Bug#1057778: qa.debian.org: [udd] some names are not stripped from blanks and quotes
Andreas Tille
Severity: normal
Hi,
in some tables names are not stripped from spaces. Example:
udd=> select '"' || u.name || '"' as name_with_spaces, uploader from uploaders u where name like '% ' or name like ' %' ;
name_with_spaces | uploader
--------------------------+-------------------------------------------
" Mehdi Dogguy" | Mehdi Dogguy <me...@debian.org>
" David Paleino" | David Paleino <da...@debian.org>
" Stéphane Glondu" | Stéphane Glondu <glo...@debian.org>
" Stefano Zacchiroli" | Stefano Zacchiroli <za...@debian.org>
" Stefano Zacchiroli" | Stefano Zacchiroli <za...@debian.org>
" Stefano Zacchiroli" | Stefano Zacchiroli <za...@debian.org>
" Stefano Zacchiroli" | Stefano Zacchiroli <za...@debian.org>
" Stefano Zacchiroli" | Stefano Zacchiroli <za...@debian.org>
"Andreas Tille " | Andreas Tille <ti...@debian.org>
" LI Daobing" | LI Daobing <lida...@debian.org>
" David Paleino" | David Paleino <da...@debian.org>
" Stefano Zacchiroli" | Stefano Zacchiroli <za...@debian.org>
" Nikita V. Youshchenko" | Nikita V. Youshchenko <yo...@debian.org>
" Nikita V. Youshchenko" | Nikita V. Youshchenko <yo...@debian.org>
" Nikita V. Youshchenko" | Nikita V. Youshchenko <yo...@debian.org>
" Nikita V. Youshchenko" | Nikita V. Youshchenko <yo...@debian.org>
" Nikita V. Youshchenko" | Nikita V. Youshchenko <yo...@debian.org>
"Colin Tuckley " | Colin Tuckley <col...@debian.org>
"Colin Tuckley " | Colin Tuckley <col...@debian.org>
"Colin Tuckley " | Colin Tuckley <col...@debian.org>
(20 rows)
in other tables names are containing quotes that should be stripped as well:
select distinct done, done_name, done_email, owner, owner_name, owner_email from archived_bugs where done_name like '%"%' or owner_name like '%"%' order by done_name;
done | done_name | done_email | owner | owner_name | owner_email
---------------------------------------------------------------------------------------------------------------+-----------------------------------------------------------------------------+-------------------------------------------------+-----------------------------------------------------------------------------------------------------------------+-----------------------------------------------------------------------------+----------------------------------------------
<der...@debian.org> | | der...@debian.org | "van...@gmail.com" <van...@gmail.com> | "van...@gmail.com" | van...@gmail.com
<twe...@debian.org> | | twe...@debian.org | "Varun Hiremath" <varunh...@gmail.com> | "Varun Hiremath" | varunh...@gmail.com
alex...@belikoff.net (Alexander L. Belikoff) | | alex...@belikoff.net | "Alexander L. Belikoff" <alex...@belikoff.net> | "Alexander L. Belikoff" | alex...@belikoff.net
an...@debian.org (Andreas B. Mundt) | | an...@debian.org | "Andreas B. Mundt" <an...@debian.org> | "Andreas B. Mundt" | an...@debian.org
antoine.ro...@gmail.com (Antoine R. Dumont (@ardumont)) | | antoine.ro...@gmail.com | "Antoine R. Dumont" <antoine.ro...@gmail.com> | "Antoine R. Dumont" | antoine.ro...@gmail.com
antoine.ro...@gmail.com (Antoine R. Dumont) | | antoine.ro...@gmail.com | "Antoine R. Dumont" <antoine.ro...@gmail.com> | "Antoine R. Dumont" | antoine.ro...@gmail.com
art...@hell.pl (Artur R. Czechowski) | | art...@hell.pl | "Artur R. Czechowski" <art...@hell.pl> | "Artur R. Czechowski" | art...@hell.pl
...
IMHO the parser should strip spaces and quotes that can be found at beginning and end of strings.
Kind regards
Andreas.
Andreas Tille
The problem was discussed on Debian-QA list. The discussion starts here:
https://lists.debian.org/debian-qa/2023/12/msg00007.html
I opened this bug report to keep a record of the discussion in BTS.
There is one statement by Lucas Nussbaum[1] defending the current
situation:
> It has been like that for about 15 years. I'm not sure changing the API
> because you think is wrong is a good idea.
I do not consider strings injected into a database are comparable to
an API. The fact that this issue exists since 15 years does not mean
that it is the right way to do. Some bugs are just discovered after
that long time since the trigger for the issue might be quite rare.
I personally do not see any potential break of any application by
stripping names to be more easily parseable. I'd be happy if someone
could convince me otherwise by for instance describing applications that
are relying on different spellings of the very same name.
Kind regards
Andreas.
[1] https://lists.debian.org/debian-qa/2023/12/msg00016.html
--
http://fam-tille.de