Changeset 61530

Timestamp:

01/26/2026 04:28:52 PM (2 months ago)

Author:

jonsurrell

Message:

Blocks: Ensure extract_full_block_and_advance() matches parse_blocks()

The behavior of WP_Block_Processor::extract_full_block_and_advance() should produce an identical output to what parse_blocks() would return on the same substring of input.

Unfortunately, when HTML spans followed inner blocks, they were being omitted in the output parse tree. This was due to an omission in the original code which would look for those blocks before advancing again after calling extract_full_block_and_advance() recursively.

This patch adds the missing check and resolves the discrepancy.

Developed in: ​https://github.com/WordPress/wordpress-develop/pull/10769
Discussed in: https://core.trac.wordpress.org/ticket/64538

Follow-up to [60939].

Reviewed by jonsurrell.
Merges [61509] to the 6.9 branch.

Props dmsnell, jonsurrell, jorbin.
Fixes #64537.

Location:

branches/6.9

Files:

• . (modified) (1 prop)
• src/wp-includes/class-wp-block-processor.php (modified) (1 diff)
• tests/phpunit/tests/block-processor/wpBlockProcessor.php (modified) (1 diff)

branches/6.9/src/wp-includes/class-wp-block-processor.php

@@ -1296 +1296 @@
                 $block['innerContent'][] = null;
             }
+
+            /*
+             * Because the parser has advanced past the closing block token, it
+             * may be matched on an HTML span. This needs to be processed before
+             * moving on to the next token at the start of the next loop iteration.
+             */
+            if ( $this->is_html() ) {
+                $chunk                   = $this->get_html_content();
+                $block['innerHTML']     .= $chunk;
+                $block['innerContent'][] = $chunk;
+            }
         }
 

branches/6.9/tests/phpunit/tests/block-processor/wpBlockProcessor.php

@@ -1298 +1298 @@
 
     /**
+     * Ensures that block extraction matches the behavior of the default block parser.
+     *
+     * @ticket 64537
+     *
+     * @dataProvider data_various_block_posts
+     *
+     * @param string $test_document An HTML document to parse as blocks.
+     */
+    public function test_extracts_equivalent_parses_as_parse_blocks( string $test_document ) {
+        $processor = new WP_Block_Processor( $test_document );
+        $blocks    = array();
+
+        while ( $processor->next_block( '*' ) ) {
+            $blocks[] = $processor->extract_full_block_and_advance();
+        }
+
+        $this->assertSame(
+            parse_blocks( $test_document ),
+            $blocks,
+            'Failed to properly parse the block structure.'
+        );
+    }
+
+    /**
+     * Data provider.
+     *
+     * @return Generator
+     */
+    public static function data_various_block_posts() {
+        yield 'Empty post' => array( '' );
+
+        yield 'Void block' => array( '<!-- wp:void /-->' );
+
+        yield 'Empty block' => array( '<!-- wp:empty --><!-- /wp:empty -->' );
+
+        yield 'Paragraph block' => array( '<!-- wp:paragraph --><p>Test</p><!-- /wp:paragraph -->' );
+
+        yield 'Paragraph block with attributes' => array(
+            '<!-- wp:paragraph {"dropCaps": true} --><p>Test</p><!-- /wp:paragraph -->',
+        );
+
+        yield 'Group with void inner' => array(
+            '<!-- wp:group --><!-- wp:void /--><!-- /wp:group -->',
+        );
+
+        /*
+         * @todo There is a hidden bug in here, which is possibly a problem in
+         *       the default parser. There are HTML spans of newlines between
+         *       these block delimiters, and without them, the parse doesn’t
+         *       match `parse_blocks()`. However, `parse_blocks()` is inconsistent
+         *       in its behavior. Whereas it produces an empty text chunk here,
+         *       in the case of a void inner block it produces none. The test is
+         *       being adjusted to step around this issue so that it can be resolved
+         *       separately, and until it’s clear if there is an implementation issue
+         *       with `parse_blocks()` itself.
+         */
+        yield 'Empty columns' => array(
+            <<<HTML
+            <!-- wp:columns -->
+            <!-- wp:column -->
+            <!-- /wp:column -->
+            <!-- /wp:columns -->
+HTML
+            ,
+        );
+
+        yield 'Contentful columns' => array(
+            <<<HTML
+            <!-- wp:columns -->
+            <ul>
+            <!-- wp:column -->
+            <li>A good point.</li>
+            <!-- wp:column /-->
+            </ul>
+            <!-- /wp:columns -->
+HTML
+            ,
+        );
+
+        yield 'Group with mixed content' => array(
+            <<<HTML
+            <!-- wp:group -->
+            <div>
+            <!-- wp:paragraph --><p>Test</p><!-- /wp:paragraph -->
+            This is freeform.
+            <!-- wp:void /-->
+            End
+            <!-- wp:footer -->
+            <footer>That&rsquo;s it!</footer>
+            <!-- /wp:footer -->
+            <!-- wp:another-void /-->
+            </div>
+            <!-- /wp:group -->
+HTML
+            ,
+        );
+
+        yield 'Nested blocks' => array(
+            <<<HTML
+            <!-- wp:a -->
+            <div>
+            <!-- wp:b -->
+            <span><!-- wp:c /--></span>
+            <!-- /wp:b -->
+            </div>
+            <!-- /wp:a -->
+HTML
+            ,
+        );
+
+        yield 'Attributes on nested blocks' => array(
+            <<<HTML
+            <!-- wp:b1 -->
+            <!-- wp:b2 {} -->
+            <!-- wp:b3 {"id":"going"} -->
+            <!-- wp:b4 {"id":"down"} -->
+            <!-- /wp:b4 -->
+            <!-- /wp:b3 -->
+            <!-- /wp:b2 -->
+            <!-- /wp:b1 -->
+HTML
+            ,
+        );
+    }
+
+    /**
      * Data provider.
      *