Content filtering

The Content filtering tab defines which items are imported. An item that fails any rule is skipped, and the log tells you which rule stopped it. Empty fields and 0 values switch a rule off.

When filters run

Filters are applied after the duplicate check and the full-text extraction, and before post templates, AI shortcodes, translation and spinning. So keyword rules see the complete article in its original language, and items you do not want never cost you AI or translation requests.

Apply filtering to

Which parts of the item the keyword rules check: post title, post content and post excerpt (all ticked by default), and post link.

Keyword rules

Separate keywords and phrases with commas. Matching ignores letter case.

Must contain any of these keywords: The item is imported if at least one keyword or phrase is found.
Must contain all these keywords: Every keyword or phrase must be found. The strictest rule.
Must contain none of these keywords: The item is skipped if any keyword or phrase is found. Use it to keep out unwanted topics, sponsored posts or brands.

How keywords are matched:

In the title, content and excerpt, a keyword matches as a whole word or phrase: car matches “Car prices rise” but not “scary” or “cars”. Add word forms separately when you need them: car, cars.
In the post link, a keyword matches any part of the URL, so /sports/ keeps only links from the sports section of a site.

Tag rules

These rules check the categories that come with the item in its <category> tags. Separate tags with commas; letter case is ignored, the whole tag must match.

Must contain any of these tags: The item must have at least one of these tags.
Must contain none of these tags: The item is skipped if it has one of these tags.

Items without any tags are not checked by these two rules.

Date rules

Must be newer than [...] day(s): Skip items published more than N days ago, according to the date in the feed. Keeps old news out when you add a feed with a long history.
Must be older than [...] day(s): Import an item only when it is at least N days old. Useful to give the original publisher a head start, or to work through archives.

Length rules

Must be longer than [...] character(s): Skip items whose content, without HTML tags, is shorter than N characters. Filters out teasers and empty items.
Must be shorter than [...] character(s): Skip items whose content is longer than N characters.

Combine the length rule with full-text extraction to make sure only real articles get through.