Add support for categories

WordPress · Nov 29, 2024 · 495be8b · 495be8b
1 parent ca5eab7
commit 495be8b
Show file tree

Hide file tree

Showing 4 changed files with 184 additions and 53 deletions.
diff --git a/packages/playground/data-liberation/src/import/WP_Topological_Sorter.php b/packages/playground/data-liberation/src/import/WP_Topological_Sorter.php
@@ -9,9 +9,8 @@
  */
 class WP_Topological_Sorter {
 
-	public $posts          = array();
-	public $categories     = array();
-	public $category_index = array();
+	public $posts      = array();
+	public $categories = array();
 
 	/**
 	 * Variable for keeping counts of orphaned posts/attachments, it'll also be assigned as temporarly post ID.
@@ -50,9 +49,8 @@ public function map_category( $byte_offset, $data ) {
 		}
 
 		$this->categories[ $data['slug'] ] = array(
-			'parent'      => $data['parent'],
-			'byte_offset' => $byte_offset,
-			'visited'     => false,
+			array_key_exists( 'parent', $data ) ? $data['parent'] : '',
+			$byte_offset,
 		);
 	}
 
@@ -85,8 +83,12 @@ public function map_post( $byte_offset, $data ) {
 
 	/**
 	 * Get the byte offset of an element, and remove it from the list.
+	 *
+	 * @param int $id The ID of the post to get the byte offset.
+	 *
+	 * @return int|bool The byte offset of the post, or false if the post is not found.
 	 */
-	public function get_byte_offset( $id ) {
+	public function get_post_byte_offset( $id ) {
 		if ( ! $this->sorted ) {
 			return false;
 		}
@@ -97,7 +99,7 @@ public function get_byte_offset( $id ) {
 			// Remove the element from the array.
 			unset( $this->posts[ $id ] );
 
-			if ( 0 === count( $this->posts ) ) {
+			if ( 0 === count( $this->categories ) && 0 === count( $this->posts ) ) {
 				// All posts have been processed.
 				$this->reset();
 			}
@@ -108,24 +110,52 @@ public function get_byte_offset( $id ) {
 		return false;
 	}
 
+	/**
+	 * Get the byte offset of an element, and remove it from the list.
+	 *
+	 * @param string $slug The slug of the category to get the byte offset.
+	 *
+	 * @return int|bool The byte offset of the category, or false if the category is not found.
+	 */
+	public function get_category_byte_offset( $slug ) {
+		if ( ! $this->sorted ) {
+			return false;
+		}
+
+		if ( isset( $this->categories[ $slug ] ) ) {
+			$ret = $this->categories[ $slug ];
+
+			// Remove the element from the array.
+			unset( $this->categories[ $slug ] );
+
+			if ( 0 === count( $this->categories ) && 0 === count( $this->posts ) ) {
+				// All categories have been processed.
+				$this->reset();
+			}
+
+			return $ret;
+		}
+
+		return false;
+	}
+
 	public function is_sorted() {
 		return $this->sorted;
 	}
 
 	/**
-	 * Sort posts topologically.
+	 * Sort elements topologically.
 	 *
-	 * Children posts should not be processed before their parent has been processed.
-	 * This method sorts the posts in the order they should be processed.
-	 *
-	 * Sorted posts will be stored as attachments and posts/pages separately.
+	 * Elements should not be processed before their parent has been processed.
+	 * This method sorts the elements in the order they should be processed.
 	 */
 	public function sort_topologically( $free_space = true ) {
 		foreach ( $this->categories as $slug => $category ) {
 			$this->topological_category_sort( $slug, $category );
 		}
 
 		$this->sort_elements( $this->posts );
+		$this->sort_elements( $this->categories );
 
 		// Free some space.
 		if ( $free_space ) {
@@ -136,6 +166,14 @@ public function sort_topologically( $free_space = true ) {
 				// Save only the byte offset.
 				$this->posts[ $id ] = $element[1];
 			}
+
+			/**
+			 * @TODO: all the elements that have not been moved can be flushed away.
+			 */
+			foreach ( $this->categories as $slug => $element ) {
+				// Save only the byte offset.
+				$this->categories[ $slug ] = $element[1];
+			}
 		}
 
 		$this->sorted = true;

diff --git a/packages/playground/data-liberation/tests/WPStreamImporterTests.php b/packages/playground/data-liberation/tests/WPStreamImporterTests.php
@@ -15,24 +15,23 @@ protected function setUp(): void {
 		}
 	}
 
