Regex in PHP preg_match_all behaves different than regex101.com

Question

I'm trying to get the numbers in an array. This is my string and my code.

$split_times = "return escape('<table class=\'split\' ><tr><td class=\'split0\'>50m</td><td class=\'split1\'>28.86</td><td class=\'split2\'>28.86</td></tr><tr><td class=\'split0\'>100m</td><td class=\'split1\'>1:01.56</td><td class=\'split2\'>32.70</td></tr><tr><td class=\'splitsep\' colspan=\'3\'></td></tr><tr><td class=\'split0\'>150m</td><td class=\'split1\'>1:36.88</td><td class=\'split2\'>35.32</td></tr><tr><td class=\'split0\'>200m</td><td class=\'split1\'>2:59:09.93</td><td class=\'split2\'>33.05</td></tr></table>')";

preg_match_all("/split1\\\'>(\d+(?:\.\d+)?)</", $split_times, $split_times_distances);
print_r($split_times_distances);

It should return an array like so:

Array
(
    [0] => Array
        (
            [0] => split1\'>28.86<
            [1] => split1\'>1:01.56<
            [2] => split1\'>1:36.88<
            [3] => split1\'>2:59:09.93<
        )

    [1] => Array
        (
            [0] => 28.86
            [1] => 1:01.56
            [2] => 1:36.88
            [3] => 2:59:09.93
        )

)

but instead, it only shows the first index of both arrays.

Your returning values which aren't even in the original content? `class=\'split`\' for example. — Nigel Ren, Jul 10 '17 at 15:55
@clearshot66 It denotes itself as a Ruby regex editor... `a Ruby regular expression editor`, maybe ruby uses PCRE, regex101 has support marked for pcre though. — chris85, Jul 10 '17 at 16:20
@clearshot66 That would depend on the regex you are running, PCRE verbs won't work there. — chris85, Jul 10 '17 at 16:23
@clearshot66 and chris85 both regex testers say I'm doing the right thing: selecting 50, 100 and 200. — Derk Jan Speelman, Jul 10 '17 at 16:24
Possibly the only think worse than trying to parse HTML with a regex ( https://stackoverflow.com/questions/1732348/regex-match-open-tags-except-xhtml-self-contained-tags/1732454 ) is trying to parse HTML embedded in Javascript using a regex. — symcbean, Jul 10 '17 at 16:30
@symcbean can you show me a better way of selecting these values in PHP? The problem here is that I already got this from a dom_xpath query since this html string was inside a onmouse attribute of some other html element on the page I was scraping — Derk Jan Speelman, Jul 10 '17 at 16:32
@chris85 this `preg_match_all` however, returns two arrays: `array:2 [▼ 0 => array:4 [▼ 0 => ">50m<" 1 => ">100m<" 3 => ">200m<" ] 1 => array:4 [▼ 0 => "50" 1 => "100" 3 => "200" ] ]` where the second array is the right one! — Derk Jan Speelman, Jul 10 '17 at 16:36
Let us [continue this discussion in chat](http://chat.stackoverflow.com/rooms/148806/discussion-between-derk-jan-speelman-and-chris85). — Derk Jan Speelman, Jul 10 '17 at 16:54

Toto · Answer 1 · 2017-07-10T18:04:21.190

1

Your regex doesn't match =\'split1\'>1:36.88<

You have to add (?:\d+:){0,2} at the begining.

$split_times = "return escape('<table class=\'split\' ><tr><td class=\'split0\'>50m</td><td class=\'split1\'>28.86</td><td class=\'split2\'>28.86</td></tr><tr><td class=\'split0\'>100m</td><td class=\'split1\'>1:01.56</td><td class=\'split2\'>32.70</td></tr><tr><td class=\'splitsep\' colspan=\'3\'></td></tr><tr><td class=\'split0\'>150m</td><td class=\'split1\'>1:36.88</td><td class=\'split2\'>35.32</td></tr><tr><td class=\'split0\'>200m</td><td class=\'split1\'>2:59:09.93</td><td class=\'split2\'>33.05</td></tr></table>')";

preg_match_all("/split1\\\'>((?:\d+:){0,2}\d+(?:\.\d+)?)</", $split_times, $split_times_distances);
//                    here __^^^^^^^^^^^^^
print_r($split_times_distances);

Output:

Array
(
    [0] => Array
        (
            [0] => split1\'>28.86<
            [1] => split1\'>1:01.56<
            [2] => split1\'>1:36.88<
            [3] => split1\'>2:59:09.93<
        )

