FLINT TABLE FILTER · HOW-TO
Choose the right source table, headers and merged-cell behavior
The macro filters one native table at a time. When its body holds a legend and a roster, a first row without header formatting, or merged cells, these settings decide what readers can filter.
What you will end with: the second table in the macro selected as the source, its header detected automatically, and its merged Week cells expanded into 6 rows, so Primary Is one of Ana returns 2 rows, both labeled Week 45.
The example
An on-call page holds a legend table, a short paragraph and a roster inside one Flint Table Filter macro. The roster merges each week’s cell down its days and spans a holiday across three columns.
| Code | Meaning |
|---|---|
| P | Primary on-call |
| B | Backup on-call |
| H | Office holiday |
| Week | Day | Primary | Backup |
|---|---|---|---|
| Week 45 | Mon | Ana | Ben |
| Tue | Ana | Chen | |
| Wed | Dev | Ben | |
| Week 46 | Mon | Chen | Ana |
| Tue | Dev | Ana | |
| Week 47 | Office holiday | ||
Choose the source table
- Edit the page and open the macro settings. Settings section starts at Source and starting view.
- In Source table (position inside this macro, starting at 1), enter
2. Blank or 1 selects the first table. - Starting view preview switches to the roster and warns that merged cells repeat covered values. Choose Save settings and publish.
Positions count the native tables placed directly in the macro body, in order. Tables inside other elements, such as an expand, are not counted and cannot be selected. Readers see Controls apply to native table 2. Choose another source table in the macro settings; other tables keep their original content. The legend stays on the page unchanged.
If the position does not exist, readers see Selected table is unavailable with the complete source below it. Correct the number in the macro settings.
Choose a header mode
Header row offers Automatic (Confluence header cells), First row is header and No header row. Automatic uses the first row only when it is formatted as Confluence header cells.
| Header row setting | First row uses header cells | First row uses ordinary cells |
|---|---|---|
| Automatic (Confluence header cells) | Columns Week, Day, Primary, Backup; 2 data rows | Columns Column 1 to Column 4; 3 data rows |
| First row is header | Columns Week, Day, Primary, Backup; 2 data rows | Columns Week, Day, Primary, Backup; 2 data rows |
| No header row | Columns Column 1 to Column 4; 3 data rows | Columns Column 1 to Column 4; 3 data rows |
A notice under the table states the mode in use: The first row is used as the header. Change Header row in the macro settings to use another mode., No header row selected. All rows are included; filters use numbered column names. or No header row detected. All rows are included; filters use numbered column names. Settings that name headers, such as a default sort or default filter, need a header row; without one, readers see a Check the table's default view warning.
How merged cells become rows
For filtering, sorting, totals and downloads, each merged cell is repeated in every row and column it covers. The roster becomes six logical rows:
| Week | Day | Primary | Backup |
|---|---|---|---|
| Week 45 | Mon | Ana | Ben |
| Week 45 | Tue | Ana | Chen |
| Week 45 | Wed | Dev | Ben |
| Week 46 | Mon | Chen | Ana |
| Week 46 | Tue | Dev | Ana |
| Week 47 | Office holiday | Office holiday | Office holiday |
Now each row answers questions on its own:
- Primary Is one of Ana returns 2 rows, Week 45 Mon and Week 45 Tue. The week label comes with each row, even though only the first row holds it in the page.
- Searching
week 45returns 3 rows. - Grouping by Week counts Week 45: 3, Week 46: 2, Week 47: 1.
- The holiday cell spans Day, Primary and Backup, so Primary’s values include Office holiday.
Readers see Merged cells repeat their rich content in each covered row and column for filtering, sorting, totals and CSV/XLSX. Totals count repeated values. The saved source and Confluence PDF/Word exports retain the original merges. If a merged cell holds a number, it is counted once for every row it covers; troubleshooting shows an example.
When merged cells cannot be expanded
Overlapping or invalid spans, or a table too large to expand safely (more than 10,000 rows, more than 1,000 columns or more than 100,000 cells after expansion), show Filtering is unavailable with the reason, for example The merged table has invalid spans or exceeds the safe logical-grid size. The original table is shown complete. Unmerge the cells or split the table to restore filtering.
Sources this release does not filter
Flint Table Filter works with native Confluence tables placed directly inside the macro. Tables produced by other macros or apps, embedded spreadsheets, page properties reports and excerpts are outside this release, as are pivots, charts, formulas, joins, imports and editing.