Re: PW#5-12: New upload procedure

[kaih@khms.westfalen.de (Kai Henningsen)]

schwarz@monet.m.isar.de (Christian Schwarz)  wrote on 13.01.98 in <[🔎] 199801132234.XAA04243@klee.painters.schwarz-online.com>:

>      So, if a package upload fixes some bugs, the maintainer should include
>      some tags in the debian/changelog file that use the following syntax
>      (Perl regexp syntax, case-insensitive):
>
>            /fixes:\s*(bug)?\#\d+\(,\s*(bug)?\#\d+\)*/i
>
>      Here is an example for those that do not speak Perl:
>
>        foo (1.0-5) unstable; urgency=low
>
>          * Recompiled with libc6, fixes: Bug#98765, Bug#98766
>          * Removed `rm -rf /' from postrm script, fixes:#99999
>
>         -- Wile E. Coyote <coyote@debian.org>  Wed, 29 Oct 1997 14:22:37
> -0600

The regexp doesn't match the example. It does match

         fixes: bug#123(, bug#345)))))))))))))))))

which is probably not what you intended :-)

In Perl regular expressions, alphanumeric characters always match  
literally unless preceded by a backslash, and non-alphanumeric characters  
always match literally if preceded by a backslash.

So, take out the backslashes before the parentheses, and you are fine.

OTOH, every () sets one of the $1 $2 ... variables; if you want to avoid  
that (that's more for the actual implementation than for the policy  
manual), use (?: ) instead of ( ) to not touch those variables.

Here's a short explanation of the corrected regex:

       /fixes:\s*(bug)?\#\d+(,\s*(bug)?\#\d+)*/i

        fixes:      matches "fixes:"
        \s*         matches zero or more whitespace characters
                    (in C: one of " \t\n\r\f")
        bug         matches "bug"
        (thing)?    matches "" or "thing"
        \#          matches "#"
        \d+         matches some digits, at least one
        ,           matches ","
        (thing)*    matches "", "thing", "thingthing", and so on
        /thing/i    matches "thing", "Thing", "tHiNg", and so on


MfG Kai