-format] Fix AlignConsecutive on PP blocks

[Jake Merdich via Phabricator via cfe-commits]

JakeMerdichAMD added a comment.

Belated 'sounds good to me'.

Other things have been keeping me busy, sorry for the delayed response on this.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D79388/new/

https://reviews.llvm.org/D79388

___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D79388: [clang-format] Fix AlignConsecutive on PP blocks

[Sylvestre Ledru via Phabricator via cfe-commits]

sylvestre.ledru added a comment.

@JakeMerdichAMD as the regression caused by this change is way more significant 
than the initial bug, I am planning to revert this patch in the next few days 
(both in master and 11).
Please let me know if you have any objection.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D79388/new/

https://reviews.llvm.org/D79388

___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D79388: [clang-format] Fix AlignConsecutive on PP blocks

[Andi via Phabricator via cfe-commits]

Abpostelnicu added a comment.

In D79388#2321231 , @MyDeveloperDay 
wrote:

> Not to my knowledge, if we are not going to fix it we should probably revert 
> it for now, we can bring it back later

I'm a strong supporter of reverting it for the moment.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D79388/new/

https://reviews.llvm.org/D79388

___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D79388: [clang-format] Fix AlignConsecutive on PP blocks

[MyDeveloperDay via Phabricator via cfe-commits]

MyDeveloperDay added a comment.

Not to my knowledge, if we are not going to fix it we should probably revert it 
for now


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D79388/new/

https://reviews.llvm.org/D79388

___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D79388: [clang-format] Fix AlignConsecutive on PP blocks

[Sylvestre Ledru via Phabricator via cfe-commits]

sylvestre.ledru added a comment.

@JakeMerdichAMD @MyDeveloperDay Do you know if the regression has been fixed?
thanks


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D79388/new/

https://reviews.llvm.org/D79388

___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D79388: [clang-format] Fix AlignConsecutive on PP blocks

[Jake Merdich via Phabricator via cfe-commits]

JakeMerdichAMD added a comment.

I'd be very surprised if any of the tests included in this change pass with 
that line commented it's meant so that things like #if and #else properly 
separate alignment after the first preprocessor run, where the whitespace 
manager doesn't have the full context of things between it.

This will break with any PP statement though, not just if-elif-else-endif. I'll 
try moving it to somewhere more specific like the cases in parsePPDirective and 
put out a fix.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D79388/new/

https://reviews.llvm.org/D79388

___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D79388: [clang-format] Fix AlignConsecutive on PP blocks

[MyDeveloperDay via Phabricator via cfe-commits]

MyDeveloperDay added inline comments.



Comment at: clang/lib/Format/UnwrappedLineParser.cpp:2996
 Line->Tokens.back().Tok->MustBreakBefore = true;
+Line->Tokens.back().Tok->MustBreakAlignBefore = true;
 MustBreakBeforeNextToken = false;

MyDeveloperDay wrote:
> If the line ends with a comment and we have align trailing comments turned on 
> then I think this breaks the alignment
If I comment out this line, then all the tests pass! and  
https://bugs.llvm.org/show_bug.cgi?id=47589 is resolved. I think we need to 
understand why we are putting this here?

```
void UnwrappedLineParser::pushToken(FormatToken *Tok) {
  Line->Tokens.push_back(UnwrappedLineNode(Tok));
  if (MustBreakBeforeNextToken) {
Line->Tokens.back().Tok->MustBreakBefore = true;
// Line->Tokens.back().Tok->MustBreakAlignBefore = true;
MustBreakBeforeNextToken = false;
  }
}
```



Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D79388/new/

https://reviews.llvm.org/D79388

___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D79388: [clang-format] Fix AlignConsecutive on PP blocks

[MyDeveloperDay via Phabricator via cfe-commits]

MyDeveloperDay added inline comments.



Comment at: clang/lib/Format/UnwrappedLineParser.cpp:2996
 Line->Tokens.back().Tok->MustBreakBefore = true;
+Line->Tokens.back().Tok->MustBreakAlignBefore = true;
 MustBreakBeforeNextToken = false;

If the line ends with a comment and we have align trailing comments turned on 
then I think this breaks the alignment


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D79388/new/

https://reviews.llvm.org/D79388

___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D79388: [clang-format] Fix AlignConsecutive on PP blocks

[MyDeveloperDay via Phabricator via cfe-commits]

MyDeveloperDay added a comment.

