ninja windows: sanitize rule names to not emit corrupt .ninja files

ninja windows: sanitize rule names to not emit corrupt .ninja files

Ninja doesn't want ( and ) in rule names.

R=thakis@chromium.org
BUG=chromium:125606
Committed: https://code.google.com/p/gyp/source/detail?r=1346

[Nico, 2012-04-30 21:30:33]

http://chromiumcodereview.appspot.com/10269012/diff/1/test/ninja/sanitize-rul...
File test/ninja/sanitize-rule-names/gyptest-sanitize-rule-names.py (right):

http://chromiumcodereview.appspot.com/10269012/diff/1/test/ninja/sanitize-rul...
test/ninja/sanitize-rule-names/gyptest-sanitize-rule-names.py:16:
test.build('sanitize-rule-names.gyp', test.ALL)
Nothing in this test looks ninja-specific. Should this just be a general test
for all generators?

[scottmg, 2012-04-30 21:33:04]

https://chromiumcodereview.appspot.com/10269012/diff/1/test/ninja/sanitize-ru...
File test/ninja/sanitize-rule-names/gyptest-sanitize-rule-names.py (right):

https://chromiumcodereview.appspot.com/10269012/diff/1/test/ninja/sanitize-ru...
test/ninja/sanitize-rule-names/gyptest-sanitize-rule-names.py:16:
test.build('sanitize-rule-names.gyp', test.ALL)
On 2012/04/30 21:30:33, Nico wrote:
> Nothing in this test looks ninja-specific. Should this just be a general test
> for all generators?

Yeah, good point. I don't think any other generators care about the details of
the rule names... So leave it in test/ninja/? Or better at top level?

[scottmg, 2012-04-30 21:36:24]

On 2012/04/30 21:33:04, scottmg wrote:
>
https://chromiumcodereview.appspot.com/10269012/diff/1/test/ninja/sanitize-ru...
> File test/ninja/sanitize-rule-names/gyptest-sanitize-rule-names.py (right):
> 
>
https://chromiumcodereview.appspot.com/10269012/diff/1/test/ninja/sanitize-ru...
> test/ninja/sanitize-rule-names/gyptest-sanitize-rule-names.py:16:
> test.build('sanitize-rule-names.gyp', test.ALL)
> On 2012/04/30 21:30:33, Nico wrote:
> > Nothing in this test looks ninja-specific. Should this just be a general
test
> > for all generators?
> 
> Yeah, good point. I don't think any other generators care about the details of
> the rule names... So leave it in test/ninja/? Or better at top level?

Moved to top level and runs on all generators.

[Nico, 2012-04-30 21:37:37]

patch set 2 lgtm (not sure what the generators will make of a .X file, but we'll
see)

[scottmg, 2012-04-30 21:39:22]

On 2012/04/30 21:37:37, Nico wrote:
> patch set 2 lgtm (not sure what the generators will make of a .X file, but
we'll
> see)

Isn't that 'rules' is for? Or do some generators do magic things based on? Can
change to .s which was the original broken thing if it fails somewhere.

[Nico, 2012-04-30 21:44:55]

On 2012/04/30 21:39:22, scottmg wrote:
> On 2012/04/30 21:37:37, Nico wrote:
> > patch set 2 lgtm (not sure what the generators will make of a .X file, but
> we'll
> > see)
> 
> Isn't that 'rules' is for? Or do some generators do magic things based on? Can
> change to .s which was the original broken thing if it fails somewhere.

I guess that makes sense :-)

[Nico, 2012-04-30 21:45:09]

(keeping stuff as is, that is)

[scottmg, 2012-04-30 21:47:36]

On 2012/04/30 21:44:55, Nico wrote:
> On 2012/04/30 21:39:22, scottmg wrote:
> > On 2012/04/30 21:37:37, Nico wrote:
> > > patch set 2 lgtm (not sure what the generators will make of a .X file, but
> > we'll
> > > see)
> > 
> > Isn't that 'rules' is for? Or do some generators do magic things based on?
Can
> > change to .s which was the original broken thing if it fails somewhere.
> 
> I guess that makes sense :-)

Too bad neither of the first two sentences made any sense, and the third one is
only barely readable. ;-)

Anyway... I'll, uh, fix it if it breaks. "Sense: That made none."

[scottmg, 2012-05-01 16:07:23]

it would seem the scons generator has the same bug:
http://build.chromium.org/p/tryserver.nacl/builders/gyp-linux/builds/221

(does anyone actually use the scons generator?)

[Nico, 2012-05-01 16:10:06]

On 2012/05/01 16:07:23, scottmg wrote:
> it would seem the scons generator has the same bug:
> http://build.chromium.org/p/tryserver.nacl/builders/gyp-linux/builds/221
> 
> (does anyone actually use the scons generator?)

Not as far as I know. Probably enough to do formats=['!scons']

[scottmg, 2012-05-01 16:10:48]

On 2012/05/01 16:07:23, scottmg wrote:
> it would seem the scons generator has the same bug:
> http://build.chromium.org/p/tryserver.nacl/builders/gyp-linux/builds/221
> 
> (does anyone actually use the scons generator?)

and when that's fixed it does indeed whine about .X being an unsupported
extension. pffft.

[scottmg, 2012-05-01 16:16:41]

On 2012/05/01 16:10:06, Nico wrote:
> On 2012/05/01 16:07:23, scottmg wrote:
> > it would seem the scons generator has the same bug:
> > http://build.chromium.org/p/tryserver.nacl/builders/gyp-linux/builds/221
> > 
> > (does anyone actually use the scons generator?)
> 
> Not as far as I know. Probably enough to do formats=['!scons']

Oh, that would have been easier. Oh well, next time it'll get the exclusion
treatment. http://chromiumcodereview.appspot.com/10263021/