str

Chris_Evans

Hi there,
Assume for a minute that I have the following in a document.

<a href="http://www.domain.com">domain</a>
<meta name="nameofmeta" content="wohoo, metacontenthere">

What I'd like to do is to extract the data inside the the "" after the href and extract the data in name and content after inside the "".

How abouts would I do this? There are several ways I can think of but such as preg but I would imagine somebody here has a better way. 🙂

It needs to be fairly standardised, for example I could get the below:

<a attributes-here href="domain.com">domain-hello</a>

Any help would be appreciated,

Chris 🙂

zapa

this is in my snippet lib:

<?php 

// Grab source code from a file or web site 
if(!($myFile=fopen("http://www.url.com","r"))) 
{ 
echo "Sorry! This feature is currently not available."; 
exit; 
} 
while(!feof($myFile)) 
{ 
// Read each line and add to $myLine 
$myLine.=fgets($myFile,255); 
} 
fclose($myFile); 
// Extract everything between start and end. You need to include these lines 
//in the headlines or pick some unique substring in the html to mark the start 
//and end of the news. 
$start="<span class=\"postbody\">"; 
$end="</span>"; 
$start_position=strpos($myLine, $start); 
$end_position=strpos($myLine, $end)+strlen($end); 
$length=$end_position-$start_position; 
$myLine=substr($myLine, $start_position, $length); 
// Display HTML 

echo $myLine; 

?>

Chris_Evans

Hi there,
It appears now as though I need to use something along these lines:

$start_position=strpos($myLine, $start);  

$end_position=strpos($myLine, $end)+strlen($end);

So I can get the start and the end position by $start = "<a" and $end = "</a>". That's fine but now how do I get just the href="domain.com" bit out of it?

Thanks,

Chris

BuzzLY

why not use preg_match? It doesn't get much more straightforward than that...

Chris_Evans

Hi there,
I've looked at preg_match but I can't see how to do it. :S

preg_match("<a href---nowwherefromhere?">somethingheretosayanything</a>", $data, $matches);

Chris.

BuzzLY

Here is an example. It should get you started in the right direction:

<?php
$input1 = '<a href="http://www.someurl.com">someurl</a>';
$input2 = '<a class="yada" href="http://www.anotherurl.com">anotherurl</a>';
$input3 = '<a style="color:#000000;" HREF="http://www.onemoreURL.com" class="yada">onemoreURL</a>';

$pattern = "§<a.*href=\"(.*?)\".*>§i";
preg_match($pattern,$input1,$matches1);
preg_match($pattern,$input2,$matches2);
preg_match($pattern,$input3,$matches3);

echo "Input1 href: " . $matches1[1] . "<br/>";
echo "Input2 href: " . $matches2[1] . "<br/>";
echo "Input3 href: " . $matches3[1] . "<br/>";
?>

Make sure you understand that the question mark (?) is a modifier that makes the match ungreedy. For more information, take a look at this page.

Chris_Evans

Hi,
Thanks very much for that BuzzLY. Out of interest, what does the § symbol do? I can't find it on the pattern syntax site.

Thanks in advance,

Chris

Weedpacket

Originally posted by Chris_Evans
Hi,
Thanks very much for that BuzzLY. Out of interest, what does the § symbol do? I can't find it on the pattern syntax site.

It's just his preferred pattern delimiter (PCRE expressions use them to show where the pattern leaves off and the modifiers (like i) begin). Traditionally / has been used, but almost any nonalphanumeric character could be used instead. The advantage of § over / is that if you use / you have to write \/ every time you want to use a slash in your pattern (which can be quite often if you're working with HTML). § is much less likely to appear.

Chris_Evans

Hi there,
Thanks very much for all your help! 🙂 I think I've finally got an understanding now of pattern syntaxes/expressions. Since I started PHP (ages ago) I've never been able to grasp these till you wrote that so thanks! 🙂

Thanks,

Chris

BuzzLY

Yeah, what Weedpacket said. Long ago I decided that I would use a symbol other than / to delimit pcre patterns. I have used @, %, and even ® and ©. But even those can appear in the pattern, albeit rarely. Since I first started using regex, I have never once had to search for §. So, it became my pattern delimiter.

Besides, it's a cool looking symbol 😃

Incidentally, in case you simply copy/pasted it, if you want to use it you just have to type ALT-0167 on your numpad. Another cool symbol to use is ¥ (ALT-0165)

Chris_Evans

Hi - that's superb. I have a small list of ALT numbers now. Due to the business we're in, my most memorable is: ° (ALT + 0176).

So basically, that first character is like some kind of seperator from the main content?

Weedpacket

Originally posted by Chris_Evans
So basically, that first character is like some kind of seperator from the main content?

Sort of; The second § indicates where the pattern ends and the modifiers begin. But since almost any character could be used, the regexp parser has to know which one is being used. So that's the first thing the regexp parser is told.

§ "Ah, so the pattern goes on up to the §"
<a.href=\"(.?)\".*> "This must be the pattern, then..."
§ "...because here's where I've been told it ends."
i "Oh, look. A modifier. How cute."

BuzzLY

I never knew regex parsers had such personality 😃

Chris_Evans

Hi there,
Ok, thanks for the info. Would you mind explaining the i in a little more detail too please? 🙂 You've been a great help!

Thanks,

Chris

BuzzLY

i is a pattern modifier, as Weedpacket pointed out.

Notice that the third input string contains "HREF", not "href". Without the i modifier, it would not have been a match.

Chris_Evans

Hi,
Thanks for that! 🙂