fix: replacement was not working for urls with special chars

Codeinwp · Aug 6, 2019 · 48a4966 · 48a4966
1 parent 1ab1739
commit 48a4966
Show file tree

Hide file tree

Showing 2 changed files with 10 additions and 2 deletions.
diff --git a/inc/manager.php b/inc/manager.php
@@ -275,7 +275,7 @@ public function process_urls_from_json( $json ) {
 	 * @return array array of urls.
 	 */
 	public function extract_urls_from_json( $content ) {
-		$regex = '/(?<!(=|\\\\)(?:"|\'|"))(?:http(?:s?):)(?:[\/\\\\|.|\w|\s|-])*\.(?:' . implode( '|', array_keys( Optml_Config::$extensions ) ) . ')(?:\??[\w|=|&|\-|\.|:]*)/';
+		$regex = '/(?<!(=|\\\\)(?:"|\'|"))(?:http(?:s?):)(?:[\/\\\\|.|\w|\s|@|%|-])*\.(?:' . implode( '|', array_keys( Optml_Config::$extensions ) ) . ')(?:\??[\w|=|&|\-|\.|:]*)/';
 		preg_match_all(
 			$regex,
 			$content,
@@ -481,7 +481,7 @@ public function process_urls_from_content( $html ) {
 	 * @return array
 	 */
 	public function extract_image_urls_from_content( $content ) {
-		$regex = '/(?:http(?:s?):)(?:[\/\\\\|.|\w|\s|-])*\.(?:' . implode( '|', array_keys( Optml_Config::$extensions ) ) . ')(?:\?{1}[\w|=|&|\-|\.|:|;]*)?/';
+		$regex = '/(?:http(?:s?):)(?:[\/\\\\|.|\w|\s|@|%|-])*\.(?:' . implode( '|', array_keys( Optml_Config::$extensions ) ) . ')(?:\?{1}[\w|=|&|\-|\.|:|;]*)?/';
 		preg_match_all(
 			$regex,
 			$content,

diff --git a/tests/test-replacer.php b/tests/test-replacer.php
@@ -171,7 +171,15 @@ public function test_replacement_with_relative_url() {
 		$this->assertContains( 'i.optimole.com', $replaced_content );
 
 	}
+	public function test_replacement_strange_chars(){
+		$content = '
+		https://www.example.org/wp-content/uploads/2018/05/@brands.png
+		https://www.example.org/wp-content/uploads/2018/05/%brands.png
+		';
+		$replaced_content = Optml_Manager::instance()->replace_content( $content );
+		$this->assertEquals( 2, substr_count( $replaced_content, 'i.optimole.com' ) );
 
+	}
 	// TODO We need to extend this to single url replacement. If we make the url extractor regex with option scheme, the parsing will take huge amount of time. We need to think alternatives.
 	public function test_replacement_without_scheme() {
 		$content          = '<div class="before-footer">