@JakeMerdichAMD  this has caused a regression 
https://bugs.llvm.org/show_bug.cgi?id=47589


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D79388/new/

https://reviews.llvm.org/D79388

___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D79388: [clang-format] Fix AlignConsecutive on PP blocks

[Jake Merdich via Phabricator via cfe-commits]

JakeMerdichAMD added a comment.

Thanks for the commit and review @MyDeveloperDay!


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D79388/new/

https://reviews.llvm.org/D79388



___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D79388: [clang-format] Fix AlignConsecutive on PP blocks

[MyDeveloperDay via Phabricator via cfe-commits]

This revision was automatically updated to reflect the committed changes.
Closed by commit rGb2eb43931757: [clang-format] Fix AlignConsecutive on PP 
blocks (authored by MyDeveloperDay).

Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D79388/new/

https://reviews.llvm.org/D79388

Files:
  clang/lib/Format/FormatToken.h
  clang/lib/Format/UnwrappedLineParser.cpp
  clang/lib/Format/WhitespaceManager.cpp
  clang/unittests/Format/FormatTest.cpp
  clang/unittests/Format/FormatTestComments.cpp

Index: clang/unittests/Format/FormatTestComments.cpp
===
--- clang/unittests/Format/FormatTestComments.cpp
+++ clang/unittests/Format/FormatTestComments.cpp
@@ -2780,6 +2780,27 @@
"   // line 2 about b\n"
"   long b;",
getLLVMStyleWithColumns(80)));
+
+  // Checks an edge case in preprocessor handling.
+  // These comments should *not* be aligned
+  EXPECT_EQ(
+  "#if FOO\n"
+  "#else\n"
+  "long a; // Line about a\n"
+  "#endif\n"
+  "#if BAR\n"
+  "#else\n"
+  "long b_long_name; // Line about b\n"
+  "#endif\n",
+  format("#if FOO\n"
+ "#else\n"
+ "long a;   // Line about a\n" // Previous (bad) behavior
+ "#endif\n"
+ "#if BAR\n"
+ "#else\n"
+ "long b_long_name; // Line about b\n"
+ "#endif\n",
+ getLLVMStyleWithColumns(80)));
 }
 
 TEST_F(FormatTestComments, AlignsBlockCommentDecorations) {
Index: clang/unittests/Format/FormatTest.cpp
===
--- clang/unittests/Format/FormatTest.cpp
+++ clang/unittests/Format/FormatTest.cpp
@@ -11457,6 +11457,29 @@
   verifyFormat("int oneTwoThree = 123; // comment\n"
"int oneTwo  = 12;  // comment",
Alignment);
+
+  // Bug 25167
+  verifyFormat("#if A\n"
+   "#else\n"
+   "int  = 12;\n"
+   "#endif\n"
+   "#if B\n"
+   "#else\n"
+   "int a = 12;\n"
+   "#endif\n",
+   Alignment);
+  verifyFormat("enum foo {\n"
+   "#if A\n"
+   "#else\n"
+   "   = 12;\n"
+   "#endif\n"
+   "#if B\n"
+   "#else\n"
+   "  a = 12;\n"
+   "#endif\n"
+   "};\n",
+   Alignment);
+
   EXPECT_EQ("int a = 5;\n"
 "\n"
 "int oneTwoThree = 123;",
Index: clang/lib/Format/WhitespaceManager.cpp
===
--- clang/lib/Format/WhitespaceManager.cpp
+++ clang/lib/Format/WhitespaceManager.cpp
@@ -377,9 +377,11 @@
 if (Changes[i].NewlinesBefore != 0) {
   CommasBeforeMatch = 0;
   EndOfSequence = i;
-  // If there is a blank line, or if the last line didn't contain any
-  // matching token, the sequence ends here.
-  if (Changes[i].NewlinesBefore > 1 || !FoundMatchOnLine)
+  // If there is a blank line, there is a forced-align-break (eg,
+  // preprocessor), or if the last line didn't contain any matching token,
+  // the sequence ends here.
+  if (Changes[i].NewlinesBefore > 1 ||
+  Changes[i].Tok->MustBreakAlignBefore || !FoundMatchOnLine)
 AlignCurrentSequence();
 
