[chora] chora image diff

Ryan Gallagher ryan@studiesabroad.com
Wed, 14 Aug 2002 11:45:45 -0500


Quoting Chuck Hagenbuch <chuck@horde.org>:

> Quoting Ryan Gallagher <ryan@studiesabroad.com>:
> 
> > This thread really peaked my interest.  I apologize ahead if comments
> > were not invited.
> 
> It's a public mailing list. On-topic conversation is implicitly invited. :)
> 
> > What about various structured document formats?  About to go out on a
> > limb here... Stop me if any repositories or diff techniques already deal 
> > with this.
> 
> That's a _very_ interesting idea. I don't know of anything that does it 
> now, but there very well might be some XML technologies for doing it (I'd 
> hesitate to apply it to non-XHTML HTML, personally). Would be neat to code 
> a diff driver for it...
> 
> > A) I'm not sure how possible it would be with simply a CVS backend.
> 
> You should be able to do it, using external code of course, as long as you 
> can retrieve two revisions.
> 
> > B) I do not have the time at the moment to investigate too much further
> > or code it.  Just an idea for a possible feature.
> 
> Well, shucks. :) But maybe you'll have time in the future, or this will 
> spark someone else's interest.
> 
> > Allowing other diff types on other filetypes seems like one of those
> > areas where "pluggable" apps (be they php or other) might come in to play.
> 
> Yeah, sounds like we might want to have a Chora_Diff class with pluggable 
> drivers...
> 

Cool, glad you liked it.  I think it may be more applicable than even I
originally imagined.  For example, there was mention of a desire to diff MS
Office formats.  Well that's quite a doozy but on the other hand, OpenOffice
uses a standards compliant XML format for all its documents.  Not sure how
useful looking at the XML _element_ diff (or whatever this might be called)
would be for the end user, but it's a start.  

This also relates to the mention of chora peeking inside archive formats. 
OpenOffice (and many others i'm sure) zip their XML so that it can be managed as
one file by the user.  

Coming back to reality, a simple structured doc diff on a single XML or SGML
file would probably be an excellent start.

I do have a long weekend coming up, (and my fav. coffeehouse just got free
wireless) so i'll investigate some more and maybe i'll get excited enough to
take back what I said about not having time to code ;-)

-- 
Ryan T. Gallagher
ryan@studiesabroad.com
International Studies Abroad
http://www.studiesabroad.com
(512)480-8522