[prev in list] [next in list] [prev in thread] [next in thread] 

List:       nedit-discuss
Subject:    Re: Revision Tracking - For Tony Balinski
From:       "RICHARD D TURPIN" <cmacleod3 () msn ! com>
Date:       2003-10-16 19:26:20
[Download RAW message or body]

Thanks.  Let me know when you have something for 5.4.  I tried it with 5.3 
and 5.4 and had lots of problems.


>From: Tony Balinski <ajbj@free.fr>
>Reply-To: discuss@nedit.org
>To: RICHARD D TURPIN <cmacleod3@msn.com>
>CC: discuss@nedit.org
>Subject: Re: Revision Tracking - For Tony Balinski
>Date: Thu, 16 Oct 2003 06:01:08 +0200
>
>Quoting RICHARD D TURPIN <cmacleod3@msn.com>:
>
> > That sounds just like what I'm looking for.  Where do I get these 
>macros?
> >
> >
> > >From: Tony Balinski <ajbj@free.fr>
> > >Quoting RICHARD D TURPIN <cmacleod3@msn.com>:
> > >
> > > > True,  Linking into CVS would probably be the best way to do it.  
>What
> > > > I'm looking for just can't be done with simple text files.  Does
> > > > anyone know of or have any good algorithms to do character
> > > > level diffs on text files?
> > >
> > >I did this for the various colored file merging macros I wrote with the
> > >old rangeset functions. Once two versions of a file were merged (with
> > >colors
> > >marking old, new and changed lines), a "rework" macro could be called. 
>This
> > >would run through the changed sections, dump the old part to one file, 
>the
> > >new part to another, break everything into one word/space/punctuation 
>per
> > >line (with newlines represented as a pair of newlines), then run diff 
>on
> > >these
> > >two temporary files. After that, they were merged using the same 
>process as
> > >before, and the lines recombined. This gave a word-level difference. 
>This
> > >highlights some changes but can be annoying since it makes lines longer
> > >where
> > >words are changed, and other lines are broken in odd places in the 
>merged
> > >result. Changes in spaces and tabs can also look very odd.
> > >
> > >Tony
>
>They were in the original macros I wrote for rangesets in 1999. I still use
>them, slightly updated. However, they are incompatible with the RS routines
>in NEdit 5.4 (which provide a better, albeit a more rigorous interface). I
>have started to adapt them for 5.4, but since I have real work, which I do
>with NEdit, I never find the time to move away from my own cobbled 5.3 and
>redo my macros - of which I have a lot!
>
>You can find an ancient copy here, as part of one of the first RS patches:
>     http://ajbj.free.fr/nedit/nedit5.0.2-tony/rangeset_v2.1.tar.gz
>Check the instructions in there.
>
>I could put up what I have for 5.4 now, but it's incomplete. Otherwise, has
>anybody done anything similar (and uses it)? The word-by-word 
>postprocessing
>of the merged file I talked about before isn't actually in that stuff 
>though.
>
>I'd like to redo the macros to work "backwards" - currently what you do is
>invoke the macro on the output of diff / diff -r executed directly in a
>NEdit window (using the (ctrl-)keypad-enter, usually). The macro prompts 
>for
>a particular pair of files taken from the lines which read
>"diff [-r] left right". Once chosen, the left file is read, and all lines
>marked as to be removed or changed marked, with the new text, taken from 
>the
>diff output, inserted and marked. Doing this processing starting with the
>right file could allow say processing the output of cvs diff -R to compare
>current checked-out file copies with cvs repository versions.
>
>Tony

_________________________________________________________________
Never get a busy signal because you are always connected  with high-speed 
Internet access. Click here to comparison-shop providers.  
https://broadband.msn.com

[prev in list] [next in list] [prev in thread] [next in thread] 

Configure | About | News | Add a list | Sponsored by KoreLogic