Re: [jamboost] Python port: current status

From: Pedro Ferreira (pedro.ferreira_at_[hidden])
Date: 2005-01-31 09:13:21

Em 31 Jan 2005, às 12:47, Alan Gutierrez escreveu:

> * Pedro Ferreira <pedro.ferreira_at_[hidden]> [2005-01-31 07:21]:
>> I attach the current snapshot of what I've been doing. There is still
>> plenty to do but I think this is a start.
>> (Volodya: do you think it is time to put this on CVS?)
>>
>> I think that (almost) all that is needed for make targets is done.
>> Please check tests/acceptance_tests/make_target.py.
>>
>> Help is much needed so, anyone willing to collaborate, please let me
>> know. Major areas are:
>>
>> - Generators (I'll start working on this one now)
>> - Tools (plenty to do for half a dozen people :-))
>> - Projects and their relationships
>> - Jamfile parser
>> - Scanners
>>
>> Feedback is very much appreciated.
>
> Doesn't Python come with a compiler, compiler in it's vast
> library? If not, what were you going to use to create the
> Jamfile parser? Maybe the existing parser generates an AST?

Volodya had proposed using bjam itself to parse Jamfiles. If I 
understood correcly, a series of stubs would forward instructions in 
Jamfiles to the appropriate calls in PyBB.
I'm not familiar with the internal of bjam so I can't really say if an 
AST exists and is easily usable.
As for the compiler compiler, yes, I think that eventually a pure 
Python parser should be created for Jamfiles or any other language.

Pedro