   FoundMatchOnLine = false;
@@ -618,6 +620,8 @@
 if (Changes[i].StartOfBlockComment)
   continue;
 Newlines += Changes[i].NewlinesBefore;
+if (Changes[i].Tok->MustBreakAlignBefore)
+  BreakBeforeNext = true;
 if (!Changes[i].IsTrailingComment)
   continue;
 
Index: clang/lib/Format/UnwrappedLineParser.cpp
===
--- clang/lib/Format/UnwrappedLineParser.cpp
+++ clang/lib/Format/UnwrappedLineParser.cpp
@@ -2968,6 +2968,7 @@
   }
   FormatTok = Tokens->getNextToken();
   FormatTok->MustBreakBefore = true;
+  FormatTok->MustBreakAlignBefore = true;
 }
 
 if (!PPStack.empty() && (PPStack.back().Kind == PP_Unreachable) &&
@@ -2992,6 +2993,7 @@
   Line->Tokens.push_back(UnwrappedLineNode(Tok));
   if (MustBreakBeforeNextToken) {
 Line->Tokens.back().Tok->MustBreakBefore = true;
+Line->Tokens.back().Tok->MustBreakAlignBefore = true;
 MustBreakBeforeNextToken = false;
   }
 }
Index: clang/lib/Format/FormatToken.h
===
--- clang/lib/Format/FormatToken.h
+++ clang/lib/Format/FormatToken.h
@@ -182,6 +182,12 @@
   /// before the token.
   bool MustBreakBefore = false;
 
+  /// Whether to not align across this token
+  ///
+  /// This happens for example when a preprocessor directive ended directly
+  /// before the token, but very rarely otherwise.
+  bool MustBreakAlignBefore = false;
+
   /// The raw text of 

[PATCH] D79388: [clang-format] Fix AlignConsecutive on PP blocks

[Jake Merdich via Phabricator via cfe-commits]

JakeMerdichAMD added a comment.

Hey, @MyDeveloperDay, can I get your assistance in committing this? It's 
probably been long enough for anyone to chime in.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D79388/new/

https://reviews.llvm.org/D79388



___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D79388: [clang-format] Fix AlignConsecutive on PP blocks

[MyDeveloperDay via Phabricator via cfe-commits]

MyDeveloperDay added a comment.

In D79388#2020320 , @JakeMerdichAMD 
wrote:

> Sure, I'll get started on that. It mainly comes from charging headfirst into 
> the edge cases, but I think I have a decent grasp of clang-format internals 
> now, and I'm definitely interested in both contributing to and reviewing for 
> it.


I'm going to start adding you to reviews, its nice to be able to increase the 
pool of reviewers we can use, I tend to work on the small defects from the bug 
tracker and I feel bad about over burdening people.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D79388/new/

https://reviews.llvm.org/D79388



___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D79388: [clang-format] Fix AlignConsecutive on PP blocks

[Jake Merdich via Phabricator via cfe-commits]

JakeMerdichAMD added a comment.

Sure, I'll get started on that. It mainly comes from charging headfirst into 
the edge cases, but I think I have a decent grasp of clang-format internals 
now, and I'm definitely interested in both contributing to and reviewing for it.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D79388/new/

https://reviews.llvm.org/D79388



___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D79388: [clang-format] Fix AlignConsecutive on PP blocks

[MyDeveloperDay via Phabricator via cfe-commits]

MyDeveloperDay added a comment.

In D79388#2019985 , @JakeMerdichAMD 
wrote:

> @MyDeveloperDay, you're right about what this issue addresses: it surprised 
> me a lot when an unrelated edit caused something to 'randomly' add spaces 
> elsewhere, since it's in a different ifdef block (whether or not it has the 
> same condition). The code that's aligned should always be across lines that 
> are visually consecutive, and those ones weren't (in most cases, they aren't 
> even logically consecutive).
>
> You're also right that I don't have commit access; this is my first commit in 
> LLVM.
>
> > I'm struggling with the simplicity of your solution ;-) (that's a 
> > compliment!)
>
> Thank you!


