In order to end up with good solutions, clang-format needs to try
"all" combinations of line breaks, evaluate them and select the
best one. Before, we have done this using a DFS with memoization
and cut-off conditions. However, this approach is very limited
as shown by the huge static initializer in the attachment of
llvm.org/PR14959.
Instead, this new implementation uses a variant of Dijkstra's
algorithm to do a prioritized BFS over the solution space.
Some numbers:
lib/Format/TokenAnnotator.cpp: 1.5s -> 0.15s
Attachment of PR14959: 10min+ (didn't finish) -> 10s
No functional changes intended.
llvm-svn: 174166
1. Never avoid bin packing in static initializers as this can
lead to terrible results.
2. If an element has to be broken over multiple lines, break after
the following comma.
This should be a step forward, but there are still many cases
especially with nested static initializers that we handle badly.
More patches will follow.
llvm-svn: 174061
The style guide only forbids this for function declarations. So,
now
someFunction(
aaaaaaaaaaaaaaaaaaaa, aaaaaaaaaaaaaaaaa, aaaaaaaaaaaa);
Is allowed in Chromium mode.
llvm-svn: 173806
Before (in good cases):
for (auto aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
: aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa) {}
for (auto aaaaaaaaaaaaaaaaaaaa : aaaaaaaaaaaaaaaaaaaa(aaaaaaaaaaaaaaaaaaaa,
aaaa)) {}
After:
for (auto aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa :
aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa) {}
for (auto aaaaaaaaaaaaaaaaaaaa :
aaaaaaaaaaaaaaaaaaaa(aaaaaaaaaaaaaaaaaaaa, aaaa)) {}
llvm-svn: 173684
Before we did not really systematically format those. Now, we format the
different cases as:
- 1 Line: a ? b : c;
- 2 Lines: short ? loooooooooong
: loooooooooong
- 2 Lines: loooooooooooooooong
? short : short
- 3 Lines: loooooooooooooooong
? loooooooooooooong
: loooooooooooooong
Not sure whether "?" and ":" should go on the new line, but it seems to
be the most consistent approach.
llvm-svn: 173683
1. Use a hanging ident for function calls nested in binary expressions.
E.g.:
int aaaaa = aaaaaaaaa && aaaaaaaaaa(
aaaaaaaaaa);
2. Slightly improve heuristic for builder type expressions and reduce
penalty for breaking before "." and "->" in those.
3. Remove mostly obsolete metric of decreasing indent level. This
fixes: llvm.org/PR14931.
Changes #1 and #2 were necessary to keep tests passing after #3.
llvm-svn: 173680
These always represent a continuation and we should increase the ident.
Before:
aaaaaaaaa(aaaaaaaaaaaaaaaaaaaaa::
aaaaaaaaaaaaaaaaaaaa);
After:
aaaaaaaaa(aaaaaaaaaaaaaaaaaaaaa::
aaaaaaaaaaaaaaaaaaaa);
llvm-svn: 173675
This combines two small changes:
1) Put a penalty on breaking after "<"
2) Only produce a hanging indent when parameters are separated by
commas.
Before:
aaaaaaaaaaaaaaaaaaaaaaaa<
aaaaaaaaaaaaaaaaa, aaaaaaaaaaaaaaaaa>(aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa);
aaaaaa(new Aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa(
aaaaaaaaaaaaaaaaaaaaaaaaaaaaa));
After:
aaaaaaaaaaaaaaaaaaaaaaaa<aaaaaaaaaaaaaaaaa, aaaaaaaaaaaaaaaaa>(
aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa);
aaaaaa(new Aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa(
aaaaaaaaaaaaaaaaaaaaaaaaaaaaa));
This changes one ObjC test, but AFAICT this is not according to any
style guide (neither before nor after). We probably should be aligning
on the ":" there according to:
http://google-styleguide.googlecode.com/svn/trunk/objcguide.xml?showone=Method_Invocations#Method_Invocations
llvm-svn: 173457
Otherwise, really long nested name specifiers can easily lead to a
violation of the column limit.