-    /**
-     * @before
+	/**
+	 * @before
 	 *
 	 * TODO: Run each test in a fresh Playground instance instead of sharing the global
 	 * state like this.
-     */
-    public function clean_up_uploads(): void
-    {
-        $files = glob( '/wordpress/wp-content/uploads/*' );
-        foreach( $files as $file ) {
-            if( is_dir( $file ) ) {
-                array_map( 'unlink', glob( "$file/*.*" ) );
-                rmdir( $file );
-            } else {
-                unlink( $file );
-            }
-        }
-    }
+	 */
+	public function clean_up_uploads(): void {
+		$files = glob( '/wordpress/wp-content/uploads/*' );
+		foreach ( $files as $file ) {
+			if ( is_dir( $file ) ) {
+				array_map( 'unlink', glob( "$file/*.*" ) );
+				rmdir( $file );
+			} else {
+				unlink( $file );
+			}
+		}
+	}
 
 	public function test_import_simple_wxr() {
 		$import = data_liberation_import( __DIR__ . '/wxr/small-export.xml' );
@@ -44,7 +43,7 @@ public function test_frontloading() {
 		$wxr_path = __DIR__ . '/wxr/frontloading-1-attachment.xml';
 		$importer = WP_Stream_Importer::create_for_wxr_file( $wxr_path );
 		$this->skip_to_stage( $importer, WP_Stream_Importer::STAGE_FRONTLOAD_ASSETS );
-		while( $importer->next_step() ) {
+		while ( $importer->next_step() ) {
 			// noop
 		}
 		$files = glob( '/wordpress/wp-content/uploads/*' );
@@ -57,17 +56,17 @@ public function test_resume_frontloading() {
 		$importer = WP_Stream_Importer::create_for_wxr_file( $wxr_path );
 		$this->skip_to_stage( $importer, WP_Stream_Importer::STAGE_FRONTLOAD_ASSETS );
 
-		$progress_url = null;
+		$progress_url   = null;
 		$progress_value = null;
-		for($i = 0; $i < 20; ++$i) {
+		for ( $i = 0; $i < 20; ++$i ) {
 			$importer->next_step();
 			$progress = $importer->get_frontloading_progress();
-			if( count( $progress ) === 0 ) {
+			if ( count( $progress ) === 0 ) {
 				continue;
 			}
-			$progress_url = array_keys( $progress )[0];
+			$progress_url   = array_keys( $progress )[0];
 			$progress_value = array_values( $progress )[0];
-			if( null === $progress_value['received'] ) {
+			if ( null === $progress_value['received'] ) {
 				continue;
 			}
 			break;
@@ -78,22 +77,22 @@ public function test_resume_frontloading() {
 		$this->assertEquals( 'https://wpthemetestdata.files.wordpress.com/2008/06/canola2.jpg', $progress_url );
 		$this->assertGreaterThan( 0, $progress_value['total'] );
 
-		$cursor = $importer->get_reentrancy_cursor();
-		$importer = WP_Stream_Importer::create_for_wxr_file( $wxr_path, [], $cursor );
+		$cursor   = $importer->get_reentrancy_cursor();
+		$importer = WP_Stream_Importer::create_for_wxr_file( $wxr_path, array(), $cursor );
 		// Rewind back to the entity we were on.
 		$this->assertTrue( $importer->next_step() );
 
 		// Restart the download of the same entity – from scratch.
-		$progress_value = [];
-		for($i = 0; $i < 20; ++$i) {
+		$progress_value = array();
+		for ( $i = 0; $i < 20; ++$i ) {
 			$importer->next_step();
 			$progress = $importer->get_frontloading_progress();
-			if( count( $progress ) === 0 ) {
+			if ( count( $progress ) === 0 ) {
 				continue;
 			}
-			$progress_url = array_keys( $progress )[0];
+			$progress_url   = array_keys( $progress )[0];
 			$progress_value = array_values( $progress )[0];
-			if( null === $progress_value['received'] ) {
+			if ( null === $progress_value['received'] ) {
 				continue;
 			}
 			break;
@@ -105,17 +104,17 @@ public function test_resume_frontloading() {
 	}
 
 	/**
-	 *
+	 * Test resume entity import.
 	 */
 	public function test_resume_entity_import() {
 		$wxr_path = __DIR__ . '/wxr/entities-options-and-posts.xml';
 		$importer = WP_Stream_Importer::create_for_wxr_file( $wxr_path );
 		$this->skip_to_stage( $importer, WP_Stream_Importer::STAGE_IMPORT_ENTITIES );
 
