[SOLVED] Goutte Selectors on Markup that may or may not be present

Goutte Selectors on Markup that may or may not be present

I'm sure this is simple but I'm struggling to get it right. I have the following markup:

<div id="container">
   <h3>Instructions</h3>
   <p>Lorem ipsum dolor sit amet, consectetur adipiscing elit.</p>
   <h3>Directions</h3>
   <p>Lorem ipsum dolor sit amet, consectetur adipiscing elit.</p>
   <h3>Warnings</h3>
   <p>Lorem ipsum dolor sit amet, consectetur adipiscing elit.</p>
</div>

Any of the three elements might be missing and they can be in any order, I want to be able to extract the text in the p tags using goutte and know which one I'm dealing with.

I've tried variations of the following without success:

$node->filter('div#container h3')->each(function (Crawler $node) {
   switch ($node->text() {
      case 'Instructions':
         //$instructions = $node->filter('p')->text();
         //$instructions = $node->closest('p')->text();
         $instructions = $node->parents()->filter('p')->text()
      break;
    //etc....
   }
});

I've also tried using xpath to get preceding-siblings but can't get it right trying things along the lines of

$node->filterXPath("/div[preceding-sibling::h3[normalize-space() = 'Instructions']]");

Solution

It doesn't seem like Crawler has a way of traversing to the next immediate sibling of an element so you may need to use XPath. Use the following-sibling:: axis with a [position() = 1] predicate to limit it to just the very next p that comes after the h3 you want:

$node->filterXPath("/div/h3[normalize-space() = 'Instructions']/following-sibling::p[position() = 1]");