Incorrect Handling of Multiple DOMElement Objects in source_block_raw Method

[veerap000]

I found an issue with the source_block_raw method in /src/parser/content-parser.php.

Current Behavior:

In the current implementation, the function only returns the first matched DOMElement when there are multiple elements within the core/html block. Specifically, it uses:
$attribute_value = trim( $crawler->outerHtml() );

This causes a problem when the core/html block contains multiple DOMElement objects. For example, consider the following HTML content:
<h2>My HTML heading</h2><p>My HTML paragraph</p>

Currently, only the first element (<h2>My HTML heading</h2>) is returned, while the rest of the elements are ignored.

Proposed Solution:

To address this, I propose modifying the function to iterate over all matched elements, collect their outer HTML, and then concatenate them together. This can be accomplished with the following change:

$attribute_values = $crawler->each(function ($node) {
    return trim( $node->outerHtml() );
});

$attribute_value = trim( implode( '', $attribute_values ) );

This approach ensures that all DOMElement objects within the core/html block are captured and returned correctly.

Example Output:
With this change, the example input:
<h2>My HTML heading</h2><p>My HTML paragraph</p>
Will be correctly handled, returning both elements as expected, instead of only the first.

Please let me know if this proposed fix is suitable or if further adjustments are needed.

[alecgeatches]

@veerap000 Thank you for the report! I think your problem explanation and solution look right to me. We'll see if we can get this fixed shortly.

[alecgeatches]

@veerap000 Merged a fix over in #82! We used a slightly different approach:

vip-block-data-api/src/parser/content-parser.php

Lines 685 to 694 in a4939bb

	if ( $crawler->count() > 0 ) {
	// $crawler's outerHtml() will only return the HTML of the first node in this raw HTML.
	// If the raw HTML contains multiple top-level nodes, we need to use the inner HTML of the wrapping
	// 'body' tag. This will also preserve internal whitespace in the HTML.
	$body_node = $crawler->closest( 'body' );
	if ( $body_node && $body_node->count() > 0 ) {
	$attribute_value = trim( $body_node->html() );
	}
	}

I noticed during testing that the trim() + implode() method would remove inner whitespace, which can be important for text like this:

<span>These</span> <span class="highlight">spaces</span> <span>are</span> <span>important</span>.

The new method still trims the outer string but will retain inner whitespace, including things like newlines and non-breaking spaces.

Thank you for reporting this! We have a small test fix to do, and should have a new release version out shortly.

[alecgeatches]

@veerap000 This has been fixed in release 1.4.2. Please let us know if you encounter any problems with the fix. Thank you for the report!