Practical guide
How to filter a web table in Chrome
A webpage table can be easy to scan at ten rows and painful at a hundred. For a standard loaded HTML table, ListSift can apply literal text or numeric rules without exporting the rows.
1. Check that the table is supported
The rows must already be loaded in the page and represented by ordinary HTML elements. Tables with merged cells, infinite loading, virtualized rows, embedded cross-origin frames, canvas rendering, or PDF content need a different workflow. ListSift processes at most the first 200 detected loaded rows.
2. Choose a row
Open ListSift, select Choose items, then click a cell in one row. The extension shows the detected row count and available columns. Check that the boundary includes only the table you intended to filter.
3. Add a local rule
Use Filter by words for a literal case-insensitive phrase. Open Numbers & sorting for at most, at least, equals, or integer divisibility. Numeric parsing supports unambiguous English-style numbers and common currency prefixes. Ranges, compact values such as 1.2k, localized decimals, dates, and missing values remain unknown.
4. Filter and inspect
Select Filter page. Rows that do not match are hidden; rows with unknown values stay visible. The results list shows Match, Unsure, and Doesn't match counts. Sorting affects this extension results list only and does not reorder the webpage table.
5. Restore the table
Select Undo to restore hidden rows and their original inline display styles. Refreshing or navigating also clears the selection. If the table changes while filtered, ListSift restores hidden rows and asks you to select the collection again.
Use local word and number rules when they can express the task. They stay on your device and do not consume AI checks.
When an AI rule helps
AI can judge a narrow description against the visible row text when literal words are insufficient. Inspect the exact extracted text first. AI is optional, requires explicit consent, and can be wrong. Unsure items remain visible.