Not with wdiff as it stands, so I think that would definitely require
changes to wdiff. I'm pretty happy with line level diffs for source
code because you don't often reflow paragraphs of text, but then again
I've never seen a token level diff. It's an interesting idea though...
Chris
Mike Samuel wrote:
> Is it possible to specify a tokenization scheme, perhaps as a regex,
> so that you can do a token level diff instead of a word level diff?
>
> 2009/11/29 Chris Pickett <chris.pickett_at_mail.mcgill.ca>:
>> Hi,
>>
>> A while ago I wrote some scripts to make Subversion better for working
>> with text documents. They wrap wdiff to show deleted words in red and
>> inserted words in green, and to trim the default wdiff output down to
>> one line each of leading and trailing context.
>>
>> They aren't perfect by any means, but they work 99% of the time for me.
>> I thought I'd publish them because I find them really useful.
>>
>> http://www.sable.mcgill.ca/~cpicke/swd.tar.gz
>>
>> Cross-posting to wdiff-bugs and SVN dev. It's primarily an enhancement
>> to wdiff, but I thought svn might have more resources. An `svn wdiff'
>> feature that mimics this would certainly be neat! Please at least CC me
>> in replies since I'm subscribed to neither.
>>
>> Chris
>>
>> ------------------------------------------------------
>> http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=2425185
>>
>> Please start new threads on the <dev_at_subversion.apache.org> mailing list.
>> To subscribe to the new list, send an empty e-mail to <dev-subscribe_at_subversion.apache.org>.
>>
------------------------------------------------------
http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=2425286
Please start new threads on the <dev_at_subversion.apache.org> mailing list.
To subscribe to the new list, send an empty e-mail to <dev-subscribe_at_subversion.apache.org>.
Received on 2009-11-29 21:32:43 CET