Re: [PATCH STABLE] i18n: use actual filename, in which function is defined, for hg.pot

2017-08-12 Thread Yuya Nishihara
On Sun, 13 Aug 2017 00:13:58 +0900, FUJIWARA Katsunori wrote:
> At Sat, 05 Aug 2017 03:16:57 +0900,
> FUJIWARA Katsunori wrote:
> > 
> > At Fri, 4 Aug 2017 00:36:28 +0900,
> > Yuya Nishihara wrote:
> > > 
> > > On Wed, 02 Aug 2017 22:36:45 +0900, FUJIWARA Katsunori wrote:
> > > > # HG changeset patch
> > > > # User FUJIWARA Katsunori 
> > > > # Date 1501599731 -32400
> > > > #  Wed Aug 02 00:02:11 2017 +0900
> > > > # Branch stable
> > > > # Node ID b6d0bcd19b4dffb2360f909f459c27104ad16565
> > > > # Parent  76b171209151fe41dbf8dbfec473cc533f3b40ca
> > > > # Available At https://bitbucket.org/foozy/mercurial-wip
> > > > #  hg pull https://bitbucket.org/foozy/mercurial-wip -r 
> > > > b6d0bcd19b4d
> > > > # EXP-Topic i18n-fix-update-pot-issues
> > > > i18n: use actual filename, in which function is defined, for hg.pot
> > > 
> > > > +funcmod = inspect.getmodule(func)
> > > > +extra = ''
> > > > +if funcmod.__package__ == funcmod.__name__:
> > > > +extra = '/__init__'
> > > > +actualpath = '%s%s.py' % (funcmod.__name__.replace('.', 
> > > > '/'), extra)
> > > > +
> > > >  src = inspect.getsource(func)
> > > > -name = "%s.%s" % (path, func.__name__)
> > > > +name = "%s.%s" % (actualpath, func.__name__)
> > > >  lineno = inspect.getsourcelines(func)[1]
> > > 
> > > Perhaps inspect.getsourcefile() can be used.
> > > 
> > 
> > Ah, I overlooked it. I'll post follow up patch on default. Thanks!
> 
> I gave up using inspect.getsourcefile() in this case, because:
> 
>   - hggettext insert result of os.getcwd() into sys.path at first, in
> order to avoid loading modules from already installed Mercurial,
> 
>   - and this makes inspect.getsourcefile() return absolute path
> 
> Even with "." instead of os.getcwd(), we should trim additional "./"
> prefix of filenames returned by inspect.getsourcefile(), in order to
> keep current order of entries in hg.pot.
> 
> If we omit trimming this additional "./", order of entries in hg.pot
> is drastically changed, and it might cause meaningless changes of *.po
> files.
> 
> Simple replacement with inspect.getsourcefile() doesn't work as we
> expected at first :-<

Fair enough. Thanks for trying that.
___
Mercurial-devel mailing list
Mercurial-devel@mercurial-scm.org
https://www.mercurial-scm.org/mailman/listinfo/mercurial-devel


Re: [PATCH STABLE] i18n: use actual filename, in which function is defined, for hg.pot

2017-08-12 Thread FUJIWARA Katsunori
At Sat, 05 Aug 2017 03:16:57 +0900,
FUJIWARA Katsunori wrote:
> 
> At Fri, 4 Aug 2017 00:36:28 +0900,
> Yuya Nishihara wrote:
> > 
> > On Wed, 02 Aug 2017 22:36:45 +0900, FUJIWARA Katsunori wrote:
> > > # HG changeset patch
> > > # User FUJIWARA Katsunori 
> > > # Date 1501599731 -32400
> > > #  Wed Aug 02 00:02:11 2017 +0900
> > > # Branch stable
> > > # Node ID b6d0bcd19b4dffb2360f909f459c27104ad16565
> > > # Parent  76b171209151fe41dbf8dbfec473cc533f3b40ca
> > > # Available At https://bitbucket.org/foozy/mercurial-wip
> > > #  hg pull https://bitbucket.org/foozy/mercurial-wip -r 
> > > b6d0bcd19b4d
> > > # EXP-Topic i18n-fix-update-pot-issues
> > > i18n: use actual filename, in which function is defined, for hg.pot
> > 
> > > +funcmod = inspect.getmodule(func)
> > > +extra = ''
> > > +if funcmod.__package__ == funcmod.__name__:
> > > +extra = '/__init__'
> > > +actualpath = '%s%s.py' % (funcmod.__name__.replace('.', 
> > > '/'), extra)
> > > +
> > >  src = inspect.getsource(func)
> > > -name = "%s.%s" % (path, func.__name__)
> > > +name = "%s.%s" % (actualpath, func.__name__)
> > >  lineno = inspect.getsourcelines(func)[1]
> > 
> > Perhaps inspect.getsourcefile() can be used.
> > 
> 
> Ah, I overlooked it. I'll post follow up patch on default. Thanks!

I gave up using inspect.getsourcefile() in this case, because:

  - hggettext insert result of os.getcwd() into sys.path at first, in
order to avoid loading modules from already installed Mercurial,

  - and this makes inspect.getsourcefile() return absolute path

Even with "." instead of os.getcwd(), we should trim additional "./"
prefix of filenames returned by inspect.getsourcefile(), in order to
keep current order of entries in hg.pot.

