[Bug 344389] feed parser doesn't expose guid/atom:id/rdf:about correctly. r=ben

2025-02-15 04:39:31 +00:00 · 2006-07-19 19:31:52 +00:00 · 2006-07-19 19:31:52 +00:00 · 3441890217
commit 3441890217
parent 9a79b08c79
10 changed files with 233 additions and 17 deletions
--- a/toolkit/components/feeds/public/nsIFeedContainer.idl
+++ b/toolkit/components/feeds/public/nsIFeedContainer.idl
@ -45,14 +45,14 @@ interface nsIArray;
 * but they have some divergent attributes and require
 * different convenience methods.
 */ 
-[scriptable, uuid(c7e65344-df1f-4a75-9615-d58357c1554b)]
+[scriptable, uuid(b7e7b6b9-a461-4817-9003-3e6b9005945f)]
 interface nsIFeedContainer : nsISupports 
 {
  /**
  * Many feeds contain an ID distinct from their URI, and
  * entries have standard fields for this in all major formats.
  */
-  attribute AString identifier;
+  attribute AString id;

  /**
   * The baseURI for the Entry or Feed.
--- a/toolkit/components/feeds/src/FeedProcessor.js
+++ b/toolkit/components/feeds/src/FeedProcessor.js
@ -59,7 +59,12 @@ const XMLNS = "http://www.w3.org/XML/1998/namespace";
 /***** Some general utils *****/
 function strToURI(link, base) {
  var base = base || null;
-  return gIoService.newURI(link, null, base);
+  try {
+    return gIoService.newURI(link, null, base);
+  }
+  catch(e) {
+    return null;
+  }
 }

 function isArray(a) {
@ -178,6 +183,19 @@ function bagHasKey(bag, key) {
  }
 }

+function makePropGetter(key) {
+  return function FeedPropGetter(bag) {
+    try {
+      return value = bag.getProperty(key);
+    }
+    catch(e) {
+    }
+    return null;
+  }
+}
+
+
+
 /**
 * XXX Thunderbird's W3C-DTF function
 *
@ -325,6 +343,7 @@ function Feed() {
  this._title = null;
  this.items = [];
  this.link = null;
+  this.id = null;
  this.baseURI = null;
 }
 Feed.prototype = {
@ -349,6 +368,7 @@ Feed.prototype = {
    _sub: ["description","dc:description","rss1:description",
           "atom03:tagline","atom:subtitle"],
    items: ["items","atom03_entries","entries"],
+    id: ["atom:id","rdf:about"],
    _title: ["title","rss1:title", "atom03:title","atom:title"],
    link:  [["link",strToURI],["rss1:link",strToURI]],
    categories: ["categories", "dc:subject"],
@ -390,7 +410,7 @@ Feed.prototype = {
      }
    }
  },
-
+  
  QueryInterface: function Feed_QI(iid) {
    if (iid.equals(Ci.nsIFeed) ||
        iid.equals(Ci.nsIFeedContainer) ||
@ -407,6 +427,7 @@ function Entry() {
  this.fields = Cc["@mozilla.org/hash-property-bag;1"].
                createInstance(Ci.nsIWritablePropertyBag2);
  this.link = null;
+  this.id = null;
  this.baseURI = null;
 }

@ -444,8 +465,10 @@ Entry.prototype = {
  mediaContent: null,

  searchLists: {
-    _title: ["title","rss1:title","atom03:title","atom:title"],
+    _title: ["title", "rss1:title", "atom03:title", "atom:title"],
    link: [["link",strToURI],["rss1:link",strToURI]],
+    id: [["guid", makePropGetter("guid")], "rdf:about",
+         "atom03:id", "atom:id"],
    _summary: ["description", "rss1:description", "dc:description",
               "atom03:summary", "atom:summary"],
    _content: ["content:encoded","atom03:content","atom:content"]
@ -453,9 +476,20 @@ Entry.prototype = {

  normalize: function Entry_normalize() {
    fieldsToObj(this, this.searchLists);
+
    // Assign Atom link if needed
    if (bagHasKey(this.fields, "links"))
      this._atomLinksToURI();
+
+    // The link might be a guid w/ permalink=true
+    if (!this.link && bagHasKey(this.fields, "guid")) {
+      var guid = this.fields.getProperty("guid");
+      if (bagHasKey(guid, "isPermaLink")) {
+        var isPermaLink = new Boolean(guid.getProperty("isPermaLink"));
+        if (isPermaLink)
+          this.link = strToURI(guid.getProperty("guid"));
+      }
+    }
  },