You are welcome, Given your obvious level of understanding could I encourage 
you to gain commit access? (its a relative simple process)

The reason I ask is we need people able to review and accept code reviews (this 
requires commit access), I wonder if you'd be interested in participating in 
future reviews?

Its always nice to have new people on board.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D79388/new/

https://reviews.llvm.org/D79388



___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D79388: [clang-format] Fix AlignConsecutive on PP blocks

[Jake Merdich via Phabricator via cfe-commits]

JakeMerdichAMD added a comment.

@MyDeveloperDay, you're right about what this issue addresses: it surprised me 
a lot when an unrelated edit caused something to 'randomly' add spaces 
elsewhere, since it's in a different ifdef block (whether or not it has the 
same condition). The code that's aligned should always be across lines that are 
visually consecutive, and those ones weren't (in most cases, they aren't even 
logically consecutive).

You're also right that I don't have commit access; this is my first commit in 
LLVM.

> I'm struggling with the simplicity of your solution ;-) (that's a compliment!)

Thank you!


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D79388/new/

https://reviews.llvm.org/D79388



___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D79388: [clang-format] Fix AlignConsecutive on PP blocks

[MyDeveloperDay via Phabricator via cfe-commits]

MyDeveloperDay accepted this revision.
MyDeveloperDay added a subscriber: VelocityRa.
MyDeveloperDay added a comment.
This revision is now accepted and ready to land.

I'm struggling with the simplicity of your solution ;-)  (that's a compliment!)

I've pulled this patch and built it locally. It certainly had no detrimental 
effects on the sources I normally check clang-format against, and in some quick 
tests

  #if FOO
  int a  = 12;
  int bb = 12;
  #else
  int a   = 12;
  #endif
  
  #if BAR
  int aaavaaa = 12; // ABC
  #else
  int a   = 12; // ABC
  #endif
  
  #if BAZ
  int aaavaaa = 12;
  #if FOO
  int aavaaa = 12;
  #endif
  #else
  int a   = 12;
  #endif
  
  #if A
  #else
  int = 12; // ABC
  #endif
  #if B
  #else
  int a   = 12; // ABC
  #endif
  #if B
  #else
  int verylongnamethatshouldalign = 12; // ABC
  #endif

Looks to be transformed into:

  #if FOO
  int a  = 12;
  int bb = 12;
  #else
  int a = 12;
  #endif
  
  #if BAR
  int aaavaaa = 12; // ABC
  #else
  int a = 12; // ABC
  #endif
  
  #if BAZ
  int aaavaaa = 12;
  #if FOO
  int aavaaa = 12;
  #endif
  #else
  int a = 12;
  #endif
  
  #if A
  #else
  int  = 12; // ABC
  #endif
  #if B
  #else
  int a = 12; // ABC
  #endif
  #if B
  #else
  int verylongnamethatshouldalign = 12; // ABC
  #endif

Which I think is your point, that the alignment shouldn't breach #endif and it 
definitely shouldn't inherit the alignment from something further down. (if 
I've understood correctly)

From my perspective this LGTM, I'm impressed.

You may want to give some others a couple of days. (I'm guessing you don't have 
commit access correct?)

I'd like to add @VelocityRa & @enyquist  as they wrote the 
AlignConsecutiveMacros work and probably have a greater insight than me into 
the problems around this area.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D79388/new/

https://reviews.llvm.org/D79388



___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D79388: [clang-format] Fix AlignConsecutive on PP blocks

[Jake Merdich via Phabricator via cfe-commits]

JakeMerdichAMD created this revision.
Herald added a project: clang.
Herald added a subscriber: cfe-commits.

Currently the 'AlignConsecutive*' options incorrectly align across
elif and else statements, even if they are very far away and across
unrelated preprocessor macros.

This failed since on preprocessor run 2+, there is not enough context
about the #ifdefs to actually differentiate one block from another,
causing them to align across different blocks or even large sections of
the file.

Eg, with AlignConsecutiveAssignments:

  \#if FOO  // Run 1
  \#else// Run 1
  int a   = 1;  // Run 2, wrong
  \#endif   // Run 1
  
  \#if FOO  // Run 1
  \#else// Run 1
  int bar = 1;  // Run 2
  \#endif   // Run 1

is read as

  int a   = 1;  // Run 2, wrong
  int bar = 1;  // Run 2

The approach taken to fix this was to add a new flag to Token that
forces breaking alignment across groups of lines (MustBreakAlignBefore)
in a similar manner to the existing flag that forces a line break
(MustBreakBefore). This flag is set for the first Token after a
preprocessor statement or diff conflict marker.