If we omit trimming this additional "./", order of entries in hg.pot
is drastically changed, and it might cause meaningless changes of *.po
files.

Simple replacement with inspect.getsourcefile() doesn't work as we
expected at first :-<


> -- 
> --
> [FUJIWARA Katsunori] fo...@lares.dti.ne.jp
> 

-- 
--
[FUJIWARA Katsunori] fo...@lares.dti.ne.jp
___
Mercurial-devel mailing list
Mercurial-devel@mercurial-scm.org
https://www.mercurial-scm.org/mailman/listinfo/mercurial-devel


Re: [PATCH STABLE] i18n: use actual filename, in which function is defined, for hg.pot

2017-08-04 Thread FUJIWARA Katsunori
At Fri, 4 Aug 2017 00:36:28 +0900,
Yuya Nishihara wrote:
> 
> On Wed, 02 Aug 2017 22:36:45 +0900, FUJIWARA Katsunori wrote:
> > # HG changeset patch
> > # User FUJIWARA Katsunori 
> > # Date 1501599731 -32400
> > #  Wed Aug 02 00:02:11 2017 +0900
> > # Branch stable
> > # Node ID b6d0bcd19b4dffb2360f909f459c27104ad16565
> > # Parent  76b171209151fe41dbf8dbfec473cc533f3b40ca
> > # Available At https://bitbucket.org/foozy/mercurial-wip
> > #  hg pull https://bitbucket.org/foozy/mercurial-wip -r 
> > b6d0bcd19b4d
> > # EXP-Topic i18n-fix-update-pot-issues
> > i18n: use actual filename, in which function is defined, for hg.pot
> 
> > +funcmod = inspect.getmodule(func)
> > +extra = ''
> > +if funcmod.__package__ == funcmod.__name__:
> > +extra = '/__init__'
> > +actualpath = '%s%s.py' % (funcmod.__name__.replace('.', '/'), 
> > extra)
> > +
> >  src = inspect.getsource(func)
> > -name = "%s.%s" % (path, func.__name__)
> > +name = "%s.%s" % (actualpath, func.__name__)
> >  lineno = inspect.getsourcelines(func)[1]
> 
> Perhaps inspect.getsourcefile() can be used.
> 

Ah, I overlooked it. I'll post follow up patch on default. Thanks!

-- 
--
[FUJIWARA Katsunori] fo...@lares.dti.ne.jp
___
Mercurial-devel mailing list
Mercurial-devel@mercurial-scm.org
https://www.mercurial-scm.org/mailman/listinfo/mercurial-devel


Re: [PATCH STABLE] i18n: use actual filename, in which function is defined, for hg.pot

2017-08-03 Thread Yuya Nishihara
On Wed, 02 Aug 2017 22:36:45 +0900, FUJIWARA Katsunori wrote:
> # HG changeset patch
> # User FUJIWARA Katsunori 
> # Date 1501599731 -32400
> #  Wed Aug 02 00:02:11 2017 +0900
> # Branch stable
> # Node ID b6d0bcd19b4dffb2360f909f459c27104ad16565
> # Parent  76b171209151fe41dbf8dbfec473cc533f3b40ca
> # Available At https://bitbucket.org/foozy/mercurial-wip
> #  hg pull https://bitbucket.org/foozy/mercurial-wip -r 
> b6d0bcd19b4d
> # EXP-Topic i18n-fix-update-pot-issues
> i18n: use actual filename, in which function is defined, for hg.pot

> +funcmod = inspect.getmodule(func)
> +extra = ''
> +if funcmod.__package__ == funcmod.__name__:
> +extra = '/__init__'
> +actualpath = '%s%s.py' % (funcmod.__name__.replace('.', '/'), 
> extra)
> +
>  src = inspect.getsource(func)
> -name = "%s.%s" % (path, func.__name__)
> +name = "%s.%s" % (actualpath, func.__name__)
>  lineno = inspect.getsourcelines(func)[1]

Perhaps inspect.getsourcefile() can be used.
___
Mercurial-devel mailing list
Mercurial-devel@mercurial-scm.org
https://www.mercurial-scm.org/mailman/listinfo/mercurial-devel


Re: [PATCH STABLE] i18n: use actual filename, in which function is defined, for hg.pot

2017-08-02 Thread Augie Fackler
On Wed, Aug 02, 2017 at 10:36:45PM +0900, FUJIWARA Katsunori wrote:
> # HG changeset patch
> # User FUJIWARA Katsunori 
> # Date 1501599731 -32400
> #  Wed Aug 02 00:02:11 2017 +0900
> # Branch stable
> # Node ID b6d0bcd19b4dffb2360f909f459c27104ad16565
> # Parent  76b171209151fe41dbf8dbfec473cc533f3b40ca
> # Available At https://bitbucket.org/foozy/mercurial-wip
> #  hg pull https://bitbucket.org/foozy/mercurial-wip -r 
> b6d0bcd19b4d
> # EXP-Topic i18n-fix-update-pot-issues
> i18n: use actual filename, in which function is defined, for hg.pot

queued, thanks
___
Mercurial-devel mailing list
Mercurial-devel@mercurial-scm.org
https://www.mercurial-scm.org/mailman/listinfo/mercurial-devel