ensure http -> https when destination is https

leonstafford · leonstafford · commit 56ccddf22e57 · 2020-05-21T04:20:51.000+09:30
diff --git a/src/HTMLProcessor.php b/src/HTMLProcessor.php
@@ -667,7 +667,21 @@ public function detectUnchangedPlaceholderURLs( string $processed_html ) : strin
         $destination_host = (string) parse_url( $this->base_url, PHP_URL_HOST );
 
         if ( strpos( $processed_html, $placeholder_host ) !== false ) {
-            return str_replace( $placeholder_host, $destination_host, $processed_html );
+            // bulk replace hosts
+            $processed_html = str_replace(
+                $placeholder_host,
+                $destination_host,
+                $processed_html
+            );
+        }
+
+        // force http -> https if destination is https
+        if ( $this->destination_protocol === 'https://' ) {
+            $processed_html = str_replace(
+                'http://' . $destination_host,
+                'https://' . $destination_host,
+                $processed_html
+            );
         }
 
         return $processed_html;
diff --git a/tests/HTMLProcessorTest.php b/tests/HTMLProcessorTest.php
@@ -119,7 +119,6 @@ public function internalLinkProvider() {
      * @covers StaticHTMLOutput\HTMLProcessor::__construct
      * @covers StaticHTMLOutput\HTMLProcessor::isInternalLink
      * @covers StaticHTMLOutput\HTMLProcessor::normalizeURL
-     * @covers StaticHTMLOutput\HTMLProcessor::normalizeURL
      * @dataProvider anchorTagProvider
      */
     public function testNormalizePartialURLInAnchor(
@@ -584,6 +583,7 @@ public function unicodeProvider() {
      * @covers StaticHTMLOutput\HTMLProcessor::detectEscapedSiteURLs
      * @covers StaticHTMLOutput\HTMLProcessor::detectIfURLsShouldBeHarvested
      * @covers StaticHTMLOutput\HTMLProcessor::detectUnchangedPlaceholderURLs
+     * @covers StaticHTMLOutput\HTMLProcessor::getBaseURLRewritePatterns
      * @covers StaticHTMLOutput\HTMLProcessor::getHTML
      * @covers StaticHTMLOutput\HTMLProcessor::getProtocolRelativeURL
      * @covers StaticHTMLOutput\HTMLProcessor::getTargetSiteProtocol
@@ -772,6 +772,15 @@ public function processlinkProvider() {
                 '<link rel="https://api.w.org/" href="http://example.com/wp-json/">' .
                 '</head><title>title</title><body>body</body></html>' . PHP_EOL,
             ],
+            'protocol relative link always rewritten to destination protocol' => [
+                false,
+                '<!DOCTYPE html><html lang="en-US"><head>' .
+                '<link rel="prefetch" href="//mydomain.com">' .
+                '</head><title>title</title><body>body</body></html>',
+                '<!DOCTYPE html>' . PHP_EOL . '<html lang="en-US"><head>' .
+                '<link rel="prefetch" href="https://mynewdomain.com">' .
+                '</head><title>title</title><body>body</body></html>' . PHP_EOL,
+            ],
         ];
     }
 