  QueryInterface: function(iid) {
@ -1134,6 +1168,7 @@ FeedProcessor.prototype = {
      this._state = "IN_" + elementInfo.fieldName.toUpperCase();
      this._docVerified(elementInfo.feedVersion);
      this._stack.push([this._feed, this._state]);
+      this._mapAttributes(this._feed, attributes);
    }
    else {
      this._state = this._processComplexElement(elementInfo, attributes);
@ -1193,7 +1228,7 @@ FeedProcessor.prototype = {
  function FP__processComplexElement(elementInfo, attributes) {
    var obj, props, key, prefix;

-    // If the container is a feed or entry, it'll need to have its 
+    // If the container is an entry/item, it'll need to have its 
    // more esoteric properties put in the 'fields' property bag, and set its
    // parent.
    if (elementInfo.containerClass) {
@ -1208,15 +1243,7 @@ FeedProcessor.prototype = {
      props = obj;
    }

-    // Cycle through the attributes, and set our properties using the
-    // prefix:localNames we find in our namespace dictionary.
-    for (var i = 0; i < attributes.length; ++i) {
-      prefix = gNamespaces[attributes.getURI(i)] ? 
-        gNamespaces[attributes.getURI(i)] + ":" : "";
-      key = prefix + attributes.getLocalName(i);
-      var val = attributes.getValue(i);
-      props.setPropertyAsAString(key, val);
-    }
+    this._mapAttributes(props, attributes);

    // We should have a container/propertyBag that's had its
    // attributes processed. Now we need to attach it to its
@ -1300,6 +1327,19 @@ FeedProcessor.prototype = {
      container.replaceElementAt(element, container.length - 1, false);
  },
  
+
+  _mapAttributes: function FP__mapAttributes(bag, attributes) {
+    // Cycle through the attributes, and set our properties using the
+    // prefix:localNames we find in our namespace dictionary.
+    for (var i = 0; i < attributes.length; ++i) {
+      prefix = gNamespaces[attributes.getURI(i)] ? 
+        gNamespaces[attributes.getURI(i)] + ":" : "";
+      key = prefix + attributes.getLocalName(i);
+      var val = attributes.getValue(i);
+      bag.setPropertyAsAString(key, val);
+    }
+  },
+
  // unknown element values are returned here. See startElement above
  // for how this works.
  returnFromExtHandler:
--- a/toolkit/components/feeds/test/xml/rfc4287/entry_id.xml
+++ b/toolkit/components/feeds/test/xml/rfc4287/entry_id.xml
@ -0,0 +1,35 @@
+<?xml version="1.0" encoding="iso-8859-1"?>
+<!--
+
+Description: atom entry id
+Expect: feed.items.queryElementAt(1, Components.interfaces.nsIFeedEntry).id == "http://foo.example.com/hmm/ok,2006,07,11";
+
+-->
+<feed xmlns="http://www.w3.org/2005/Atom"
+      xmlns:foo="http://www.example.org"
+      foo:quux="quuux">
+
+  <title>hmm</title>
+
+  <author>
+    <email>hmm@example.com</email>
+    <name>foo</name>
+  </author>
+  <generator version="1.1" uri="http://example.org">Hmm</generator>
+  <author>
+    <email>bar@example.com</email>
+    <name>foo</name>
+  </author>
+  <rights type="xhtml">
+    <div xmlns="http://www.w3.org/1999/xhtml"><i>test</i> rights</div>
+  </rights>
+  
+
+  <entry></entry>
+
+  <entry foo:bar="baz">
+    <title>test</title>
+    <id>http://foo.example.com/hmm/ok,2006,07,11</id>
+  </entry>
+
+</feed>
--- a/toolkit/components/feeds/test/xml/rfc4287/feed_id.xml
+++ b/toolkit/components/feeds/test/xml/rfc4287/feed_id.xml
@ -2,7 +2,7 @@
 <!--

