Merge pull request #137 from alexdebril/issue/134

Issue/134

README.md

@@ -197,6 +197,18 @@ $feedIo = new FeedIo($client, $logger);
 
 ```
 
+## Dealing with missing timezones
+
+Sometimes you have to consume feeds in which the timezone is missing from the dates. In some use-cases, you may need to specify the feed's timezone to get an accurate value, so feed-io offers a workaround for that : 
+
+```php
+$feedIo->getDateTimeBuilder()->setFeedTimezone(new \DateTimeZone($feedTimezone));
+$result = $feedIo->read($feedUrl);
+$feedIo->getDateTimeBuilder()->resetFeedTimezone();
+``` 
+
+Don't forget to reset `feedTimezone` after fetching the result, or you'll end up with all feeds located in the same timezone.
+
 ## Online documentation and API reference
 
 The whole documentation and API reference is available at https://feed-io.net/documentation.html

examples/force-timezone.php

@@ -0,0 +1,44 @@
+<?php
+/*
+ * This file is part of the feed-io package.
+ *
+ * (c) Alexandre Debril <alex.debril@gmail.com>
+ *
+ * For the full copyright and license information, please view the LICENSE
+ * file that was distributed with this source code.
+ */
+
+require __DIR__.DIRECTORY_SEPARATOR.'bootstrap.php';
+
+$feedIo = \FeedIo\Factory::create()->getFeedIo();
+
+$feedIo->getDateTimeBuilder()->setFeedTimezone(new \DateTimeZone('-0500'));
+$result = $feedIo->read('http://news.php.net/group.php?group=php.announce&format=rss');
+
+echo "First item's pubDate raw value in the feed"  .PHP_EOL;
+
+$domItems = $result->getDocument()->getDOMDocument()->getElementsByTagName('item');
+
+/** @var \DOMElement $firstDomItem */
+$firstDomItem = $domItems->item(0);
+$pubDate = $firstDomItem->getElementsByTagName('pubDate')->item(0);
+
+var_dump($pubDate->nodeValue);
+
+$pubDateTime = new \DateTime($pubDate->nodeValue);
+
+echo "here is its timestamp : {$pubDateTime->getTimestamp()}" . PHP_EOL;
+
+$feed = $result->getFeed();
+
+/** @var \FeedIo\Feed\ItemInterface $item */
+$feed->rewind();
+$item = $feed->current();
+
+echo "var_dump the first item's pubDate after parsing. It's the same date converted in your local configuration's timezone" . PHP_EOL;
+var_dump($item->getLastModified());
+echo "here is its timestamp : {$item->getLastModified()->getTimestamp()}" . PHP_EOL;
+
+if($pubDateTime->getTimestamp() === $item->getLastModified()->getTimestamp()) {
+    echo "HOURAY, both timestamps match !" . PHP_EOL;
+}

examples/missing-timezone.php

@@ -0,0 +1,39 @@
+<?php
+/*
+ * This file is part of the feed-io package.
+ *
+ * (c) Alexandre Debril <alex.debril@gmail.com>
+ *
+ * For the full copyright and license information, please view the LICENSE
+ * file that was distributed with this source code.
+ */
+
+require __DIR__.DIRECTORY_SEPARATOR.'bootstrap.php';
+
+echo 'see https://github.com/alexdebril/feed-io/issues/134 for full explanations' . PHP_EOL;
+
+$dateTimeBuilder = new \FeedIo\Rule\DateTimeBuilder();
+$pubDate = 'Wed, 02 Aug 2017 07:21:29';
+
+echo 'the publish date\'s timezone is America/chicago but $dateTimeBuilder ignores it. The date below is wrong'  . PHP_EOL;
+
+$dateTime = $dateTimeBuilder->convertToDateTime($pubDate);
+
+var_dump($dateTime);
+echo "timestamp : {$dateTime->getTimestamp()}" . PHP_EOL;
+
+echo 'after setting the feed\'s timezone in the builder\'s attributes, the date is right' . PHP_EOL;
+
+$dateTimeBuilder->setFeedTimezone(new \DateTimeZone('America/Chicago'));
+
+$dateTime = $dateTimeBuilder->convertToDateTime($pubDate);
+
+var_dump($dateTime);
+echo "timestamp : {$dateTime->getTimestamp()}" . PHP_EOL;
+
+echo 'if I really want the date in its original timezone, I can set it' . PHP_EOL;
+
+$dateTime->setTimezone(new \DateTimeZone('America/Chicago'));
+
+var_dump($dateTime);
+echo "timestamp : {$dateTime->getTimestamp()}" . PHP_EOL;

src/FeedIo/Rule/DateTimeBuilder.php

@@ -38,7 +38,12 @@ class DateTimeBuilder
     /**
      * @var \DateTimeZone
      */
-    protected $timezone;
+    protected $feedTimezone;
+
+    /**
+     * @var \DateTimeZone
+     */
+    protected $serverTimezone;
 
     /**
      * @var LoggerInterface
@@ -121,7 +126,7 @@ public function convertToDateTime($string)
     {
         $string = trim($string);
         foreach ([$this->getLastGuessedFormat(), $this->guessDateFormat($string) ] as $format) {
-            $date = \DateTime::createFromFormat($format, $string);
+            $date = $this->newDate($format, $string);
             if ($date instanceof \DateTime) {
                 $date->setTimezone($this->getTimezone());
 
@@ -145,25 +150,82 @@ public function stringToDateTime($string)
         if (false === strtotime($string)) {
             throw new \InvalidArgumentException('Impossible to convert date : '.$string);
         }
-        $date = new \DateTime($string);
+        $date = new \DateTime($string, $this->getFeedTimezone());
         $date->setTimezone($this->getTimezone());
 
         return $date;
     }
 
+    /**
+     * @return \DateTimeZone
+     */
+    public function getFeedTimezone()
+    {
+        return $this->feedTimezone;
+    }
+
+    /**
+     * Specifies the feed's timezone. Do this it the timezone is missing
+     *
+     * @param \DateTimeZone $timezone
+     */
+    public function setFeedTimezone(\DateTimeZone $timezone)
+    {
+        $this->feedTimezone = $timezone;
+    }
+
+    /**
+     * Resets feedTimezone to null.
+     */
+    public function resetFeedTimezone()
+    {
+        $this->feedTimezone = null;
+    }
+
+    /**
+     * @return \DateTimeZone
+     */
+    public function getServerTimezone()
+    {
+        return $this->serverTimezone;
+    }
+
+    /**
+     * @param \DateTimeZone $timezone
+     */
+    public function setServerTimezone(\DateTimeZone $timezone)
+    {
+        $this->serverTimezone = $timezone;
+    }
+
     /**
      * @return \DateTimeZone
      */
     public function getTimezone()
     {
-        return $this->timezone;
+        return $this->getServerTimezone();
     }
 
     /**
      * @param \DateTimeZone $timezone
      */
     public function setTimezone(\DateTimeZone $timezone)
     {
-        $this->timezone = $timezone;
+        $this->setServerTimezone($timezone);
+    }
+
+
+    /**
+     * @param $format
+     * @param $string
+     * @return \DateTime
+     */
+    protected function newDate($format, $string)
+    {
+        if (!! $this->getFeedTimezone()) {
+            return \DateTime::createFromFormat($format, $string, $this->getFeedTimezone());
+        }
+
+        return \DateTime::createFromFormat($format, $string);
     }
 }