This is the historical archive of the now-inactive 'grubstreet' list.
Discussion on OpenGuides development has now moved to OpenGuides-Dev. Discussion on The Open Guide to London now takes place on OpenGuides-London.

Re: [grubstreet] Documentation for the search?

[prev] [thread] [next] [lurker] [Date index for 2002/12/18]

From: Kate L Pugh
Subject: Re: [grubstreet] Documentation for the search?
Date: 00:26 on 18 Dec 2002

Ivor wrote:
> Not quite sure how it's parsing that - I would expect it to reject a
> semicolon. The word search just matches words made up of \w characters. It
> specifically treats punctuation including ';' as non-word characters.

Are you sure?  A search on 'takeaway;' returns one page; a search on
'takeaway' many more.

> Also not sure what you are trying to achieve. Will a search on
> takeaway&fulham not do the trick?

For that specific case, yes, but it was just an example.  I'm after a
more general solution to the problem of searching on words that can be
part of other words.  For example, if you search for 'pie' you get the
Oxford Tube page, which has a Piers but no pie.  Really I want to
match on /\bpie\b/i, but the new search doesn't let you use regexes.

> To look for an exact string which includes punctuation characters,
> you need a pair of double quotes round the whole string. Then,
> everything is _literal_ between the "" and case sensitive.

A search on
  ( takeaway |" takeaway;").*fulham
still fails.  Also, I don't want case sensitivity.

>> Also I have a bug report; searching on just '(' or ')' gives an
>> internal server error.
> 
> Not bug but feature. Brackets are used for expression grouping, and
> will be a syntax error if not properly matched. If you want to match
> a single open or close bracket, use the double quotes.

I think it is a bug; it's an error that should be handled.  Well, I'd
consider it a bug, but it's your code.

Kake

-- 
grubstreet mailing list
http://london.openguides.org/old-list-archives/

Re: [grubstreet] Documentation for the search? Ivor Williams 00:11 on 18 Dec 2002
From: "Kate L Pugh" <kake@earth.li>
- Kate L Pugh 00:26 on 18 Dec 2002
  Are you sure? A search on 'takeaway;' returns one page; a search on
Kate L Pugh 00:41 on 18 Dec 2002
As just discovered with Ivor's help over IRC, my confusion over this
- Ivor Williams 01:00 on 18 Dec 2002
  From: "Kate L Pugh" <kake@xxxxx.xx>
  - Kate L Pugh 01:13 on 18 Dec 2002
    It already works like this. It uses LWP::Simple to get the