nntplib: abstraction of threads

[Rakesh]

For a particular application of mine, I need to get the messages from
usenet , (and group them by each thread) . My startup python code looks
as follows.

<--- Startup code to read messages from a newsgroup -->

import nntplib, cStringIO, rfc822, sys

SRVR = '<my_news_serve r>' # Your news server
newsgroup = 'comp.lang.c' # Group of your choice

def inpdflt(s, d):
resp = raw_input("%s [%s]: " % (s, d))
return resp or d

news = nntplib.NNTP(SR VR)
resp, estimate, first, last, name = news.group(news group)

if estimate == '0':
sys.exit("No messages in " + newsgroup)

#
# Get (article number, subject, poster, date, id, references, size,
lines)
# for each of the articles between first and last
#
xover = news.xover(firs t, last)

# loop through articles, extracting headers
for x in xover[1]:
# x == (article number, subject, poster, date, id, references,
size, lines)
try:
hdrs = news.head(x[0])[3]
mesg = rfc822.Message( cStringIO.Strin gIO("\r\n".join (hdrs)))
print '%s\n+++%s' % (mesg.getheader ("from"),
mesg.getheader( "subject"))
except nntplib.NNTPErr or:
pass
news.quit()

<-- End newsgroup -->

I am getting all the messages of the newsgroup stored in the newsgroup
server.
What I want is to *group the messages belonging to each thread* .

How would I do that ?

Eg:

Topic 1
|
-- Re: Topic:1
-- Re: Topic: 1
|
-- Re: Re: Topic 1

Topic 2
|
-- Re: Topic:2


Total number of messages 6, but number of threads = 2,
I want to get an abstraction something similar to this.

[Werner Amann]

Re: nntplib: abstraction of threads

Rakesh schrieb:
[color=blue]
> What I want is to *group the messages belonging to each thread* .[/color]

Hello

Why not sort with Message-ID and References?
Attention - it is a Newbie-Solution.

import nntplib

hamster = nntplib.NNTP('1 27.0.0.1', 119, 'user', 'pass')
resp, count, first, last, name = hamster.group(' comp.lang.pytho n')
resp, items = hamster.xover(f irst,last)

start_dic = {}
re_dic = {}
numb = 1

for id,subject,auth or,date,message _id,references, size,lines in items:
if 'Re:' not in subject:
start_dic[subject] = (author, message_id)
else:
re_dic[numb] = (subject, author, references)
numb += 1

resp = hamster.quit()

for a in start_dic:
print a
print start_dic[a][0]
for b in re_dic:
if start_dic[a][1] in re_dic[b][2]:
print '|'
print ' ->', re_dic[b][0]
print ' ', re_dic[b][1]
print

--
Werner Amann

[Steve Holden]

Re: nntplib: abstraction of threads

Werner Amann wrote:
[color=blue]
> Rakesh schrieb:
>
>[color=green]
>>What I want is to *group the messages belonging to each thread* .[/color]
>
>
> Hello
>
> Why not sort with Message-ID and References?
> Attention - it is a Newbie-Solution.
>
> import nntplib
>
> hamster = nntplib.NNTP('1 27.0.0.1', 119, 'user', 'pass')
> resp, count, first, last, name = hamster.group(' comp.lang.pytho n')
> resp, items = hamster.xover(f irst,last)
>
> start_dic = {}
> re_dic = {}
> numb = 1
>
> for id,subject,auth or,date,message _id,references, size,lines in items:
> if 'Re:' not in subject:
> start_dic[subject] = (author, message_id)
> else:
> re_dic[numb] = (subject, author, references)
> numb += 1
>
> resp = hamster.quit()
>
> for a in start_dic:
> print a
> print start_dic[a][0]
> for b in re_dic:
> if start_dic[a][1] in re_dic[b][2]:
> print '|'
> print ' ->', re_dic[b][0]
> print ' ', re_dic[b][1]
> print
>[/color]
Better still, do a Google search on "mail threading algorithm",
implement the algorithm described in

http://www.jwz.org/doc/threading.html

and post your implementation back to the newsgroup :-)

regards
Steve
--
Steve Holden http://www.holdenweb.com/
Python Web Programming http://pydish.holdenweb.com/
Holden Web LLC +1 703 861 4237 +1 800 494 3119

[Rakesh]

Re: nntplib: abstraction of threads


Steve Holden wrote:[color=blue]
> Werner Amann wrote:
>[color=green]
> > Rakesh schrieb:
> >
> >[color=darkred]
> >>What I want is to *group the messages belonging to each thread* .[/color]
> >
> >
> > Hello
> >
> > Why not sort with Message-ID and References?
> > Attention - it is a Newbie-Solution.
> >
> > import nntplib
> >
> > hamster = nntplib.NNTP('1 27.0.0.1', 119, 'user', 'pass')
> > resp, count, first, last, name = hamster.group(' comp.lang.pytho n')
> > resp, items = hamster.xover(f irst,last)
> >
> > start_dic = {}
> > re_dic = {}
> > numb = 1
> >
> > for id,subject,auth or,date,message _id,references, size,lines in[/color][/color]
items:[color=blue][color=green]
> > if 'Re:' not in subject:
> > start_dic[subject] = (author, message_id)
> > else:
> > re_dic[numb] = (subject, author, references)
> > numb += 1
> >
> > resp = hamster.quit()
> >
> > for a in start_dic:
> > print a
> > print start_dic[a][0]
> > for b in re_dic:
> > if start_dic[a][1] in re_dic[b][2]:
> > print '|'
> > print ' ->', re_dic[b][0]
> > print ' ', re_dic[b][1]
> > print
> >[/color]
> Better still, do a Google search on "mail threading algorithm",
> implement the algorithm described in
>
> http://www.jwz.org/doc/threading.html[/color]

Thanks a lot for the link.
[color=blue]
>
> and post your implementation back to the newsgroup :-)[/color]

Sure I would. I would definitely do the same.
I am a python newbie and am reading nntp spec (rfc) right now.
Once I get a working version I would definitely post the same.

[JanC]

Re: nntplib: abstraction of threads

Steve Holden schreef:
[color=blue]
> Better still, do a Google search on "mail threading algorithm",
> implement the algorithm described in
>
> http://www.jwz.org/doc/threading.html
>
> and post your implementation back to the newsgroup :-)[/color]

<http://www.google.com/search?q=jwz+py thon&btnI=I'm+F eeling+Lucky>
<http://www.amk.ca/python/code/jwz>

--
JanC

"Be strict when sending and tolerant when receiving."
RFC 1958 - Architectural Principles of the Internet - section 3.9