REGEX Question: Get filenames and alt-tags from html

[Smasal]

I have 2 problems with regular expressions:

1) I want to get the image's filenames from a website:

e.g. src="http://bytes.com/images/world.jpg", src=images/world.jpg and src='images/world.jpg' should result in "world"

$imgfilenames=preg_match_all('/^[0-9A-Za-z_ ](.jpg|.gif|.JPG|.GIF|.png|.PNG)$/i', $html, $matches);

2) I want to get the alt descriptions from images in a website

$alttags=preg_match_all('/<img[^>]*alt="([^"]*)"/i', $html, $matches);

But both my expressions don't work because I don't really get it ;-).

[Atli]

Your first expression should work, except that the first character class is defined as just a single character. If you want a class to cover more than one character, it needs to be trailed by one of:

• * - Any number of characters, (including none).
• ? - 0 or 1 characters.
• + - One or more characters.
• {high,low} - A specific range of characters.

For example:
[code=regexp][a-zA-Z0-9]+[/code]
This matches one or more alphanumeric character. If you remove the + it only matches one (which is what your expression does).

See regular-expressions.inf o for more details.

P.S.
It is usually better to parse HTML documents using a DOM parser, or one of the XML parsers. - Regular expressions are poorly suited to parse large chunks of HTML, because HTML is not a "regular" format.