[ic] search query

[Philip S. Hempel]

On Thu, 2003-05-22 at 17:40, Philip S. Hempel wrote:
> On Thu, 2003-05-22 at 15:38, Stefan Hornburg wrote:
> > On 22 May 2003 14:56:56 -0400
> > "Philip S. Hempel" <pshempel@linuxhardcore.com> wrote:
> > 
> > > I am trying to get a search to function like the sql query below does
> > > 
> > > SELECT  `page`  FROM  `tree`  WHERE 1  AND  `mgroup`  =  'Products' AND 
> > > `msort`  =  '0' AND  `page`  !=  ' '
> > > 

<snip>

> > > [loop
> > >     search="
> > >     fi=tree
> > >     st=db
> > >     tf=sort
> > >     ac=0
> > >     ac=0
> > >     co=yes
> > > 
> > >     sf=page
> > >     op=ne
> > >     se=   <<- this is where I am lost how do you describe null or empty?
> > > 
> > >     sf=msort
> > >     se=0
> > >     op=rm
> > > 
> > >     sf=mgroup
> > >     se=Products
> > > 
> > > "]
> > > 
> > > ====
> > > 
> > > The goal is to not return on field "page" being empty.
> > 
> > What about op=rm se=\S ?
> > 
> 
> For some reason it only returns a few areas that have data in them. I
> would have to think that there are non-white space characters in the
> fields that are not returned.
> 
> If there was a term to just say something like the user tags do in using
> the if statement i.e. if-loop-field do something it would be a little
> easier.
> > > 
<snip>
> > 
> > http://www.icdevgroup.org/i/dev/docfly.html?mv_arg=icdatabase06%2e04

Now tell me why both
	op=rm
	se=\W
	or se=\S
	return the same results.

Now using se=\A (match only on the beginning of a string) works giving
me the results I am looking for.
I thought that \A required a string to process and never thought of
using it till now.

So what is the deal with the match, is it a default to match anything at
the beginning then?

Thanks, and appreciate your tolerance with my ignorance of regular
expressions. These things I have yet to get down pat.

-- 
Philip S. Hempel
debian/rules

http://linuxhardcore.com/