diff --git a/libraries/common_syndication_parser.inc b/libraries/common_syndication_parser.inc
index 47b7850004f9761ccd98252babc92639b4337795..d2296eb34c44c08f3416c63ee689290aab79894d 100644
--- a/libraries/common_syndication_parser.inc
+++ b/libraries/common_syndication_parser.inc
@@ -86,11 +86,7 @@ function _parser_common_syndication_atom10_parse($feed_XML) {
"georss" => "http://www.georss.org/georss",
);
- $base = $feed_XML->xpath("@base");
- $base = (string) array_shift($base);
- if (!valid_url($base, TRUE)) {
- $base = FALSE;
- }
+ $base = _parser_common_syndication_atom10_parse_base_url($feed_XML);
// Detect the title
$parsed_source['title'] = isset($feed_XML->title) ? _parser_common_syndication_title("{$feed_XML->title}") : "";
@@ -98,20 +94,13 @@ function _parser_common_syndication_atom10_parse($feed_XML) {
$parsed_source['description'] = isset($feed_XML->subtitle) ? "{$feed_XML->subtitle}" : "";
$parsed_source['link'] = _parser_common_syndication_link($feed_XML->link);
- if (valid_url($parsed_source['link']) && !valid_url($parsed_source['link'], TRUE) && !empty($base)) {
+ if ($base && !valid_url($parsed_source['link'], TRUE) && valid_url($parsed_source['link'])) {
$parsed_source['link'] = $base . $parsed_source['link'];
}
$parsed_source['items'] = array();
foreach ($feed_XML->entry as $news) {
-
- $original_url = NULL;
- $guid = !empty($news->id) ? "{$news->id}" : NULL;
- if (valid_url($guid, TRUE)) {
- $original_url = $guid;
- }
-
$georss = (array)$news->children($ns["georss"]);
$geoname = '';
if (isset($georss['featureName'])) {
@@ -164,13 +153,6 @@ function _parser_common_syndication_atom10_parse($feed_XML) {
$body .= "{$news->summary}";
}
- if (!empty($news->content['src'])) {
- // some src elements in some valid atom feeds contained no urls at all
- if (valid_url("{$news->content['src']}", TRUE)) {
- $original_url = "{$news->content['src']}";
- }
- }
-
$original_author = '';
if (!empty($news->source->author->name)) {
$original_author = "{$news->source->author->name}";
@@ -182,8 +164,6 @@ function _parser_common_syndication_atom10_parse($feed_XML) {
$original_author = "{$feed_XML->author->name}";
}
- $original_url = _parser_common_syndication_link($news->link);
-
$item = array();
$item['title'] = _parser_common_syndication_title($title, $body);
$item['description'] = $body;
@@ -201,17 +181,32 @@ function _parser_common_syndication_atom10_parse($feed_XML) {
$item['timestamp'] = _parser_common_syndication_parse_date("{$news->updated}");
}
- $item['url'] = trim($original_url);
- if (valid_url($item['url']) && !valid_url($item['url'], TRUE) && !empty($base)) {
- $item['url'] = $base . $item['url'];
+ $item['guid'] = (string) $news->id;
+
+ $item['url'] = _parser_common_syndication_link($news->link);
+
+ if (!$item['url'] && !empty($news->content['src']) && valid_url($news->content['src'], TRUE)) {
+ $item['url'] = (string) $news->content['src'];
}
- // Fall back on URL if GUID is empty.
- if (!empty($guid)) {
- $item['guid'] = $guid;
+
+ if (!strlen($item['url']) && $item['guid'] && valid_url($item['guid'], TRUE)) {
+ $item['url'] = $item['guid'];
}
- else {
+
+ if (!valid_url($item['url'], TRUE) && valid_url($item['url'])) {
+ if ($item_base = _parser_common_syndication_atom10_parse_base_url($news)) {
+ $item['url'] = $item_base . $item['url'];
+ }
+ elseif ($base) {
+ $item['url'] = $base . $item['url'];
+ }
+ }
+
+ // Fall back on URL if GUID is empty.
+ if (!strlen($item['guid'])) {
$item['guid'] = $item['url'];
}
+
$item['geolocations'] = array();
if ($lat && $lon) {
$item['geolocations'] = array(
@@ -226,9 +221,61 @@ function _parser_common_syndication_atom10_parse($feed_XML) {
$item['domains'] = isset($additional_taxonomies['ATOM Domains']) ? $additional_taxonomies['ATOM Domains'] : array();
$parsed_source['items'][] = $item;
}
+
return $parsed_source;
}
+/**
+ * Finds the base URL of an Atom document.
+ *
+ * @param SimpleXMLElement $xml
+ * The XML document.
+ *
+ * @return string|false
+ * Returns the base URL or false on failure.
+ */
+function _parser_common_syndication_atom10_parse_base_url(SimpleXMLElement $xml) {
+ $base = $xml->attributes('xml', TRUE)->base;
+ if (!$base) {
+ $base = $xml['base'];
+ }
+
+ if ($base && valid_url($base, TRUE)) {
+ return rtrim($base, '/') . '/';
+ }
+
+ // Try to build a base from the self link.