-		for($i = 0; $i < 11; ++$i) {
+		for ( $i = 0; $i < 11; ++$i ) {
 			$this->assertTrue( $importer->next_step() );
-			$cursor = $importer->get_reentrancy_cursor();
-			$importer = WP_Stream_Importer::create_for_wxr_file( $wxr_path, [], $cursor );
+			$cursor   = $importer->get_reentrancy_cursor();
+			$importer = WP_Stream_Importer::create_for_wxr_file( $wxr_path, array(), $cursor );
 			// Rewind back to the entity we were on.
 			// Note this means we may attempt to insert it twice. It's
 			// the importer's job to detect that and skip the duplicate
@@ -125,6 +124,18 @@ public function test_resume_entity_import() {
 		$this->assertFalse( $importer->next_step() );
 	}
 
+	public function test_sort_categories() {
+		$wxr_path = __DIR__ . '/wxr/mixed-categories.xml';
+		$importer = WP_Stream_Importer::create_for_wxr_file( $wxr_path );
+		$this->skip_to_stage( $importer, WP_Stream_Importer::STAGE_TOPOLOGICAL_SORT );
+
+		while ( $importer->next_step() ) {
+			if ( $importer->get_next_stage() === WP_Stream_Importer::STAGE_FRONTLOAD_ASSETS ) {
+				break;
+			}
+		}
+	}
+
 	private function skip_to_stage( WP_Stream_Importer $importer, string $stage ) {
 		do {
 			while ( $importer->next_step() ) {

diff --git a/packages/playground/data-liberation/tests/WPTopologicalSorterTests.php b/packages/playground/data-liberation/tests/WPTopologicalSorterTests.php
@@ -23,8 +23,8 @@ public function test_parent_after_child() {
 		$sorter->sort_topologically();
 
 		$this->assertEquals( array( 2 => 20, 1 => 10 ), $sorter->posts );
-		$this->assertEquals( 10, $sorter->get_byte_offset( 1 ) );
-		$this->assertEquals( 20, $sorter->get_byte_offset( 2 ) );
+		$this->assertEquals( 10, $sorter->get_post_byte_offset( 1 ) );
+		$this->assertEquals( 20, $sorter->get_post_byte_offset( 2 ) );
 		$this->assertFalse( $sorter->is_sorted() );
 	}
 
@@ -37,7 +37,7 @@ public function test_child_after_parent() {
 		$sorter->sort_topologically();
 
 		$this->assertEquals( array( 1 => 10, 2 => 20, 3 => 30 ), $sorter->posts );
-		$this->assertEquals( 10, $sorter->get_byte_offset( 1 ) );
+		$this->assertEquals( 10, $sorter->get_post_byte_offset( 1 ) );
 	}
 
 	public function test_orphaned_post() {
@@ -48,8 +48,8 @@ public function test_orphaned_post() {
 		$sorter->sort_topologically();
 
 		$this->assertEquals( array( 1 => 10, 2 => 20 ), $sorter->posts );
-		$this->assertEquals( 10, $sorter->get_byte_offset( 1 ) );
-		$this->assertEquals( 20, $sorter->get_byte_offset( 2 ) );
+		$this->assertEquals( 10, $sorter->get_post_byte_offset( 1 ) );
+		$this->assertEquals( 20, $sorter->get_post_byte_offset( 2 ) );
 	}
 
 	public function test_chain_parent_child_after() {
@@ -80,9 +80,9 @@ public function test_get_byte_offsets_consume_array() {
 
 		$this->assertEquals( array( 3 => 30, 2 => 20, 1 => 10 ), $sorter->posts );
 
-		$this->assertEquals( 10, $sorter->get_byte_offset( 1 ) );
-		$this->assertEquals( 20, $sorter->get_byte_offset( 2 ) );
-		$this->assertEquals( 30, $sorter->get_byte_offset( 3 ) );
+		$this->assertEquals( 10, $sorter->get_post_byte_offset( 1 ) );
+		$this->assertEquals( 20, $sorter->get_post_byte_offset( 2 ) );
+		$this->assertEquals( 30, $sorter->get_post_byte_offset( 3 ) );
 		$this->assertCount( 0, $sorter->posts );
 	}
 

