extracting pattern-matched info?

[Anon]

SOrry that the title is a bit nebulous, but I can't figure out a way to solve this problem -

I have many downloaded files (could all be fopen files, but there you have it) and would like to scen them for information that lies between $pattern1 and $pattern2 .

Ok, simple example to try and extract a url from embedded links:

file contains:
<a href="http://www.somesite.com">SomeSite.com</a>

$pattern1="<a href=\"";
$pattern2="\">";
$pattern3="</a>";

So, how would one go about putting the url "http://www.somesite.com" "SomeSite.com" and the entire link "<a href="http.....</a>" into variables (or array),

I have gotten completely lost trying to make the entire file an exploded variable and trying to walk the array!

HELP (please!)

[tasistro]

Ok this of the top of my head so test it before going live

BTW won't work on hrefs split along many lines

<?

$fcontents=file("%your_file_here");
while (list ($line_num, $line) = each ($fcontents)) {
    $pos=0;
    $foundpos=0;
    while ($pos<strlen($line)) {   // while not end of string
        // look for start of required string
        $pos=strpos(strtoupper($line),"<A HREF=\"",$foundpos);
        if ($pos===false)
        {   // pos not found goto end of line
            $pos=strlen($line);
        }   // pos not found
        else
        {   // found pos
            $foundpos=$pos+1;
            // if I've got start then look for end of required string
            $pos=strpos(strtoupper($line),"</A>",$foundpos);
            // get all text between start and end
            $urls[]=substr($line,$foundpos-1,$pos+5-$foundpos);  // check the index here to see if I'm not missing a char or two
            $foundpos=$pos+1;
        }   // found pos
    }   // while not end of string
}  // fcontents while

while (list ($line_num, $line) = each ($urls)) {  // for all found urls
    strtok($line,"\"");
    $href[]=strotk("\""); //this line should echo the data you want  for domain/path
    strtok($line,">");
    $caption[]=strotk("<"); //this line should echo the data you want  for text inside <a> </a>
}  // for all found urls

?>

Saludos
Gerardo

[Anon]

Hi Gerardo -

I appreciate the time you've taken to show me how to traverse the code to extract . It works great (well, when the caption is text, but if it's <img src=> it's not..but I can figure that out.

I'm still a little baffled by the actual 'extraction' once start and end positions are found! I will try and <shudder> get my head around it as I need to extract prices, so need to edit it.

Cheers

--
ps if others use it, there are a couple of syntax errors at the bottome
strtok() not strotk(), but the fatal error will alert you to them.