 Description: atom author name works
-Expect: feed.fields.getProperty('atom:id') == 'urn:uuid:60a76c80-d399-11d9-b93C-0003939e0af6'
+Expect: feed.fields.getProperty('atom:id') == 'urn:uuid:60a76c80-d399-11d9-b93C-0003939e0af6' && feed.id == 'urn:uuid:60a76c80-d399-11d9-b93C-0003939e0af6'

 -->

@ -24,4 +24,4 @@ Expect: feed.fields.getProperty('atom:id') == 'urn:uuid:60a76c80-d399-11d9-b93C-
    <summary>Some text.</summary>
  </entry>
  
-</feed>
+</feed>
--- a/toolkit/components/feeds/test/xml/rss1/feed_id.xml
+++ b/toolkit/components/feeds/test/xml/rss1/feed_id.xml
@ -0,0 +1,25 @@
+<?xml version="1.0" encoding="utf-8"?>
+<!--
+
+Description: RSS1 feed w/ feed rdf:about
+Expect: feed.id == 'http://www.xml.com/xml/news.rss'
+
+-->
+<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 rdf:about="http://www.xml.com/xml/news.rss">
+    <title>Test</title>
+    <link>http://xml.com/pub</link>
+    <dc:description>another description</dc:description>  
+  </channel>
+  <item rdf:about="http://example.com/hmm">
+   <title>XML: A Disruptive Technology</title> 
+   <link>http://c.moreover.com/click/here.pl?r123</link>
+   <dc:description>
+      XML is placing increasingly heavy loads on the existing technical
+      infrastructure of the Internet.
+    </dc:description>
+  </item> 
+</rdf:RDF>
--- a/toolkit/components/feeds/test/xml/rss1/item_id.xml
+++ b/toolkit/components/feeds/test/xml/rss1/item_id.xml
@ -0,0 +1,32 @@
+<?xml version="1.0" encoding="utf-8"?>
+<!--
+
+Description: RSS1 feed w/ item rdf:about
+Expect: feed.items.queryElementAt(0, Components.interfaces.nsIFeedEntry).id == 'http://example.com/hmm'
+
+-->
+<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 rdf:about="http://www.xml.com/xml/news.rss">
+    <title>Test</title>
+    <link>http://xml.com/pub</link>
+    <dc:description>another description</dc:description>  
+  </channel>
+  <item rdf:about="http://example.com/hmm">
+   <title>XML: A Disruptive Technology</title> 
+   <link>http://c.moreover.com/click/here.pl?r123</link>
+   <dc:description>
+      XML is placing increasingly heavy loads on the existing technical
+      infrastructure of the Internet.
+    </dc:description>
+   <!-- <dc:publisher>The O'Reilly Network</dc:publisher>
+    <dc:creator>Simon St.Laurent (mailto:simonstl@simonstl.com)</dc:creator>
+    <dc:rights>Copyright &#169; 2000 O'Reilly &amp; Associates, Inc.</dc:rights>
+    <dc:subject>XML</dc:subject>-->
+  </item> 
+  <item>
+
+  </item>
+</rdf:RDF>
--- a/toolkit/components/feeds/test/xml/rss2/item_guid_bogus_url.xml
+++ b/toolkit/components/feeds/test/xml/rss2/item_guid_bogus_url.xml
@ -0,0 +1,21 @@
+<?xml version="1.0" encoding="iso-8859-1"?>
+<!--
+
+Description: item copes with bogus guid
+Expect: feed.items.queryElementAt(0, Components.interfaces.nsIFeedEntry).link == null;
+
+-->
+<rss version="2.0" >
+<channel>
+<item>
+
+<enclosure length="24986239" type="audio/mpeg" url="http://dallas.example.com/joebob_050689.mp3" /> 
+<author>jbb@dallas.example.com (Joe Bob Briggs)</author>
+<comments>http://example.org</comments>
+<title>test</title>
+<guid isPermaLink="true">xorg</guid>
+<category domain="foo">bar</category>
+