+ foreach ($xml->xpath('*[local-name() = "link" and @rel="self" and @href]') as $self) {
+ if (valid_url($self['href'], TRUE)) {
+ return _parser_common_syndication_string_url_path((string) $self['href']);
+ }
+ }
+
+ // Try to build a base from the alternate link.
+ foreach ($xml->xpath('*[local-name() = "link" and @rel="alternate" and @href]') as $alternate) {
+ if (valid_url($alternate['href'], TRUE)) {
+ return _parser_common_syndication_string_url_path((string) $alternate['href']);
+ }
+ }
+
+ return FALSE;
+}
+
+/**
+ * Removes the path parts of an absolute URL.
+ *
+ * @param string $url
+ * The absolute URL.
+ *
+ * @return string
+ * The absolute URL with the path stripped.
+ */
+function _parser_common_syndication_string_url_path($url) {
+ $pos = strpos($url, '/', strpos($url, '//') + 2);
+
+ return $pos ? substr($url, 0, $pos + 1) : $url . '/';
+}
+
/**
* Parse RDF Site Summary (RSS) 1.0 feeds in RDF/XML format.
*
@@ -586,7 +633,7 @@ function _parser_common_syndication_link($links) {
}
}
}
- return $to_link;
+ return trim($to_link);
}
/**
diff --git a/tests/common_syndication_parser.test b/tests/common_syndication_parser.test
index d8eabb4cad5237f518a6004aed76e36a50975b0e..d68ca4b522f97a9f5f89e69b57778cb398051a81 100644
--- a/tests/common_syndication_parser.test
+++ b/tests/common_syndication_parser.test
@@ -33,6 +33,7 @@ class CommonSyndicationParserTestCase extends DrupalWebTestCase {
$this->_testRSS2();
$this->_testAtomGeoRSS();
$this->_testAtomGeoRSSWithoutAuthor();
+ $this->_testAtomEntriesWithoutBaseUrl();
}
/**
@@ -91,6 +92,32 @@ class CommonSyndicationParserTestCase extends DrupalWebTestCase {
$feed = common_syndication_parser_parse($string);
}
+ /**
+ * Tests if the base url is prepended for entries without base url.
+ *
+ * For example, the url in the following entry should be parsed as
+ * 'http://www.example.com/node/123' and not as 'node/123'.
+ * @code
+ * <entry>
+ * <link href="node/123"/>
+ * </entry>
+ * @endcode
+ */
+ protected function _testAtomEntriesWithoutBaseUrl() {
+ $string = $this->readFeed('entries-without-base-url.atom');
+ $feed = common_syndication_parser_parse($string);
+
+ // Assert that all items got the base url assigned.
+ $expected = array(
+ 'http://www.example.com/node/1281496#comment-11669575',
+ 'http://www.example.com/node/1281496#comment-10080648',
+ 'http://www.example.com/node/1281496#comment-10062564',
+ );
+ foreach ($feed['items'] as $key => $item) {
+ $this->assertEqual($expected[$key], $item['url']);
+ }
+ }
+
/**
* Helper to read a feed.
*/
diff --git a/tests/feeds/entries-without-base-url.atom b/tests/feeds/entries-without-base-url.atom
new file mode 100644
index 0000000000000000000000000000000000000000..043c78b6a8dc15e27dee1f7d484fe9d36eb9bf1d
--- /dev/null
+++ b/tests/feeds/entries-without-base-url.atom
@@ -0,0 +1,45 @@
+<?xml version="1.0" encoding="utf-8" ?>
+<feed xmlns="http://www.w3.org/2005/Atom" xml:lang="en">
+ <title>Feeds issue #1281496</title>
+ <link rel="self" type="application/atom+xml" href="http://www.example.com/feeds/entries-without-base-url.atom"/>
+ <link rel="alternate" type="text/html" href="http://www.example.com/node/1281496"/>
+ <updated>2016-10-29T20:35:56Z</updated>
+ <author>
+ <name>dcotruta</name>
+ </author>
+ <id>http://www.example.com/node/1281496</id>
+
+ <entry>
+ <title>Re-spin the patch</title>
+ <link rel="alternate" type="text/html" href="node/1281496#comment-11669575" />
+ <id>comment-11669575</id>
+ <updated>2016-09-28T17:08:00Z</updated>
+ <summary>Re-spin the patch for feeds 7.x-2.0-beta2.</summary>
+ <author>
+ <name>natew</name>
+ </author>
+ </entry>
+
+ <entry>
+ <title>Thanks twistor, I just tried the latest patch</title>
+ <link rel="alternate" type="text/html" href="node/1281496#comment-10080648" />
+ <id>comment-10080648</id>
+ <updated>2015-07-02T19:33:00Z</updated>
+ <summary>Thanks twistor, I just tried the latest patch and this works for me. The feed items get imported and the proper url is set.</summary>
+ <author>
+ <name>natew</name>
+ </author>
+ </entry>
+
+ <entry>
+ <title>Probably missed a string cast somewhere.</title>
+ <link rel="alternate" type="text/html" href="node/1281496#comment-10062564" />
+ <id>comment-10062564</id>
+ <updated>2015-06-26T19:52:00Z</updated>
+ <summary>Probably missed a string cast somewhere.</summary>
+ <author>
+ <name>twistor</name>
+ </author>
+ </entry>
+
+</feed>