You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This might be user error rather than a bug but I can't get an exclusion filter to work.
Last Week Tonight posts videos on Youtube that are geolocked. Conveniently the titles follow the common way of labeling episodes with the season and episode number, Sxx Eyy.
I've tried various formulations of what I understand a regex field filter would be. I started with !title:/^S[0-9]+ E[0-9]+:/, which is what I expect would be the correct filter to exclude these videos. I've tried escaping various things like the + metacharacters but haven't gotten it to work. Problem with testing is that I assume the filter only applies to new entries and I don't know when the next episode will be posted, so I don't find out the filter doesn't work until another of these geolocked episodes shows up in the feed.
I assume the title field filter tries to match the string in the following span:
<spanclass="entry-title-link" aria-expanded="false" aria-current="true" role="link" tabindex="0">S12 E02: DOGE, National Parks & Content Moderation: 2/23/25: Last Week Tonight with John Oliver</span>
The text was updated successfully, but these errors were encountered:
Your filter expression looks okay to me and works for me in tests:
--- a/tests/Helpers/FilterTest.php+++ b/tests/Helpers/FilterTest.php@@ -492,6 +492,15 @@ final class FilterTest extends TestCase {
),
true,
];
++ yield 'Not(Title): Match' => [+ '!title:/^S[0-9]+ E[0-9]+:/',+ self::mkItem(+ /* title: */ 'S12 E02: DOGE, National Parks & Content Moderation: 2/23/25: Last Week Tonight with John Oliver',+ /* content: */ ''+ ),+ false,+ ];
}
/**
What selfoss version are you using? It is only available in 2.20 preview builds, not 2.19.
The only other reason it would not be working I see is if there are extra spaces at the beginning of the string.
As for testing, I have a preview feature on todo list and have it implemented as CLI script but still need to design a UI. For now, there is a (bit tiresome) option of creating new source, deleting its items in the database before each filter change and refreshing.
That must be my problem. I noticed the filter field and read these docs on it. Maybe the docs can have a version indication/selector of some sort, or at least a warning they're not for the latest stable release?
This might be user error rather than a bug but I can't get an exclusion filter to work.
Last Week Tonight posts videos on Youtube that are geolocked. Conveniently the titles follow the common way of labeling episodes with the season and episode number,
Sxx Eyy
.I've tried various formulations of what I understand a regex field filter would be. I started with
!title:/^S[0-9]+ E[0-9]+:/
, which is what I expect would be the correct filter to exclude these videos. I've tried escaping various things like the+
metacharacters but haven't gotten it to work. Problem with testing is that I assume the filter only applies to new entries and I don't know when the next episode will be posted, so I don't find out the filter doesn't work until another of these geolocked episodes shows up in the feed.I assume the
title
field filter tries to match the string in the following span:The text was updated successfully, but these errors were encountered: