How do I Extract Attachment from Newsgroup Message

[snewman18]

I'm parsing NNTP messages that have XML file attachments. How can I
extract the encoded text back into a file? I looked for a solution
with mimetools (the way I'd approach it for email) but found nothing.

Here's a long snippet of the message:

>>n.article('11 6431')

('220 116431 <D8*******@news .ap.orgarticle' , '116431',
'<D8*******@new s.ap.org>', ['MIME-Version: 1.0', 'Message-ID:
<D8*******@news .ap.org>', 'Content-Type: Multipart/Mixed;', '
boundary="------------Boundary-00=_A5NJCP3FX6Y 5BI3BH890"', 'Date: Thu,
24 May 2007 07:41:34 -0400 (EDT)', 'From: Newsclip <ne******@ap.or g>',
'Path: newsclip.ap.org !flounder.ap.or g!flounder', 'Newsgroups:
ap.spanish.onli ne,ap.spanish.o nline.business' , 'Keywords: MUN ECO
PETROLEO PRECIOS', 'Subject: MUN ECO PETROLEO PRECIOS', 'Summary: ',
'Lines: 108', 'Xref: newsclip.ap.org ap.spanish.onli ne:938298
ap.spanish.onli ne.business:116 431', '', '', '--------------
Boundary-00=_A5NJCP3FX6Y 5BI3BH890', 'Content-Type: Text/Plain',
'Content-Transfer-Encoding: 8bit', 'Content-Description: text,
unencoded', '', '(AP) Precios del crudo se mueven sin rumbo claro',
'Por GEORGE JAHN', 'VIENA', 'Los precios

.... (truncated for length) ...

'', '___', '', 'Editores: Derrick Ho, periodista de la AP en Singapur,
contribuy\xf3 con esta informaci\xf3n. ', '', '', '--------------
Boundary-00=_A5NJCP3FX6Y 5BI3BH890', 'Content-Type: Text/Xml', 'Content-
Transfer-Encoding: base64', 'Content-Description: text, base64
encoded', '',
'PD94bWwgdmVyc2 lvbj0iMS4wIiBlb mNvZGluZz0iVVRG LTgiPz4KPCFET0N UWVBFIG5pdGYgU1 lT',
'VEVNICJuaXRmLm R0ZCI+CjxuaXRmP gogPGhlYWQ
+CiAgPG1ldGEgbm FtZT0iYXAtdHJhb nNyZWYi',
'IGNvbnRlbnQ9Il NQMTQ3MiIvPgogI DxtZXRhIG5hbWU9 ImFwLW9yaWdpbiI gY29udGVudD0ic3 Bh',
'bm9sIi8+CiAgPG 1ldGEgbmFtZT0iY XAtc2VsZWN0b3Ii IGNvbn

[kyosohma]

On May 31, 8:54 am, "snewma...@gmai l.com" <snewma...@gmai l.comwrote:

I'm parsing NNTP messages that have XML file attachments. How can I
extract the encoded text back into a file? I looked for a solution
with mimetools (the way I'd approach it for email) but found nothing.

Here's a long snippet of the message:

>n.article('116 431')

('220 116431 <D8PANK...@news .ap.orgarticle' , '116431',
'<D8PANK...@new s.ap.org>', ['MIME-Version: 1.0', 'Message-ID:
<D8PANK...@news .ap.org>', 'Content-Type: Multipart/Mixed;', '
boundary="------------Boundary-00=_A5NJCP3FX6Y 5BI3BH890"', 'Date: Thu,
24 May 2007 07:41:34 -0400 (EDT)', 'From: Newsclip <[e-mail] g>',
'Path: newsclip.ap.org !flounder.ap.or g!flounder', 'Newsgroups:
ap.spanish.onli ne,ap.spanish.o nline.business' , 'Keywords: MUN ECO
PETROLEO PRECIOS', 'Subject: MUN ECO PETROLEO PRECIOS', 'Summary: ',
'Lines: 108', 'Xref: newsclip.ap.org ap.spanish.onli ne:938298
ap.spanish.onli ne.business:116 431', '', '', '--------------
Boundary-00=_A5NJCP3FX6Y 5BI3BH890', 'Content-Type: Text/Plain',
'Content-Transfer-Encoding: 8bit', 'Content-Description: text,
unencoded', '', '(AP) Precios del crudo se mueven sin rumbo claro',
'Por GEORGE JAHN', 'VIENA', 'Los precios

... (truncated for length) ...

'', '___', '', 'Editores: Derrick Ho, periodista de la AP en Singapur,
contribuy\xf3 con esta informaci\xf3n. ', '', '', '--------------
Boundary-00=_A5NJCP3FX6Y 5BI3BH890', 'Content-Type: Text/Xml', 'Content-
Transfer-Encoding: base64', 'Content-Description: text, base64
encoded', '',
'PD94bWwgdmVyc2 lvbj0iMS4wIiBlb mNvZGluZz0iVVRG LTgiPz4KPCFET0N UWVBFIG5pdGYgU1 lT',
'VEVNICJuaXRmLm R0ZCI+CjxuaXRmP gogPGhlYWQ
+CiAgPG1ldGEgbm FtZT0iYXAtdHJhb nNyZWYi',
'IGNvbnRlbnQ9Il NQMTQ3MiIvPgogI DxtZXRhIG5hbWU9 ImFwLW9yaWdpbiI gY29udGVudD0ic3 Bh',
'bm9sIi8+CiAgPG 1ldGEgbmFtZT0iY XAtc2VsZWN0b3Ii IGNvbn

This looks like what you might be looking for:
http://mail.python.org/pipermail/pyt...ne/265018.html

Not sure if you'll need this or not, but here's some info on encoding/
decoding files:
http://www.jorendorff.com/articles/unicode/python.html

There are lots of ways to parse xml. I use the minidom module myself.

Mike

[Dave Borne]

I looked for a solution

with mimetools (the way I'd approach it for email) but found nothing.

....

<D8*******@news .ap.org>', 'Content-Type: Multipart/Mixed;', '
boundary="------------Boundary-00=_A5NJCP3FX6Y 5BI3BH890"', 'Date: Thu,

....

Playing with

data = n.article('1164 31')[3]

and email.message_f rom_string, there seems to be a problem with the
content type being split up. I was able to get a multipart message by
using

msg = email.message_f rom_string('\n' .join(data).rep lace(';\n', ';'))

(and adding an ending boundary to your sample data).
This is a bit hackish and could cause problems if there are
semicolons inside the message body (no warranties expressed or
implied, etc.)

Hope this helps,
-Dave