pkg-info: supplying word boundary in regex

[Matthew Seaman]

On 2015/01/12 15:40, parv at pair.com wrote:
> (Running pkg 1.4.4 on FreeBSD 8-STABLE here.)
> 
> Could somebody tell me please the syntax for word boundary in regex
> for pkg-info ...
> 
>   pkg info -x '...'
> 
> 
> ... ?
> 
> The pkg-info manual page says ...
> 
>   -x, --regex
>     Treat pkg-name as a regular expression according to the "modern"
>     or "extended" syntax of re_format(7).
> 
> ... and re_format man page says ...
> 
>   There are two special cases= of bracket expressions: the bracket
>   expressions `[[:<:]]' and `[[:>:]]' match the null string at the
>   beginning and end of a word respectively.  A word is defined as a
>   sequence of word characters which is neither preceded nor followed
>   by word characters.  A word character is an alnum character (as
>   defined by ctype(3)) or an underscore.
> 
> ... then specifying a word boundary as "x[[:>:]]" (to match "x" at the
> end of a "word") causes "Invalid regex" error. For example, to get
> result only for "tex" (avoiding packages with "text" as string in a
> package name[0]) ...
> 
>    # pkg info -x 'tex[[:>:]]'
>    pkg: sqlite error while executing iterator in file \
>    pkgdb_iterator.c:905: Invalid regex

Normally I'd suggest trying:

     # env DEBUG=4 pkg info -x 'tex[[:>:]]'

which should display exactly what SQL is being run.  However it appears
that the error occurs before Sqlite does any querying -- it must be
while compiling the RE.

> 
> Does pkg overstate its support for re_format(7) then?
> 
> 
>   - parv
> 
> 
>   [0] Yes, I realize if a name happens to be "text-tex" I would also
>       get a result & that would be expected. Common case is that
>       packages here have only either "text" or "tex" exclusively in
>       the name.
> 

There is always "pkg info -x 'tex$' to anchor the RE at the end of a
string.  However that doesn't really answer your question.

The re_format(7) support comes straight out of libc -- it's compiled as
a loadable module for sqlite, so that queries of the form:

    SELECT foo FROM bar WHERE foo REXEP 'some-re' ;

should be using re_format(7) style REGEXPs.  Admittedly, we haven't
tested all the odd corner cases for regular expression syntax.  Cn you
raise an issue on github please?

	Cheers,

	Matthew




-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 882 bytes
Desc: OpenPGP digital signature
URL: <http://lists.freebsd.org/pipermail/freebsd-questions/attachments/20150112/0ec01541/attachment.sig>