diff options
author | Kevin Decherf <kevin@kdecherf.com> | 2020-12-09 22:13:09 +0100 |
---|---|---|
committer | Benjamin Brahmer <info@b-brahmer.de> | 2020-12-15 20:02:21 +0100 |
commit | 42ea24f2f41ce04588aa929e5ffdaf1dbeb1a700 (patch) | |
tree | 657089d619d1ee76d8f86dbf8553f9ce86c6867d /lib/Fetcher | |
parent | 1345cedd7e226c23a9fff6275a0fb71e444aa1b3 (diff) |
Remove LastModified-based cursor when updating feeds
We remove the call to readSince() as some feeds push new articles with
pubDate prior to the lastModified time stored for these feeds (e.g.
lemonde.fr). As we go through all items of a feed again and again, we
prevent the constant update of an item's lastModified timestamp by
keeping the previous one if its fingerprint does not change.
Fixes #921
Signed-off-by: Kevin Decherf <kevin@kdecherf.com>
Diffstat (limited to 'lib/Fetcher')
-rwxr-xr-x | lib/Fetcher/FeedFetcher.php | 16 |
1 files changed, 1 insertions, 15 deletions
diff --git a/lib/Fetcher/FeedFetcher.php b/lib/Fetcher/FeedFetcher.php index fa6d9a346..b9526165d 100755 --- a/lib/Fetcher/FeedFetcher.php +++ b/lib/Fetcher/FeedFetcher.php @@ -111,21 +111,7 @@ class FeedFetcher implements IFeedFetcher } $url = $url2->getNormalizedURL(); $this->reader->resetFilters(); - if (empty($lastModified) || !is_string($lastModified)) { - $resource = $this->reader->read($url); - } else { - $resource = $this->reader->readSince($url, new DateTime($lastModified)); - } - - $response = $resource->getResponse(); - if (!$response->isModified()) { - $this->logger->debug('Feed {url} was not modified since last fetch. old: {old}, new: {new}', [ - 'url' => $url, - 'old' => print_r($lastModified, true), - 'new' => print_r($response->getLastModified(), true), - ]); - return [null, []]; - } + $resource = $this->reader->read($url); $location = $resource->getUrl(); $parsedFeed = $resource->getFeed(); |