Quoting Robert Haas <robertmhaas(at)gmail(dot)com>:

>
> I'm not sure I understand how this more convenient than just using
> xpath() with exists()?
>

It will save a lot of complexity in WHERE clauses. For example using  
exists() in xpath() you might construct something like:

WHERE array_dims(xpath('exists(/foo/bar)','<bar><foo/></bar>'::xml) IS  
NOT NULL ...

Whereas a dedicated xpath_exists() would look like:

WHERE xpath_exists('/foo/bar','<bar><foo/></bar>'::xml) ....

I accept this example is quite basic, but I hope it illustrates the  
added usability. I think xml in sql is complex enough, especially when  
you start considering namespaces, that anything we can do to simplify  
common use cases can only help improve the uptake of postgres xml.

Thanks,

--
Mike Fowler

In response to

• Re: Adding xpath_exists function at 2010-05-06 15:43:09 from Robert Haas

Responses

• Re: Adding xpath_exists function at 2010-05-07 02:40:59 from Robert Haas

pgsql-hackers by date

Next:	From: Takahiro Itagaki	Date: 2010-05-07 02:19:33
	Subject: Re: pg_stat_transaction patch
Previous:	From: Bruce Momjian	Date: 2010-05-06 21:32:07
	Subject: Re: max_standby_delay considered harmful

pgsql-rrreviewers by date

Next:	From: Robert Haas	Date: 2010-05-07 02:40:59
	Subject: Re: Adding xpath_exists function
Previous:	From: Robert Haas	Date: 2010-05-06 15:43:09
	Subject: Re: Adding xpath_exists function