When to Use the Table Block

• Show a list of records users can scan, sort, filter and page through.

• Let users act on records via per-row buttons or bulk actions over many selected rows.

• Let users edit records in place, with edits batched into a single save.

• Give users saved views, grouping or a summary footer over the data.

Note: Table vs Repeatable: Use a Repeatable when every record should render as its own free-form card of blocks. Use a Table when records belong in columns and users need sorting, filtering, pagination, selection or inline editing.

Data Binding

A table's rows come from one of two Source modes, chosen under Inspector → Content → Source.

Primary Key

The Primary Key field uniquely identifies each row. It is required for row selection, inline editing, drag-reordering, and saved views to work correctly. Set it under Inspector → Content → Primary Key.

Warning: Changing the data source rebuilds columns. Selecting or changing the data source regenerates the column list from the response and replaces any custom configuration — column names, formats, widths and editability you set up are lost. Only text, number, and boolean fields from the response become columns automatically. Reconfigure columns after switching sources.

Content Properties

Appearance Properties

Note: Empty State slot vs Empty Message: A block in the Empty State slot replaces both the Empty Message and the "All records are filtered" message. If a search or filter is active with zero results, the fixed message "All records are filtered" shows instead of your custom Empty Message — only the slot overrides it.

Add-ons

Add-ons live under Inspector → Content → Add-ons. Each opens its own detail panel.

Pagination

Warning: Mapped sources need Total Records. With a Mapped source and page-based pagination, the table cannot know the total row count — bind Total Records yourself or the page navigation has no total to count against.

Events

Configure these under Inspector → Content → Interactions (On Save under the Save Action add-on).

Note: Clicking a row action also sets {{ id.selectedRow }} and {{ id.selectedRowKey }} before the action's own event runs, so the handler and anything bound to selection always sees the clicked row.

Methods

Wire any event to a Trigger Component Method action, pick the table under Block ID, the method under Method Name, and fill Method Payload when required. Methods only run while the table is visible — calling a method on a hidden table does nothing and the call is not queued.

Exposed State

Read these from anywhere on the page as {{ id.* }} where id is the table's block ID.

Warning: context bindings (currentRow, currentRowIndex) only resolve inside the table — in a row action's settings, a cell value, or a block placed in a custom cell. From a button elsewhere on the page, use {{ id.selectedRow }} instead.

Common Patterns

Behavior & Gotchas

Bulk actions lock selection to Multiple

While any bulk action is configured, the Single and None selection modes are disabled in the inspector. Remove the bulk actions first if you need Single or None.

No built-in way to add or delete rows

The table has no add-row or delete-row control and no method to insert or remove a row. With a Mapped source, change the bound data itself. With an Object source, run a create/delete call from a button or row action, then call refetchData.

The header stays put; the grid scrolls

The header row is sticky by default. The grid scrolls vertically inside a capped container while the header stays visible. Horizontal scrolling appears automatically when columns are wider than the block. Turn on a column's Grow setting so columns share leftover width instead of overflowing.

Methods do nothing on hidden tables

Calling any method on a hidden table (behind a visibility condition, on an inactive tab, inside a closed drawer) silently does nothing — the call is not queued. Trigger methods after the table is shown, not before.

User State Settings remembers layout and paging, not selection

With User State Settings on (the default), each user's page number, filters, sort, search, column order and widths are persisted. Row selection and pending edits are not saved. Turn on Reset When Hidden to clear state on hide.

Frequently Asked Questions

What is the difference between an Object source and a Mapped source?

An Object source points the table at a platform object (a stored business entity with records): the table runs the query, applies toolbar filter/search/sort server-side, and computes pagination totals for you. A Mapped source means you bind the rows yourself — from an API, a variable, another block — and you own the query, the paging inputs ({{ id.content.page.offset }} and {{ id.content.page.size }}), and the Total Records count. Use Object when your data is a platform object; use Mapped for external APIs and computed data.

When I select a row and open a drawer, why does the drawer show nothing?

Binding the drawer's fields to {{ table1.selectedRow }} fills the content but does not open the drawer. You need two steps: bind the content, and wire the table's On Select Row event to the drawer's Show method via a Trigger Component Method action. Without that method call, the drawer never opens.

Why does my mapped table show the same rows on every page?

Your data source is returning rows in an unstable order, or you are not passing the current offset to it. The table never slices data itself — it asks the source for each page by offset and size via {{ id.content.page.offset }} and {{ id.content.page.size }}. Bind those into your API call's parameters and sort the source deterministically (add a tiebreaker like an id) so page 2 today is the same as page 2 later.

Can I have more than one Table on the same page?

Yes — each Table block has its own Block ID and its own independent state. You can wire them together (bind the second table's data source to the first table's {{ table1.selectedRow.id }} as a filter parameter) to build master-detail layouts.

Why are my columns missing after I switched the data source?

Changing the data source regenerates the column list from the new response and replaces everything you configured — column names, formats, widths, editability are all reset. This is by design. Reconfigure columns after switching sources, not before.

Related Blocks