[Gambas-user] Bugtracker protocol

[Tobias Boege]

Hi Benoit,

for my command line bugtracker client, I have written some kind of ncurses
interface today which is at least able to fetch the session cookie after
logging in by now.

Then I was digging around further to see how to retrieve some bug reports
for display. My questions are:

 - Is there, by chance, an interface of the bugtracker from which I can
   retrieve raw (i.e. non-HTML) data?
 - If there isn't, I will have to use gb.xml.html to parse the webpages but:
     - the primary listing of bug reports I have found is in an HTML table
       without many "id" or "name" attributes (good would be a scheme like
       id = "bugXY-summary" for the <td> which holds the summary of bug #XY)
       which is unpleasant to parse (I have to hardcode column indices),
     - there are many french (accented) words in the HTML attributes.

My preferred solution would be to have read-only access to the table of
reports (I suppose there is a structured database of reports accessible via
standard gb.db methods). I am OK with using gb.net.curl's HttpForm to submit
stuff like reports or comments (authentication is necessary there anyway).

What do you think?

Regards,
Tobi

PS: Also I found [0] which is, if I connected the view pieces I understood
correctly, really impressive. Is this search engine actually employed by
the bugtracker? Would be good if the words could be in English :-)

[0] http://gambaswiki.org/bugtracker/help/search

-- 
"There's an old saying: Don't change anything... ever!" -- Mr. Monk