[Info-vax] DCL Differences algorithm

[John Wallace]

On Nov 9, 6:25 pm, "John Reagan" <johnrrea... at earthlink.net> wrote:
> "John Wallace" <johnwalla... at yahoo.co.uk> wrote in message
>
> news:049953d1-118c-494e-be3d-437c6bc85f9d at l32g2000yqc.googlegroups.com...
> Based on that experience, I'm not 100% convinced that VMS DIFF uses
> the industry standard algorithm, but (a) I could be wrong (b) it may
> not matter for your needs.
>
> When I looked years ago, the code simply read records with RMS and used
> CH$EQL (or STR$COMPARE_CASE_BLIND or whatever the name is if you are
> ignoring case).  No funky 'hashing' or whatever.  As Fred said, all in
> BLISS, not too complicated.  Should be on the listing CDs for those of you
> with access.
>
> Personally, for files with interesting differences, I used CMS DIFF instead
> of DCL DIFF.  I don't know what algorithm it uses, but it seems to not get
> as upset when I added a routine in the middle of a source module...  I'll
> guess that CMS DIFF just accesses the same difference engine used for
> REPLACE and edits to the CMS element files.
>
> John

Comparing of the records is relatively trivial. Getting back into sync
after a difference is the fun part, which is where the "dynamic
programming" (and hashing) stuff comes in in UNIX diff, or "the
sliding window" in VMS DIFF. Other algorithms are available.