+<description>I'm headed for France. I wasn't gonna go this year, but then last week &lt;a href="http://www.imdb.com/title/tt0086525/"&gt;Valley Girl&lt;/a&gt; came out and I said to myself, Joe Bob, you gotta get out of the country for a while.</description></item>
+</channel>
+</rss>
--- a/toolkit/components/feeds/test/xml/rss2/item_guid_isPermaLink.xml
+++ b/toolkit/components/feeds/test/xml/rss2/item_guid_isPermaLink.xml
@ -0,0 +1,20 @@
+<?xml version="1.0" encoding="iso-8859-1"?>
+<!--
+
+Description: item guid works
+Expect: var link = feed.items.queryElementAt(0, Components.interfaces.nsIFeedEntry).link; link.spec == 'http://www.example.org/';
+
+-->
+<rss version="2.0" >
+<channel>
+<item>
+
+<enclosure length="24986239" type="audio/mpeg" url="http://dallas.example.com/joebob_050689.mp3" /> 
+<author>jbb@dallas.example.com (Joe Bob Briggs)</author>
+<title>test</title>
+<guid isPermaLink="true">http://www.example.org/</guid>
+<category domain="foo">bar</category>
+
+<description>I'm headed for France. I wasn't gonna go this year, but then last week &lt;a href="http://www.imdb.com/title/tt0086525/"&gt;Valley Girl&lt;/a&gt; came out and I said to myself, Joe Bob, you gotta get out of the country for a while.</description></item>
+</channel>
+</rss>
--- a/toolkit/components/feeds/test/xml/rss2/item_guid_normalized.xml
+++ b/toolkit/components/feeds/test/xml/rss2/item_guid_normalized.xml
@ -0,0 +1,21 @@
+<?xml version="1.0" encoding="iso-8859-1"?>
+<!--
+
+Description: item guid works
+Expect: feed.items.queryElementAt(0, Components.interfaces.nsIFeedEntry).id == 'asdf';
+
+-->
+<rss version="2.0" >
+<channel>
+<item>
+
+<enclosure length="24986239" type="audio/mpeg" url="http://dallas.example.com/joebob_050689.mp3" /> 
+<author>jbb@dallas.example.com (Joe Bob Briggs)</author>
+<comments>http://example.org</comments>
+<title>test</title>
+<guid>asdf</guid>
+<category domain="foo">bar</category>
+
+<description>I'm headed for France. I wasn't gonna go this year, but then last week &lt;a href="http://www.imdb.com/title/tt0086525/"&gt;Valley Girl&lt;/a&gt; came out and I said to myself, Joe Bob, you gotta get out of the country for a while.</description></item>
+</channel>
+</rss>
--- a/toolkit/components/feeds/test/xml/rss2/item_guid_with_link.xml
+++ b/toolkit/components/feeds/test/xml/rss2/item_guid_with_link.xml
@ -0,0 +1,22 @@
+<?xml version="1.0" encoding="iso-8859-1"?>
+<!--
+
+Description: item prefers link to guid
+Expect: feed.items.queryElementAt(0, Components.interfaces.nsIFeedEntry).link.spec == 'http://link.example.org/';
+
+-->
+<rss version="2.0" >
+<channel>
+<item>
+
+<enclosure length="24986239" type="audio/mpeg" url="http://dallas.example.com/joebob_050689.mp3" /> 
+<author>jbb@dallas.example.com (Joe Bob Briggs)</author>
+<comments>http://example.org</comments>
+<title>test</title>
+<guid isPermaLink="true">http://www.example.org</guid>
+<link>http://link.example.org/</link>
+<category domain="foo">bar</category>
+
+<description>I'm headed for France. I wasn't gonna go this year, but then last week &lt;a href="http://www.imdb.com/title/tt0086525/"&gt;Valley Girl&lt;/a&gt; came out and I said to myself, Joe Bob, you gotta get out of the country for a while.</description></item>
+</channel>
+</rss>