Possible alternatives might be hashing preprocessor state to detect if
two lines come from the same block (but the way that ifdefs are
sometimes deferred makes that tricky) or showing the preprocessor
statements on all passes instead of just the first one (seems harder).

Fixes #25167,#31281


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D79388

Files:
  clang/lib/Format/FormatToken.h
  clang/lib/Format/UnwrappedLineParser.cpp
  clang/lib/Format/WhitespaceManager.cpp
  clang/unittests/Format/FormatTest.cpp
  clang/unittests/Format/FormatTestComments.cpp

Index: clang/unittests/Format/FormatTestComments.cpp
===
--- clang/unittests/Format/FormatTestComments.cpp
+++ clang/unittests/Format/FormatTestComments.cpp
@@ -2780,6 +2780,27 @@
"   // line 2 about b\n"
"   long b;",
getLLVMStyleWithColumns(80)));
+
+  // Checks an edge case in preprocessor handling.
+  // These comments should *not* be aligned
+  EXPECT_EQ(
+  "#if FOO\n"
+  "#else\n"
+  "long a; // Line about a\n"
+  "#endif\n"
+  "#if BAR\n"
+  "#else\n"
+  "long b_long_name; // Line about b\n"
+  "#endif\n",
+  format("#if FOO\n"
+ "#else\n"
+ "long a;   // Line about a\n" // Previous (bad) behavior
+ "#endif\n"
+ "#if BAR\n"
+ "#else\n"
+ "long b_long_name; // Line about b\n"
+ "#endif\n",
+ getLLVMStyleWithColumns(80)));
 }
 
 TEST_F(FormatTestComments, AlignsBlockCommentDecorations) {
Index: clang/unittests/Format/FormatTest.cpp
===
--- clang/unittests/Format/FormatTest.cpp
+++ clang/unittests/Format/FormatTest.cpp
@@ -11429,6 +11429,29 @@
   verifyFormat("int oneTwoThree = 123; // comment\n"
"int oneTwo  = 12;  // comment",
Alignment);
+
+  // Bug 25167
+  verifyFormat("#if A\n"
+   "#else\n"
+   "int  = 12;\n"
+   "#endif\n"
+   "#if B\n"
+   "#else\n"
+   "int a = 12;\n"
+   "#endif\n",
+   Alignment);
+  verifyFormat("enum foo {\n"
+   "#if A\n"
+   "#else\n"
+   "   = 12;\n"
+   "#endif\n"
+   "#if B\n"
+   "#else\n"
+   "  a = 12;\n"
+   "#endif\n"
+   "};\n",
+   Alignment);
+
   EXPECT_EQ("int a = 5;\n"
 "\n"
 "int oneTwoThree = 123;",
Index: clang/lib/Format/WhitespaceManager.cpp
===
--- clang/lib/Format/WhitespaceManager.cpp
+++ clang/lib/Format/WhitespaceManager.cpp
@@ -377,9 +377,11 @@
 if (Changes[i].NewlinesBefore != 0) {
   CommasBeforeMatch = 0;
   EndOfSequence = i;
-  // If there is a blank line, or if the last line didn't contain any
-  // matching token, the sequence ends here.
-  if (Changes[i].NewlinesBefore > 1 || !FoundMatchOnLine)
+  // If there is a blank line, there is a forced-align-break (eg,
+  // preprocessor), or if the last line didn't contain any matching token,
+  // the sequence ends here.
+  if (Changes[i].NewlinesBefore > 1 ||
+  Changes[i].Tok->MustBreakAlignBefore || !FoundMatchOnLine)
 AlignCurrentSequence();
 
   FoundMatchOnLine = false;
@@ -618,6 +620,8 @@
 if (Changes[i].StartOfBlockComment)
   continue;
 Newlines += Changes[i].NewlinesBefore;
+if (Changes[i].Tok->MustBreakAlignBefore)
+  BreakBeforeNext = true;
 if (!Changes[i].IsTrailingComment)
   continue;
 
Index: