PHP and regular expressions

[Damo]

Hi,
I'm new to this group and regular expressions. I want to extract text
from a newspaper website using regular expressions and php
I'm using this regular expression at the moment

$regexp = "%<table cellspacing=\"0 \" cellpadding=\"0 \"(.+) <img
height=\"1\" alt=\"Today's News\" />%s";

Each news story is in between those tags , So if I extract those
chunks of html using

preg_match($reg exp,$document,$ matches);

where $document is a handle on the file. I can store them in %matches
for further processing. Alas it does not work and i cannot figure out
why

Can anyone help?
Thanks

[BKDotCom]

__where $document is a handle on the file__
the 2nd parameter should be a string
read the file to a string...

Do you really want the space around "(.+)" ?
you can clean this up by enclosing it in '' rather than ""
$regexp = '%<table cellspacing="0" cellpadding="0" (.+) <img
height="1" alt="Today's News" />%s';

On Feb 28, 12:03 pm, "Damo" <cormacdeba...@ gmail.comwrote:

Hi,
I'm new to this group and regular expressions. I want to extract text
from a newspaper website using regular expressions and php
I'm using this regular expression at the moment

$regexp = "%<table cellspacing=\"0 \" cellpadding=\"0 \"(.+) <img
height=\"1\" alt=\"Today's News\" />%s";

Each news story is in between those tags , So if I extract those
chunks of html using

preg_match($reg exp,$document,$ matches);

where $document is a handle on the file. I can store them in %matches
for further processing. Alas it does not work and i cannot figure out
why

Can anyone help?
Thanks

[NC]

On Feb 28, 10:03 am, "Damo" <cormacdeba...@ gmail.comwrote:

>
I'm using this regular expression at the moment

$regexp = "%<table cellspacing=\"0 \" cellpadding=\"0 \"(.+) <img
height=\"1\" alt=\"Today's News\" />%s";

Each news story is in between those tags , So if I extract those
chunks of html using

preg_match($reg exp,$document,$ matches);

where $document is a handle on the file. I can store them in %matches
for further processing. Alas it does not work

How exactly does it not work? What gets into $matches and what value
does
preg_match() return?

Can anyone help?

First, do you really need the whitespace around (.+)? Second,
$document
must be a string, not a handle on the file. Third, your regular
expression
as written is greedy; is this intentional?

Cheers,
NC

[Damo]

>
First, do you really need the whitespace around (.+)?

Does the presence/absence of whitespace make a difference... As I said
I'm new to regex

Second,$documen t must be a string, not a handle on the file.

$document is a handle on a URL taht I was reading in , so ye it was
just a string

Third, your regular expression as written is greedy; is this intentional?

There was no ? at the end of my regular expression it was just (.+)

Cheers,
NC

I got it working since, Thanks for all you help.
cheers

[Rik]

Damo <co***********@ gmail.comwrote:

>First, do you really need the whitespace around (.+)?

Does the presence/absence of whitespace make a difference... As I said
I'm new to regex

Yes, it will match that whitespace unless the /x modifier is set.

>Second,$docume nt must be a string, not a handle on the file.

$document is a handle on a URL taht I was reading in , so ye it was
just a string

>Third, your regular expression as written is greedy; is this
intentional?

There was no ? at the end of my regular expression it was just (.+)

Yes, so it's greedy. It will match as much as possible untill the second
match.

consider:
'<a>foo</a>bar<a>baz</a>foz'

'|<a>.+</a>|' will match '<a>foo</a>bar<a>baz</a>'
'|<a>.+?</a>|' will match '<a>foo</a>'

For a lot of info about regular expressions:
<http://www.regularexpr essions.info>

In your case, I'd possibly use:

$regexp = "%<table[^>]*>(.+?)<img%si" ;

(the /i modifier will make the dot match linebreaks, which is possibly the
breaking point for your regex).

Highly depends on the actual markup wether this will work though...
--
Rik Wasmus

[BKDotCom]

On Mar 1, 4:07 pm, "Damo" <cormacdeba...@ gmail.comwrote:

Second,$documen t must be a string, not a handle on the file.

$document is a handle on a URL that I was reading in , so ye it was just a string

Do we have different definitions of what a handle is?

[OmegaJunior]

On Fri, 02 Mar 2007 06:03:00 +0100, BKDotCom <bk***********@ yahoo.com>
wrote:

On Mar 1, 4:07 pm, "Damo" <cormacdeba...@ gmail.comwrote:

Second,$documen t must be a string, not a handle on the file.

$document is a handle on a URL that I was reading in , so ye it was
just a string

Do we have different definitions of what a handle is?

Yup.
http://nl3.php.net/manual/en/resource.php
--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/

[Damo]

I'm slowly getting the hang of regex's, Instead ive starting a new
thread i thought id ask here. I have this regex:

$regexp = "%<li><a href=\"(.+?)\"> .+[^>]+>([^<]+)</a>.+<span class=
\"date\">(.+ ?)</span>.+<div class=\"abstrac t\">(.+?)</div>%s";

It matches a the URL of a link, the text of a link, a date and the
abstract of a story.
If I use preg_match_all and theres multiple matches to the pattern in
a html string, it inly stores the last one in $matches

this is how i'm printing them:

preg_match_all( $regexp,$docume nt,$matches);
$numMatches = count($matches[1]);
for ($i=0;$i <= $numMatches-1; $i++)
{
echo $matches[1][$i]."<br>";
echo $matches[2][$i]."<br>";
echo $matches[3][$i]."<br>";
echo $matches[4][$i]."<br>";
}

I have no idea why its skipping the earlier matches.
Thanks

[Rik]

On Fri, 02 Mar 2007 18:16:38 +0100, Damo <co***********@ gmail.comwrote:

I'm slowly getting the hang of regex's, Instead ive starting a new
thread i thought id ask here. I have this regex:

$regexp = "%<li><a href=\"(.+?)\"> .+[^>]+>([^<]+)</a>.+<span class=

-------------------------------------^greedy dot--------^again

\"date\">(.+ ?)</span>.+<div class=\"abstrac t\">(.+?)</div>%s";

------------------------^greedy dot

It matches a the URL of a link, the text of a link, a date and the
abstract of a story.
If I use preg_match_all and theres multiple matches to the pattern in
a html string, it inly stores the last one in $matches

$regexp = '%<li>\s*
<a[^>]*?href="([^"]+)"[^>]*>
(.*?)
</a>
..*?
<span[^>]*?class="date"[^>]*>
(.+?)
</span>
..+?
<div[^>]*?class="abstra ct"[^>]*>
(.+?)
</div>%six';

--
Rik Wasmus