diff --git a/packages/playground/data-liberation/tests/wxr/mixed-categories.xml b/packages/playground/data-liberation/tests/wxr/mixed-categories.xml
@@ -0,0 +1,82 @@
+<?xml version="1.0" encoding="UTF-8" ?>
+<rss version="2.0"
+	xmlns:excerpt="http://wordpress.org/export/1.2/excerpt/"
+	xmlns:content="http://purl.org/rss/1.0/modules/content/"
+	xmlns:wfw="http://wellformedweb.org/CommentAPI/"
+	xmlns:dc="http://purl.org/dc/elements/1.1/"
+	xmlns:wp="http://wordpress.org/export/1.2/"
+>
+
+<channel>
+	<title>Mixed Categories</title>
+	<link>https://playground.wordpress.net/scope:funny-chic-valley</link>
+	<description></description>
+	<pubDate>Fri, 29 Nov 2024 12:36:23 +0000</pubDate>
+	<language>en-US</language>
+	<wp:wxr_version>1.2</wp:wxr_version>
+	<wp:base_site_url>https://playground.wordpress.net/scope:funny-chic-valley</wp:base_site_url>
+	<wp:base_blog_url>https://playground.wordpress.net/scope:funny-chic-valley</wp:base_blog_url>
+
+	<wp:author>
+		<wp:author_id>1</wp:author_id>
+		<wp:author_login><![CDATA[admin]]></wp:author_login>
+		<wp:author_email><![CDATA[[email protected]]]></wp:author_email>
+		<wp:author_display_name><![CDATA[admin]]></wp:author_display_name>
+		<wp:author_first_name><![CDATA[]]></wp:author_first_name>
+		<wp:author_last_name><![CDATA[]]></wp:author_last_name>
+	</wp:author>
+
+	<wp:category>
+		<wp:term_id>5</wp:term_id>
+		<wp:category_nicename><![CDATA[parent-of-previous-category]]></wp:category_nicename>
+		<wp:category_parent><![CDATA[]]></wp:category_parent>
+		<wp:cat_name><![CDATA[Parent of previous category]]></wp:cat_name>
+	</wp:category>
+	<wp:category>
+		<wp:term_id>1</wp:term_id>
+		<wp:category_nicename><![CDATA[uncategorized]]></wp:category_nicename>
+		<wp:category_parent><![CDATA[]]></wp:category_parent>
+		<wp:cat_name><![CDATA[Test #1]]></wp:cat_name>
+	</wp:category>
+	<wp:category>
+		<wp:term_id>3</wp:term_id>
+		<wp:category_nicename><![CDATA[test-3]]></wp:category_nicename>
+		<wp:category_parent><![CDATA[uncategorized]]></wp:category_parent>
+		<wp:cat_name><![CDATA[Test #3]]></wp:cat_name>
+	</wp:category>
+	<wp:category>
+		<wp:term_id>2</wp:term_id>
+		<wp:category_nicename><![CDATA[test-child]]></wp:category_nicename>
+		<wp:category_parent><![CDATA[parent-of-previous-category]]></wp:category_parent>
+		<wp:cat_name><![CDATA[Test child]]></wp:cat_name>
+	</wp:category>
+	<wp:term>
+		<wp:term_id>5</wp:term_id>
+		<wp:term_taxonomy><![CDATA[category]]></wp:term_taxonomy>
+		<wp:term_slug><![CDATA[parent-of-previous-category]]></wp:term_slug>
+		<wp:term_parent><![CDATA[]]></wp:term_parent>
+		<wp:term_name><![CDATA[Parent of previous category]]></wp:term_name>
+	</wp:term>
+	<wp:term>
+		<wp:term_id>1</wp:term_id>
+		<wp:term_taxonomy><![CDATA[category]]></wp:term_taxonomy>
+		<wp:term_slug><![CDATA[uncategorized]]></wp:term_slug>
+		<wp:term_parent><![CDATA[]]></wp:term_parent>
+		<wp:term_name><![CDATA[Test #1]]></wp:term_name>
+	</wp:term>
+	<wp:term>
+		<wp:term_id>3</wp:term_id>
+		<wp:term_taxonomy><![CDATA[category]]></wp:term_taxonomy>
+		<wp:term_slug><![CDATA[test-3]]></wp:term_slug>
+		<wp:term_parent><![CDATA[uncategorized]]></wp:term_parent>
+		<wp:term_name><![CDATA[Test #3]]></wp:term_name>
+	</wp:term>
+	<wp:term>
+		<wp:term_id>2</wp:term_id>
+		<wp:term_taxonomy><![CDATA[category]]></wp:term_taxonomy>
+		<wp:term_slug><![CDATA[test-child]]></wp:term_slug>
+		<wp:term_parent><![CDATA[parent-of-previous-category]]></wp:term_parent>
+		<wp:term_name><![CDATA[Test child]]></wp:term_name>
+	</wp:term>
+</channel>
+</rss>