Filtering output

Long output — a test run, a build log, tail -f — is easier to read when you can filter it down. Every block has its own filter bar that narrows the output to matching lines, live.

Opening the filter

Select a block and click the 🔽 icon in its header, or pick Toggle block filter from the right-click menu. The filter bar docks under the block's header. Filters are per-block: each block remembers its own filter independently, and the bar stays open until you close it.

Using it

• Query field — type and the output narrows immediately to lines containing your text. A live match count shows how many lines matched.
• ✕ — clears the query.
• **.*** — regex mode. Your query is compiled as a regular expression; matching lines are kept.
• Aa — case-sensitive matching. Both toggles highlight in the accent color when active.

While the filter has a non-empty query, matched substrings inside the surviving lines are highlighted, so you can spot the hit inside a long line at a glance.

# e.g. filter a test run down to failures
cargo test
# open the filter, type: FAILED
# or regex mode: error(\[E\d+\])?

Works on streaming output

The filter is re-evaluated every frame, so it works on running blocks too. Open a filter on a tail -f or a long-running build and new lines are filtered as they stream in — you effectively get a live grep over any command without restarting it.

Filtering vs. searching

• The block filter searches one block's output and hides non-matching lines.
• Global search (ctrl+shift+F) searches files in your project via ripgrep. See the command palette for how to reach it.

Related

• Block actions — all header icons and menu items
• Blocks — block anatomy and limits