Implement feed discovery; fixes #110

lib/Database.php

@@ -406,7 +406,7 @@ class Database {
         }
     }
 
-    public function subscriptionAdd(string $user, string $url, string $fetchUser = "", string $fetchPassword = ""): int {
+    public function subscriptionAdd(string $user, string $url, string $fetchUser = "", string $fetchPassword = "", bool $discover = true): int {
         if (!Arsse::$user->authorize($user, __FUNCTION__)) {
             throw new User\ExceptionAuthz("notAuthorized", ["action" => __FUNCTION__, "user" => $user]);
         }
@@ -417,7 +417,7 @@ class Database {
             $feedID = $this->db->prepare('INSERT INTO arsse_feeds(url,username,password) values(?,?,?)', 'str', 'str', 'str')->run($url, $fetchUser, $fetchPassword)->lastId();
             try {
                 // perform an initial update on the newly added feed
-                $this->feedUpdate($feedID, true);
+                $this->feedUpdate($feedID, true, $discover);
             } catch (\Throwable $e) {
                 // if the update fails, delete the feed we just added
                 $this->db->prepare('DELETE from arsse_feeds where id is ?', 'int')->run($feedID);
@@ -548,7 +548,7 @@ class Database {
         return array_column($feeds, 'id');
     }
     
-    public function feedUpdate($feedID, bool $throwError = false): bool {
+    public function feedUpdate($feedID, bool $throwError = false, bool $discover = false): bool {
         $tr = $this->db->begin();
         // check to make sure the feed exists
         if (!ValueInfo::id($feedID)) {
@@ -564,7 +564,7 @@ class Database {
         // here. When an exception is thrown it should update the database with the
         // error instead of failing; if other exceptions are thrown, we should simply roll back
         try {
-            $feed = new Feed((int) $feedID, $f['url'], (string) Date::transform($f['modified'], "http", "sql"), $f['etag'], $f['username'], $f['password'], $scrape);
+            $feed = new Feed((int) $feedID, $f['url'], (string) Date::transform($f['modified'], "http", "sql"), $f['etag'], $f['username'], $f['password'], $scrape, $discover);
             if (!$feed->modified) {
                 // if the feed hasn't changed, just compute the next fetch time and record it
                 $this->db->prepare("UPDATE arsse_feeds SET updated = CURRENT_TIMESTAMP, next_fetch = ? WHERE id is ?", 'datetime', 'int')->run($feed->nextFetch, $feedID);

lib/Feed.php

@@ -21,7 +21,7 @@ class Feed {
     public $newItems = [];
     public $changedItems = [];
 
-    public function __construct(int $feedID = null, string $url, string $lastModified = '', string $etag = '', string $username = '', string $password = '', bool $scrape = false) {
+    public function __construct(int $feedID = null, string $url, string $lastModified = '', string $etag = '', string $username = '', string $password = '', bool $scrape = false, bool $discover = false) {
         // set the configuration
         $userAgent = Arsse::$conf->fetchUserAgentString ?? sprintf('Arsse/%s (%s %s; %s; https://code.jkingweb.ca/jking/arsse) PicoFeed (https://github.com/fguillot/picoFeed)',
             VERSION, // Arsse version
@@ -36,7 +36,7 @@ class Feed {
         $this->config->setClientUserAgent($userAgent);
         $this->config->setGrabberUserAgent($userAgent);
         // fetch the feed
-        $this->download($url, $lastModified, $etag, $username, $password);
+        $this->download($url, $lastModified, $etag, $username, $password, $discover);
         // format the HTTP Last-Modified date returned
         $lastMod = $this->resource->getLastModified();
         if (strlen($lastMod)) {
@@ -65,10 +65,11 @@ class Feed {
         $this->nextFetch = $this->computeNextFetch();
     }
 