    [1] => Array
        (
            [0] => 28.86
            [1] => 1:01.56
            [2] => 1:36.88
            [3] => 2:59:09.93
        )

)

edited Jul 10 '17 at 18:04

answered Jul 10 '17 at 17:15

Toto

89,455
62
89
125

Will this also match values like "1:59:08.69" ? – Derk Jan Speelman Jul 10 '17 at 17:16
@DerkJanSpeelman: No, it will not. For such matches you have to change into `(?:\d+:){0,2}`, see my edit – Toto Jul 10 '17 at 17:19
Id like to select the entire content of the element, no matter what is says. – Derk Jan Speelman Jul 10 '17 at 17:23
Id like so select **all characters** between these two strings: "split0\'>" and "" – Derk Jan Speelman Jul 10 '17 at 17:24
@DerkJanSpeelman: Sorry, I don't get you, it's completely different from OP. – Toto Jul 10 '17 at 17:25
Please note the "2:59:09.93" – Derk Jan Speelman Jul 10 '17 at 17:29
@chris85 I just want to select the value between two strings – Derk Jan Speelman Jul 10 '17 at 17:33
I want to select it by classname like I said in our discussion. I want to select all characters between split0\'> and – Derk Jan Speelman Jul 10 '17 at 17:36

score 1 · Answer 2 · answered Jul 10 '17 at 20:04

You have already extracted your string from an onMouse... attribute using DOMDocument, why not continue? Without using a dedicated Javascript parser, it's easy to extract Javascript strings, then all you have to do is to remove escaped quotes to obtain the "raw" string:

$onMouseAttr = "return escape('<table class=\'split\' ><tr><td class=\'split0\'>50m</td><td class=\'split1\'>28.86</td><td class=\'split2\'>28.86</td></tr><tr><td class=\'split0\'>100m</td><td class=\'split1\'>1:01.56</td><td class=\'split2\'>32.70</td></tr><tr><td class=\'splitsep\' colspan=\'3\'></td></tr><tr><td class=\'split0\'>150m</td><td class=\'split1\'>1:36.88</td><td class=\'split2\'>35.32</td></tr><tr><td class=\'split0\'>200m</td><td class=\'split1\'>2:59:09.93</td><td class=\'split2\'>33.05</td></tr></table>')";

# first step: extracting the strings

$stringPattern = <<<'EOD'
~ " ( [^"\\]* (?:\\.[^"\\]*)* ) "  |  ' ( [^'\\]* (?:\\.[^'\\]*)* ) ' ~xsS
EOD;

if ( preg_match_all($stringPattern, $onMouseAttr, $matches, PREG_SET_ORDER) ) {

    foreach ($matches as $match) {
        # unescape the string for the correct quote
        $html = isset($match[2]) ? str_replace("\\'", "'", $match[2])
                                 : str_replace('\\"', '"', $match[1]);

        # extract the nodes you want with DOMDocument/DOMXPath
        $dom = new DOMDocument;
        $dom->loadHTML($html);
        $xp = new DOMXPath($dom);
        $nodeList = $xp->query('//td[@class="split1"]');
        foreach ($nodeList as $node) {
            # display them
            echo $node->nodeValue, PHP_EOL;
            # or store them
            # $results[] = $node->nodeValue;
        }
    }
}

Thank you for your answer, this is a solid way to do this indeed. What I came up with eventually was this: `preg_match_all("/split0\\\'>(.*?)m", $split_times, $split_times_distances)`. I do have another (what I think is) a really great question about XML and its textContent parameter, I will post another comment here to notify you! — Derk Jan Speelman, Jul 10 '17 at 21:12
[This is my XML question](https://stackoverflow.com/questions/45022059/xml-php-different-ouputs-of-attribute-inside-query-and-outside-query-but-with) — Derk Jan Speelman, Jul 10 '17 at 21:40

score 0 · Accepted Answer · answered Jul 10 '17 at 21:13

What I did was this: simply select all characters between two strings. In this case based on the <td> classname.

print_r(preg_match_all("/split1\\\'>(.*?)</", $split_times, $split_times_distances));

Output:

Array
(
    [0] => Array
        (
            [0] => split1\'>28.86<
            [1] => split1\'>1:01.56<
            [2] => split1\'>1:36.88<
            [3] => split1\'>2:59:09.93<
        )

    [1] => Array
        (
            [0] => 28.86
            [1] => 1:01.56
            [2] => 1:36.88
            [3] => 2:59:09.93
        )

)

Regex in PHP preg_match_all behaves different than regex101.com

3 Answers3

Output: