Re: [ubl-hisc] Latest Input Spec progress

G. Ken Holman wrote:

> This looks good, Micah ... I have a few comments on the Python, 
> perhaps others will as well.
>
>  (1) - would the building of the record classes be done more easily 
> from the XML using DOM or SAX instead of parsing the flat text?  I get 
> the feeling that may have made the job more direct than recreating a 
> structure from a string that was created from the structure.  Since it 
> is working it doesn't have to change, but I'm curious why you didn't 
> go directly to the structures.

The Python approach has served me well so far. Alot of it is just 
figuring out what needs to be done--flexibility is key. If I do a good 
job, the final solution will look like something that would be a good 
fit for other technologies!

>  (2) - can the set of rules be split into a separate Python source 
> file that is imported by this file ... or can that separate Python 
> source file import this file, such that the collection of rules is 
> practically the only thing in the file except for comments and 
> necessary wrapper declarations and invocations?  That way we can point 
> to the file and a lay person need not go through all of the Pythonese 
> to get to the table of rule statements.

Done.

> Thanks!

-m

-- 
  Available for consulting. XForms, web forms, information overload.
  Micah Dubinko                           mailto:
  Brain Attic, L.L.C.                        http://brainattic.info
  Yahoo IM: mdubinko                                +1 623 298 5172
  Learn XForms today: http://xformsinstitute.com