Filters

A feed's filters decide which of its entries you see. They run in the engine the moment an entry arrives, so a match never reaches the unread list.

Keep and block

Open a feed, then Filters in the top bar's menu. There are two lists:

• Block rules hide every entry that matches any of them.
• Keep rules hide every entry that matches none of them. With a keep rule, only what matches comes through.

A rule is a field — Title, Content, Link or Author — and a pattern, as contains (a phrase, case does not matter) or regex (a regular expression). Block rules are checked first; an entry that survives them is then checked against the keep rules.

Ignore posts like this… in an entry's menu is the quick way to a block rule: it opens the dialog prefilled from that entry's title.

What happens to a hidden entry

A hidden entry is stored and marked read with a note that a rule did it. That note matters when the rules change:

• a new or changed rule is re-applied to the entries already stored, so you need not wait for the next fetch;
• a rule that is removed gives its hidden entries back as unread, because a rule hid them, not a person;
• an entry you read yourself is never touched by a rule change.

Entries from before you subscribed are left alone.

The grammar

Under the dialog the rules are plain text in Miniflux's grammar, one per line, so rules written for Miniflux work as they are:

EntryTitle=(?i)sponsored
EntryURL=/tag/podcast/
EntryAuthor=^Newsdesk$
EntryContent=giveaway

The pattern is a regular expression; (?i) makes it case-insensitive. A pattern the engine cannot read — some constructs of Go's regular expressions have no JavaScript equivalent — is reported as a warning when you save and skipped, never applied half-way.

Duplicate entries

The same dialog has Duplicate entries: By URL hides an entry whose link was seen before in this feed, By URL + title requires both to repeat. For feeds that re-post the same story under a new id.