Thread: regular expression question
I have a question regarding regular expression with postgres.
I am unable to find any reference to match on word boundaries (\b in
perl). If this functionality exists in postgres I am interested in
knowing how to use it, an example would probably suffice.
If it does not, which is what I suspect, I am interested in knowing how
others have dealt with this problem. It seems silly/ineffecient to just
try all cases, i.e. field ~* (^term| term | term$), but I can't think of a
better way.
Thanks,
Parker.
----------------------------------------------------------------
"Tom would prefer you use Emacs, because it makes Justin really
mad. Justin would prefer you use vi, because he thinks Tom
suffers from some vicious malaise. You'll probably be lame and
end up using Pico."
-Shoutcast Server Documentation
Parker Thompson dijo: > I have a question regarding regular expression with postgres. > > I am unable to find any reference to match on word boundaries (\b in > perl). If this functionality exists in postgres I am interested in > knowing how to use it, an example would probably suffice. > > If it does not, which is what I suspect, I am interested in knowing how > others have dealt with this problem. It seems silly/ineffecient to just > try all cases, i.e. field ~* (^term| term | term$), but I can't think of a > better way. You can use [[:<:]] and [[:>:]] as word boundaries. -- Alvaro Herrera (<alvherre[a]atentus.com>) "La vida es para el que se aventura"
I wish everyone would use the same syntax...it's getting to the point where you need a reference book for everybody's idea of how regex should be done. -d Alvaro Herrera wrote: >Parker Thompson dijo: > > > >>I have a question regarding regular expression with postgres. >> >>I am unable to find any reference to match on word boundaries (\b in >>perl). If this functionality exists in postgres I am interested in >>knowing how to use it, an example would probably suffice. >> >>If it does not, which is what I suspect, I am interested in knowing how >>others have dealt with this problem. It seems silly/ineffecient to just >>try all cases, i.e. field ~* (^term| term | term$), but I can't think of a >>better way. >> >> > >You can use [[:<:]] and [[:>:]] as word boundaries. > > >
David Ford dijo: > I wish everyone would use the same syntax...it's getting to the point > where you need a reference book for everybody's idea of how regex should > be done. In Linux (well, at least in Mandrake) you can look at regex(7) for the POSIX 1003.2 standard regexes, plus some things Henry Spencer added. His implementation is the one used in Postgres internally, so it has some value. -- Alvaro Herrera (<alvherre[a]atentus.com>) "El dia que dejes de cambiar dejaras de vivir"
Yes, and honestly I rather prefer the POSIX syntax in most cases, but I do like some of the shorthands, i.e. < (or \<) is so much easier to type than "[[:<:]]", and the same for > v.s. "[[:>:]]". That's just silly. This by the way is an extension. -d Alvaro Herrera wrote: >David Ford dijo: > > > >>I wish everyone would use the same syntax...it's getting to the point >>where you need a reference book for everybody's idea of how regex should >>be done. >> >> > >In Linux (well, at least in Mandrake) you can look at regex(7) for the >POSIX 1003.2 standard regexes, plus some things Henry Spencer added. His >implementation is the one used in Postgres internally, so it has some >value. > > >
On Tue, Jun 18, 2002 at 12:48:10AM -0400, David Ford wrote: > Yes, and honestly I rather prefer the POSIX syntax in most cases, but I > do like some of the shorthands, i.e. < (or \<) is so much easier to type > than "[[:<:]]", and the same for > v.s. "[[:>:]]". That's just silly. > This by the way is an extension. I must admit that perl has some nice regex extensions and it would be nice if postgres could use them (via libpcre). Then you could use \b, \s, \w, etc. However, Having looked into the regex language proposed for Perl 6, I think that's going way over the top. Perl 5 is fine. -- Martijn van Oosterhout <kleptog@svana.org> http://svana.org/kleptog/ > There are 10 kinds of people in the world, those that can do binary > arithmetic and those that can't.
Martijn van Oosterhout <kleptog@svana.org> writes:
> I must admit that perl has some nice regex extensions and it would be nice
> if postgres could use them (via libpcre). Then you could use \b, \s, \w,
> etc.
If you want perl-style regexps, write yourself a one-liner plperl
function. End of problem...
regards, tom lane