-    protected function download(string $url, string $lastModified = '', string $etag = '', string $username = '', string $password = ''): bool {
+    protected function download(string $url, string $lastModified, string $etag, string $username, string $password, bool $discover): bool {
+        $action = $discover ? "discover" : "download";
         try {
             $this->reader = new Reader($this->config);
-            $this->resource = $this->reader->download($url, $lastModified, $etag, $username, $password);
+            $this->resource = $this->reader->$action($url, $lastModified, $etag, $username, $password);
         } catch (PicoFeedException $e) {
             throw new Feed\Exception($url, $e);
         }
@@ -361,13 +362,13 @@ class Feed {
 
     protected function computeLastModified() {
         if (!$this->modified) {
-            return $this->lastModified;
+            return $this->lastModified; // @codeCoverageIgnore
         }
         $dates = $this->gatherDates();
         if (sizeof($dates)) {
             return Date::normalize($dates[0]);
         } else {
-            return null;
+            return null; // @codeCoverageIgnore
         }
     }
 

tests/Feed/TestFeed.php

@@ -133,6 +133,15 @@ class TestFeed extends Test\AbstractTest {
         $this->assertSame($categories, $f->data->items[5]->categories);
     }
 
+    public function testDiscoverAFeedSuccessfully() {
+        $this->assertInstanceOf(Feed::class, new Feed(null, $this->base."Discovery/Valid", "", "", "", "", false, true));
+    }
+
+    public function testDiscoverAFeedUnsuccessfully() {
+        $this->assertException("subscriptionNotFound", "Feed");
+        new Feed(null, $this->base."Discovery/Invalid", "", "", "", "", false, true);
+    }
+
     public function testParseEntityExpansionAttack() {
         $this->assertException("xmlEntity", "Feed");
         new Feed(null, $this->base."Parsing/XEEAttack");

tests/docroot/Feed/Discovery/Feed.php

@@ -0,0 +1,12 @@
+<?php return [
+    'mime'    => "application/rss+xml",
+    'content' => <<<MESSAGE_BODY
+<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
+<channel>
+    <title>Test feed</title>
+    <link>http://example.com/</link>
+    <description>Example newsfeed title</description>
+</channel>
+</rss>
+MESSAGE_BODY
+];

tests/docroot/Feed/Discovery/Invalid.php

@@ -0,0 +1,8 @@
+<?php return [
+    'mime'    => "text/html",
+    'content' => <<<MESSAGE_BODY
+<html>
+<title>Example article</title>
+</html>
+MESSAGE_BODY
+];

tests/docroot/Feed/Discovery/Valid.php

@@ -0,0 +1,9 @@
+<?php return [
+    'mime'    => "text/html",
+    'content' => <<<MESSAGE_BODY
+<html>
+<title>Example article</title>
+<link rel="alternate" type="application/rss+xml" href="http://localhost:8000/Feed/Discovery/Feed">
+</html>
+MESSAGE_BODY
+];

tests/lib/Database/SeriesSubscription.php

@@ -135,7 +135,7 @@ trait SeriesSubscription {
         Phake::when(Arsse::$db)->feedUpdate->thenReturn(true);
         $this->assertSame($subID, Arsse::$db->subscriptionAdd($this->user, $url));
         Phake::verify(Arsse::$user)->authorize($this->user, "subscriptionAdd");
-        Phake::verify(Arsse::$db)->feedUpdate($feedID, true);
+        Phake::verify(Arsse::$db)->feedUpdate($feedID, true, true);
         $state = $this->primeExpectations($this->data, [
             'arsse_feeds'         => ['id','url','username','password'],
             'arsse_subscriptions' => ['id','owner','feed'],
@@ -153,7 +153,7 @@ trait SeriesSubscription {
             Arsse::$db->subscriptionAdd($this->user, $url);
         } catch (FeedException $e) {
             Phake::verify(Arsse::$user)->authorize($this->user, "subscriptionAdd");
-            Phake::verify(Arsse::$db)->feedUpdate($feedID, true);
+            Phake::verify(Arsse::$db)->feedUpdate($feedID, true, true);
             $state = $this->primeExpectations($this->data, [
                 'arsse_feeds'         => ['id','url','username','password'],
                 'arsse_subscriptions' => ['id','owner','feed'],