I'm sure I'm speaking out of turn here, but what good is it to rely on the
composition of generated source code? It would be a true shame if Chameleon
internals became APIs.
On 3/1/10 10:47 AM, Uli Fouquet wrote:
> Hi there,
>
> Malthe Borch wrote:
>> On 27 February 2010 12:10, Uli Fouquet<u...@gnufix.de> wrote:
>>> While trying to make megrok.chameleon compatible with zopetoolkit and
>>> groktoolkit package versions, I noticed that latest z3c.pt (1.1.0) and
>>> trunk are not completely compatible with Chameleon>= 1.1.0. The tests
>>> fail.
>>
>> Do you think you can find time to try and reproduce this in straight
>> Chameleon?
>
> Sure, the following doctest example added to core/codegen.txt passes
> with Chameleon 1.0.8::
>
> Syntax extension: ``econtext`` references
> -----------------------------------------
>
> >>> suite = Suite("""\
> ... result = _path(options, True, 'test')
> ... """)
> >>> suite.source
> "result = econtext['_path'](econtext['options'], True, 'test')\n"
>
> But with Chameleon trunk gives::
>
> ----------------------------------------------------------------------
> File "...chameleon/trunk/src/chameleon/core/codegen.txt", line 48, in
> codegen.txt
> Failed example:
> suite.source
> Expected:
> "result = econtext['_path'](econtext['options'], True, 'test')\n"
> Got:
> "result =
> econtext['econtext']['_path'](econtext['econtext']['options'], True,
> 'test')\n"
>
> [snip]
>
>>> 'options/test'
>>> result = econtext['econtext']['_path'](
>>> econtext['econtext']['options'],
>>> econtext['econtext']['request'], True, 'test')
>>
>> Seems like the code that does the name transformation
>> (codegen.py:visit_Name) should include ``econtext`` to the list of
>> globally known symbols.
>
> Ah, I see. Thanks for the hint! Looks like this is really the crucial
> point.
>
> When I add ``'econtext'`` to ``CONSTANTS`` (just to inject it somewhere,
> probably not the right place), all tests pass, including the snippet
> above. This is true for both versions, the current trunk and 1.0.8 :-)
>
> [snip]
>
> Thanks for looking into this! If you think the mentioned behaviour is
> really a bug in Chameleon, I'd be glad to assist in fixing it.
>
> Best regards,
>
>
>
>
> _______________________________________________
> Repoze-dev mailing list
> Repoze-dev@lists.repoze.org
> http://lists.repoze.org/listinfo/repoze-dev
--
Chris McDonough
Agendaless Consulting, Fredericksburg VA
The repoze.bfg Web Application Framework Book: http://bfg.repoze.org/book
_______________________________________________
Repoze-dev mailing list
Repoze-dev@lists.repoze.org
http://lists.repoze.org/listinfo/repoze-dev
