Improve parsing of link relationships

* Parse the rel attribute in accordance with the WHATWG spec:
  https://infra.spec.whatwg.org/#split-on-ascii-whitespace
* Only list unique rel values in the rel-urls output, fixes #159:
  microformats/microformats2-parsing#30
* Sort the unique rel values alphabetically:
  microformats/microformats2-parsing#29
* Correctly merge attribute values into the resulting object.

Author: Zegnat

Mf2/Parser.php

@@ -1250,16 +1250,15 @@ public function parseRelsAndAlternates() {
 
 		// Iterate through all a, area and link elements with rel attributes
 		foreach ($this->xpath->query('//a[@rel and @href] | //link[@rel and @href] | //area[@rel and @href]') as $hyperlink) {
-			if ($hyperlink->getAttribute('rel') == '') {
+			// Parse the set of rels for the current link
+			$linkRels = array_unique(array_filter(preg_split('/[\t\n\f\r ]/', $hyperlink->getAttribute('rel'))));
+			if (count($linkRels) === 0) {
 				continue;
 			}
 
 			// Resolve the href
 			$href = $this->resolveUrl($hyperlink->getAttribute('href'));
 
-			// Split up the rel into space-separated values
-			$linkRels = array_filter(explode(' ', $hyperlink->getAttribute('rel')));
-
 			$rel_attributes = array();
 
 			if ($hyperlink->hasAttribute('media')) {
@@ -1299,13 +1298,27 @@ public function parseRelsAndAlternates() {
 				$rels[$rel][] = $href;
 			}
 
-			if (!in_array($href, $rel_urls)) {
-				$rel_urls[$href] = array_merge(
-					$rel_attributes, 
-					array('rels' => $linkRels)
-				);
+			if (!array_key_exists($href, $rel_urls)) {
+				$rel_urls[$href] = array('rels' => array());
 			}
 
+			// Add the attributes collected only if they were not already set
+			$rel_urls[$href] = array_merge(
+				$rel_attributes,
+				$rel_urls[$href]
+			);
+
+			// Merge current rels with those already set
+			$rel_urls[$href]['rels'] = array_merge(
+				$rel_urls[$href]['rels'],
+				$linkRels
+			);
+		}
+
+		// Alphabetically sort the rels arrays after removing duplicates
+		foreach ($rel_urls as $href => $object) {
+			$rel_urls[$href]['rels'] = array_unique($rel_urls[$href]['rels']);
+			sort($rel_urls[$href]['rels']);
 		}
 
 		if (empty($rels) and $this->jsonMode) {
@@ -1314,8 +1327,8 @@ public function parseRelsAndAlternates() {
 
 		if (empty($rel_urls) and $this->jsonMode) {
 			$rel_urls = new stdClass();
-		}		
-		
+		}
+
 		return array($rels, $rel_urls, $alternates);
 	}