RE: Subject line processing in Gnats 4.0

[Dirk Bergstrom]

paul traina writes:
> Negative Dirk,
> I had it behave that way before, and it sucked.  Every time 
> someone sent in anything like "OS/2" it would append it to PR #2.

mr. traina makes a good point (he does that rather often).  i had been 
wrestling with this, with no great success.  it's why the idea had been sitting 
in my files for months.  now that my memory has been jogged, i remember a 
couple of solutions i'd had in mind.  in order of increasing radicalism:

*) accept FW: as well as RE: -- change the subject pattern thusly:

- "(.*re[ \t]*(\\[[0-9]+\\])?:)?[ \t]*([-a-z0-9_+.]*[:/][ \t]*([0-9]+))"
+ "(.*(re|fwd?)[ \t]*(\\[[0-9]+\\])?:)?[ \t]*([-a-z0-9_+.]*[:/][ \t]*([0-9]+))"

this will require adjusting what parenthetical expression is saved out of the 
match.

*) gnats currently accepts things like "sw-foo: 1234" instead of "sw-foo/1234". 
 i'm guessing that the former is an antiquated form, since gnats doesn't emit 
it anymore.  people don't use it, but they *do* use "PR1234".  adjust the regex 
appropriately:

+ "(.*(re|fwd?)[ \t]*(\\[[0-9]+\\])?:)?[ \t]*(([-a-z0-9_+.]*/)|pr)([0-9]+))"

this will require some futher munging of code, since the stuff that pulls out 
and uses <category>/<number> will have to learn to contend with PR<number>.  
since it really isn't necessary to use category to pull up the PR, it might be 
best to change the regex to save only the number, and not the category/number 
pair.  then we could just hack out the additional parsing code.

i think that the first change is the minimum we should do.

--
Dirk Bergstrom               address@hidden
_____________________________________________
Juniper Networks Inc.,          Computer Geek
Tel: 707.433.0564           Fax: 707.433.0769