PHP 正则表达式 html 具有固定标记的数据属性

Question

I have the following fixed pattern markup scenarios

<div class="myclass" id="id123" data-foo="bar">content</div>
<div class="myclass" id="id123" data-foo="bar" >content</div>
<div class="myclass" id="id123" data-foo="bar" data-baz="qux">content</div>
<div class="myclass" id="id123" data-foo="bar" data-baz="qux" >content</div>

I'm trying to parse the following values out

id123
bar
qux (if it ever exists)

I was able to figure out how to get the different scenarios, but I'm haven't trouble coming up with one final rule that would work for all scenarios.

/<div class="myclass" id="(.*)" data-foo="(.*)"(data-baz="(.*)")?>/

I seem to be missing some basic regex principle. I tried bounding and ending and whitespace but not luck.

Answer 1

1. I do not endorse using regex to parse html, but you say that you are optimizing for speed and that the markup is predictably structured.
2. You just need to use lazy quantifiers with those dots and show a little more care regarding the optional spaces

Code: ( Demo )

$text = <<<TEXT
<div class="myclass" id="id123" data-foo="bar">content</div>
<div class="myclass" id="id123" data-foo="bar" >content</div>
<div class="myclass" id="id123" data-foo="bar" data-baz="qux">content</div>
<div class="myclass" id="id123" data-foo="bar" data-baz="qux" >content</div>
TEXT;

preg_match_all('~<div class="myclass" id="(.*?)" data-foo="(.*?)" ?(?:data-baz="(.*?)" ?)?>~', $text, $matches);
var_export(array_slice($matches, 1));

Output:

  0 => 
  array (
    0 => 'id123',
    1 => 'id123',
    2 => 'id123',
    3 => 'id123',
  ),
  1 => 
  array (
    0 => 'bar',
    1 => 'bar',
    2 => 'bar',
    3 => 'bar',
  ),
  2 => 
  array (
    0 => '',
    1 => '',
    2 => 'qux',
    3 => 'qux',
  ),
)

You can improve the regex efficiency by not using lazy quantifiers. If you know that the attribute values will not contain double-quotes, then you can use a this negated character class with a greedy quantifier: [^"]* .