ex-webby

If so...

how do you match a whole word from a string.

eg. word = "test"

String = "test, test testing....test;"

I need it to match all of those apart from the "test" part of "testing".

Thanks in advance

Simon

JV

Depends what you're using, but in Perl it's /\btest\b/

(\b for word boundary)

HTH,
James

Fosters

what language Simon?

ex-webby

JV

thank you very much. It's actually in PHP.. don't suppose you know if this will use the same format do you? Apologies, I am useless at this!

Cheers

Simon

ex-webby

the actual requirement is to replace the word with another word wherever it appears... but only the full word.

Cheers
Simon

JV

Not sure about PHP, but this might help:

http://uk2.php.net/manual/en/function.preg-replace.php

RallyMarshal

Then cant you do a simple replace function that looks for [space]WORD[space] rather than a regular expression that might strip out content from within another string?

stevencotton

That's what \b does. You can't look for /\stest\s/ since that won't change something like "test, ".

ex-webby

JV

Perfect! Thank you very much, that's done the job

All the best

Simon

icantthinkofone

I think the 'regexp' (purist!??) way would be to use a character class eg {., }word{., } which will match any of the supplied options (rather than the /b); regexp being language independant.

That said, if php supplies a /b style option, use it, it's a lot easier the generating a complex regexp!

stevencotton

Character classes are done with square brackets rather than curly braces, which are used for matching repetition. The regex purist will use what's available in the implementation, rather than sticking to one way of doing things If the PHP regular expression parser is POSIX compliant it should be ok.

ex-webby

OK..

it gets more complicated....

I'm using the \b thing which works perfectly... thank you...

But.. I want it to ignore the text if it is within certain specific tags...

best example would be in html...

testing test <a href="http://www.test.com">test</a> test <img src="http://www.test.com/image.jpg">

in this example, I would want it to ignore the two "test"s between "<a href" and "</a>" and also ignore the test inbetween "<img" and ">"

is this possible, or does it get way too complex?

Cheers

Simon

Fosters

fussy b@stard!

ex-webby

LOL

Fosters

how about a routine that scans the page and replaces everything between "<a href" and "/a>" with spaces. same with "<img" tags and then do your /b thingie?

kinda like a mid(rah, rah, rah)=space$(1) in vb

ex-webby

Was trying to use a regular expression to avoid parsing all the text each time for performance.

If it can't be done in one regular expression, I guess I'll have to run two processes on it.


Cheers
Simon

stevencotton

Regex matching is anything but efficient so you're already imposing a performance hit using them, even if you precompile it. You can't do what you want with one regular expression, you'd be better off with a mixture of some kind of SAX parser (so you can ignore all the HTML tags) and regex the substitution.

Hmm, saying that, I just did this:

my $word_to_replace = 'test';
my $new_word = 'april fools day';

$html =~ s!(>[^<]*?\b)$word_to_replace\b!$new_word!gi;

dsmith

or maybe I wont try to talk about things I know nowt about

Deano