[Resolved] Perl Regular Expressions

[jsvoyager]

I'm trying to turn
[[cell1][cell2]]
into
<tr><td>cell1</td><td>cell2</td></tr>
using regular expressions. I'm really bad at them so I really have no clue what I'm doing. Anyone got any ideas? If you could also provide a little commentary with your expression disecting each part, that'd be nice so I could learn how to do this.

thanks

[iceraider]

$var = "The cell [[cell1][cell2]]"; // The string with the cell in it

$result = str_replace('[[cell1][cell2]]', '<tr><td>cell1</td><td>cell2</td></tr>', $var); // Store the new string in $result

[jsvoyager]

but that only replaces a literal cell1 and cell2 doesn't it. I meant like
[[whatever data is in here][and in here]]
to
<tr><td>whatever data is in here</td><td>and in here</td>

Since I posted I have been able to make this regex, but it only kinda works:

$result = preg_replace(
	  "/\\[\\[(\w+)\\]\\[(\w+)\\]\\]/i",
	  "<tr><td>\$1</td><td>\$2</td></tr>",
	  $output);

It only works so far as the content between the brackets does not have a space and there are only two cells. I'd like to make it work with spaces at the very least and with any number of cells if possible. Thanks

[rehfeld]

try this

$result = preg_replace(
	  "/\[\[(.*?)\]\[(.*?)\]\]/s",
	  "<tr><td>\$1</td><td>\$2</td></tr>",
	  $output);

i used the s modifier, . normally matches any character except a new line. the s modifier makes . match any character including a new line

or this

$result = preg_replace(
	  "/\[\[([^\]]+)\]\[([^\]]+)\]\]/",
	  "<tr><td>\$1</td><td>\$2</td></tr>",
	  $output);

using a negated character class. one or more characters that are not a literal ]

[mtmosier]

To replace any number of cells (and rows) it gets a bit more complicated.

$data = 'Random data.  [[cell1][cell2][cell3]][[cell4][cell5][cell6]]  More randomness.';

if (preg_match_all('/\[(\[.*?\])\]/s', $data, $matches)) {
    $count = count($matches[0]);
    for ($i = 0; $i < $count; $i++) {
        $workstr = $matches[1][$i];
        $tmp = '';
        while ($tmp != $workstr) {
            $tmp = $workstr;
            $workstr = preg_replace('/\[([^\]]*)\]/s', '<td>\1</td>', $workstr, 1);
        }

    $data = preg_replace('/\[\[.*?\]\]/s', '<tr>'.$workstr.'</tr>', $data, 1);
}
}

Let me try to explain this...

The preg_match_all is used to separate out one row at a time, where anything starting with "[[" and ending with "]]" is a row. The ".*?" in this says to match anything number of characters in a non-greedy manner. You can read about greediness on this page if you like.

Then for each of these rows, it goes through and changes each cell one at a time into the proper format.

Finally it again matches the single row and replaces it with the new cell content it just worked out.

[jsvoyager]

Thank you guys soooooo much for the help. I decided to add a feature, and that was that [align=”right”|Cell's Content] becomes <td align=”right”>Cell's Content</td>. I've almost got it working. Here's my code:

	if (preg_match_all('/\[(\[.*?\])\]/s', $data, $matches)) {
	    $count = count($matches[0]);
	    for ($i = 0; $i < $count; $i++) {
	        $workstr = $matches[1][$i];
	        $tmp = '';
	        while ($tmp != $workstr) {
	            $tmp = $workstr;
	            $workstr = preg_replace('/\[([^\|]*\|)?([^\]]*)\]/s', '<td \1>\2</td>', $workstr, 1);
	        }

        $data = preg_replace('/\[\[.*?\]\]/s', '<tr>'.$workstr."</tr>\n", $data, 1);
    }
$data = "\n<table width=\"100%\" border=\"0\">\n$data</table>\n\n";
}

EDIT: Put final $data... line in if statement

The only problem is that it outputs the pipe (ie, <td align=”right”|>Content</td>). Any thoughts on how to get rid of this? Also, I's thinking about making it so that you can some how also pass the attributes for the whole table somehow, but I really have no idea how. If anyone has ideas on how to solve either of these problems, I'd love to hear them. Thanks.

[rehfeld]

personally i wouldnt start trying to integrate html attributes into it.

i would just alter the content of the cell

[<div align="right">blah</div>]

you bound to run into a situation where youll need to use an attribute other than align, and unless you plan on just modifying the regex each time to support the new attribute, i would just alter the cells content(if its possible)

[jsvoyager]

the main reason I's going to incorporate this is that sometimes I need to use the colspan attr. I don't think there's a way around that one.

[jsvoyager]

I've decided to just add the following line above the last $data = .... line:

$data = str_replace("|>", ">", $data);

After all, anyone using a > as a character should be escaping it :-). If you have a more elegant solution, please let me know.

[mtmosier]

Will this string ever contain anything other than your table definition? If not you could put your table attributes at the very beginning, similar to how you're doing your cell attributes.

Anyway, here's how I would get rid of that bar.

$workstr = preg_replace('/\[(([^\|]*)\|)?([^\]]*)\]/s', '<td \2>\3</td>', $workstr, 1);

Edit: No reason to repost all the code. Sorry.

[jsvoyager]

good idea, I can't believe I didn't think of that!
Regarding the table attrs, the string does contain other data (this is actually all a Smarty Filter). I guess if I want a table with special attr, I'll just do it the long way; 90% of my tables use these attr.