Implement feed language

lib/Parser/XML/Construct.php

@@ -209,4 +209,22 @@ abstract class Construct {
     protected function getIdDC(): ?string {
         return $this->fetchString("dc:identifier", ".+");
     }
+
+    protected function getLangXML(): ?string {
+        // walk up the tree looking for the nearest language tag
+        $el = $this->subject;
+        do {
+            $out = $this->fetchString("@xml:lang", ".+", false, $el);
+            $el = $el->parentNode;
+        } while (is_null($out) && $el);
+        return $out;
+    }
+
+    protected function getLangDC(): ?string {
+        return $this->fetchString("dc:language", ".+");
+    }
+
+    protected function getLangRss2(): ?string {
+        return $this->fetchString("language", ".+");
+    }
 }

lib/Parser/XML/Feed.php

@@ -83,7 +83,7 @@ class Feed extends Construct implements \MensBeam\Lax\Parser\Feed {
         $feed->meta->url = $this->url;
         $feed->sched = $this->getSchedule();
         $feed->id = $this->getId();
-        //$feed->lang = $this->getLang();
+        $feed->lang = $this->getLang();
         //$feed->url = $this->getUrl();
         //$feed->link = $this->getLink();
         //$feed->title = $this->getTitle();
@@ -115,6 +115,10 @@ class Feed extends Construct implements \MensBeam\Lax\Parser\Feed {
         return $sched;
     }
 
+    public function getLang(): ?string {
+        return $this->getLangXML() ?? $this->getLangDC() ?? $this->getLangRss2();
+    }
+
     public function getUrl(): ?Url {
         return $this->getUrlAtom() ?? $this->getUrlRss1() ?? $this->getUrlPod();
     }
@@ -152,10 +156,6 @@ class Feed extends Construct implements \MensBeam\Lax\Parser\Feed {
         return $this->getEntriesAtom() ?? $this->getEntriesRss1() ?? $this->getEntriesRss2() ?? [];
     }
 
-    public function getLang(): ?string {
-        return null;
-    }
-
     public function getIcon(): ?Url {
         return null;
     }

tests/cases/XML/feed-atom.yaml

@@ -39,3 +39,18 @@ Bogus ID before good:
         format: atom
         version: '1.0'
         id: 'http://example.com/'
+
+Feed language:
+    input: >
+        <feed xmlns="http://www.w3.org/2005/Atom" xml:lang="en"/>
+    output:
+        format: atom
+        version: '1.0'
+        lang: en
+
+Bogus feed language:
+    input: >
+        <feed xmlns="http://www.w3.org/2005/Atom" xml:lang=""/>
+    output:
+        format: atom
+        version: '1.0'

tests/cases/XML/feed-rss0.yaml

@@ -24,3 +24,13 @@ Minimal feed without channel 2:
     output:
         format: rdf
         version: '0.90'
+
+Feed language: # demonstrate walking up the DOM
+    input: >
+        <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns="http://channel.netscape.com/rdf/simple/0.9/" xml:lang="fr">
+            <channel xml:lang=""/>
+        </rdf:RDF>
+    output:
+        format: rdf
+        version: '0.90'
+        lang: fr

tests/cases/XML/feed-rss1.yaml

@@ -150,3 +150,16 @@ Syndication schedule base 1:
         sched:
             interval: PT24H
             base: '2020-03-01T20:21:12-04:00'
+
+DC language:
+    input: >
+        <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns="http://purl.org/rss/1.0/" xmlns:dc="http://purl.org/dc/elements/1.1/">
+            <channel>
+                <dc:language/>
+                <dc:language>de</dc:language>
+            </channel>
+        </rdf:RDF>
+    output:
+        format: rdf
+        version: '1.0'
+        lang: de

tests/cases/XML/feed-rss2.yaml

@@ -177,7 +177,6 @@ Skip all hours:
             expired: true
             skip: 16777215 # 0b111111111111111111111111
 
-
 Skip all days and hours:
     input: >
         <rss><channel>
@@ -222,3 +221,12 @@ Skip all days and hours:
         sched:
             expired: true
             skip: 2147483647 # 0b1111111111111111111111111111111
+
+Feed language:
+    input: >
+        <rss><channel>
+            <language>ja</language>
+        </channel></rss>
+    output:
+        format: rss
+        lang: ja