Not sure about the rules for indentation in those cases, so input is
appreciated (see tests.).
llvm-svn: 173438
Before:
int aaaa = aaaaa().aaaaa() // force break
.aaaaa();
After:
int aaaa = aaaaa().aaaaa() // force break
.aaaaa();
The other indent is just wrong and confusing.
llvm-svn: 173273
Before:
bool aaaa = aaaaaaaaaaa(
aaaaaaaaaaaaaaaaa);
After:
bool aaaa = aaaaaaaaaaa(
aaaaaaaaaaaaaaaaa);
The other indentation was a nice attempt but doesn't work in many cases.
Not sure what the right long term solution is as the "After: " is still
not nice. We either need to figure out what to do in the cases where it
"doesn't work" or come up with a third solution, e.g. falling back to:
bool aaaa =
aaaaaaaaaaa(
aaaaaaaaaaaaaaaaa);
which should always work and nicely highlight the structure.
llvm-svn: 173268
Layouting would prevent breaking before + in
a[b + c] = d;
Regression detected by code review.
Also fixes an invalid-read found by the valgrind bot.
llvm-svn: 173262
Having seen more cases, this actually was not a good thing to do in the
first place. We can still improve on what we do now, but breaking after
the "=" is good in many cases.
Before:
aaaaaaaaaaaaa = aa->aaaaaaaaaaaaaaaaaaaa(
aaaaaaaaaaaaaaaaaaaaaaaa(aaaaaaaaaaaaaa));
After:
aaaaaaaaaaaaa =
aa->aaaaaaaaaaaaaaaaaaaa(aaaaaaaaaaaaaaaaaaaaaaaa(aaaaaaaaaaaaaa));
llvm-svn: 173257
Before: if (int * a = &b) ...
After: if (int *a = &b) ...
Also changed all the existing tests to test the expressions in question
both in a declaration and in an expression context.
llvm-svn: 173256
We will need a more principled solution, but we should not leave this
unfixed until we come up with one.
Before: void f() { int * a; }
After: void f() { int *a; }
llvm-svn: 173252
This only affects styles where BinPackParameters is false.
With AllowAllParametersOnNextLine:
aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa(
aaaaaaaaaa, aaaaaaaaaa, aaaaaaaaaa, aaaaaaaaaaa, aaaaaaaaaaa);
Without it:
aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa(aaaaaaaaaa,
aaaaaaaaaa,
aaaaaaaaaa,
aaaaaaaaaaa,
aaaaaaaaaaa);
llvm-svn: 173246
This gives us the ability to guess better defaults for whether a *
between identifiers is a pointer dereference or binary operator.
Now correctly formats:
void f(a *b);
void f() { f(a * b); }
llvm-svn: 173243
Changing nextToken() in the UnwrappedLineParser to get the next
non-comment token. This allows us to correctly layout a whole class of
snippets, like:
if /* */(/* */ a /* */) /* */
f() /* */; /* */
else /* */
g();
Fixes a bug in the formatter where we would assume there is a previous
non-comment token.
Also adds the indent level of an unwrapped line to the debug output in
the parser.
llvm-svn: 173168
We used to align trailing comments belong to different things.
Before:
void f() { // some function..
}
int a; // some variable..
After:
void f() { // some function..
}
int a; // some variable..
llvm-svn: 173100
Very similar to what we do for record definitions:
- tighten down what is an enum definition, so that we don't mistake a
function for an enum
- allow common idioms around declarations (we'll want to handle that
more centrally in the future)
We now correctly format:
enum X f() {
a();
return 42;
}
llvm-svn: 173075
We now only put empty blocks into a single line, if all of:
- all tokens of the structural element fit into a single line
- we're not in a control flow statement
Note that we usually don't put record definitions into a single line, as
there's usually at least one more token (the semicolon) after the
closing brace. This doesn't hold when we are in a context where there is
no semicolon, like "enum E {}".
There were some missing tests around joining lines around the corner
cases of the allowed number of columns, so this patch adds some.
llvm-svn: 173055
Before: template <template <typename T>, typename P > class X;
After: template <template <typename T>, typename P> class X;
More importantly, the token annotations for the second ">" are now computed
correctly.
llvm-svn: 173047
Daniel Jasper