I wrote my own Unix shell script + comm program script to upload and
download files for yarn to import and export. It was more complex than
Olmenu, because my accoupt is used by 3 separate people. Additionally,
I wanted to be able to select which newsgroups to upload and to
dynamically filter articles each time I uploaded a soup packet from
my ISP to my PC.
To do the latter I used uqwk's summary mode option and a user defined
.newsrc file: I could select which news groups were contained in .newsrc,
uqwk would then produce a summary file containing the subject lines of
selected articles -- then I could use unix software (e.g., grep) or
even any editor to further filter the subject lines. Finally, uqwk grabs
the real articles corresponding to the subject lines and I get a soup
packet.
Well all this is a bit of bother, AND uqwk took quite a while (minutes)
to produce a summary file for even one fairly large newsgroup (for instance,
ba.jobs.offered). I believe this is because uqwk opens each and every file
corresponding to an article, and then graps the header.
The news reader, trn, on the other hand simply reads a pre-existing
separate overfile containing article headers (much like yarn itself).
As a result, trn can search for a pattern in the headers of
a lot of articles and then download a resulting mailbox file in a flash!
Finally, I noticed yarn's:
import -r
option which will import an ordinary ascii file in mailbox format
into its mail inbox. Mail filter rules can then forward the articles
to the apprpriate news group. I would like to recommend this method of
getting selected news articles into yarn:
1) Use trn to select articles based on a pattern in the header and
or body
2) save the articles in a mailbox file
3) upload the mailbox file to your pc
4) run import -r mailbox-file
5) Use yarn's mail filter to send the articles to the proper newsgroup
Whew! Sorry I said that so long winded-ly.