elm-table

Features API

TanStack's Features API Reference lists the feature objects you register on a table, the built-in function registries, and the state types each feature owns. elm-table has no feature registry: every feature is compiled into the package and stays inert until its slice of State is non-empty or its stage is called. So each *Feature variable maps to a group of functions rather than to a value. This page maps TanStack's entries onto ours; full signatures and doc comments live on the generated Table module page and the three function modules.

Feature registry and core features

TanStack elm-table Notes
stockFeatures — Nothing to register. Every feature is always available.
coreFeatures — Same reason.
coreTablesFeature config, initialState See the Table API.
coreColumnsFeature column, group, display and the column readers See the Column API.
coreRowsFeature the row* readers and getValue See the Row API.
coreRowModelsFeature rows and the six stage functions See Row Models.
coreCellsFeature getAllCells, visibleCells See the Cell API.
coreHeadersFeature headerGroups and the header* readers See the Header API.

Cell spanning

TanStack elm-table Notes
cellSpanningFeature cellSpanIndex, cellRowSpan, cellColSpan, cellIsCovered, columnCanSpan Turned on with withCellSpanning; per column with withSpanRows, withSpanRowsWhen, withSpanColumns, withEnableCellSpanning.

There is no state slice: spans are derived from the rows you render. The index is an explicit value, CellSpanIndex, because nothing is memoized. See Cell Spanning.

Column filtering

TanStack elm-table Notes
columnFilteringFeature filteredRowModel and the column-filter queries setColumnFilter, setColumnFilters, resetColumnFilters, getCanFilter, getIsFiltered, getFilterValue, getFilterIndex.
ColumnFilter ColumnFilter { id : String, value : Value }. TanStack's value is unknown.
ColumnFiltersState State.columnFilters : List ColumnFilter Same list, same order.
FilterFn FilterFn Carries filter, resolveFilterValue, resolveDataValue, and autoRemove, like TanStack's created filter fn.
FilterFns — No registry. A filter function is passed as a value with withFilterFn, or as a row predicate with withCustomFilter.
FilterFnOption — Same reason. No filter function set means auto; getAutoFilterFn picks it and getFilterFn resolves it.
FilterMeta — No per-row filter meta. Ranking a fuzzy match is the caller's job; see Fuzzy Filtering.
BuiltInFilterFn the values of Table.FilterFn Names below.

See Column Filtering.

Column faceting

TanStack elm-table Notes
columnFacetingFeature facetedRowModel, facetedUniqueValues, facetedMinMax, globalFacetKey No state slice. Facets are computed from the row model you pass in.

See Faceting.

Row aggregation

TanStack elm-table Notes
rowAggregationFeature aggregationValue, aggregationValueOf, cellIsAggregated, rowAggregatedValues Values are computed by groupedRowModel and stored on the group row.
AggregationFns — No registry. Set one per column with withAggregationFn.
AggregationFnOption — Same reason. Unset means auto: getAutoAggregationFn picks sum for numbers and extent for dates, and getAggregationFn resolves it. TanStack's keyed list form, which produces an object of several results, has no counterpart.
AggregationResult Value One Value per column, never a keyed object.
AggregationContext — An aggregation is List Value -> Value. It gets the values, not the table.
AggregationValueContext — Same reason.
AggregationValueResult — There is no "handled" marker, because there is no column-level value override to signal it from.
AggregationMergeContext — A merge is List Value -> Value over the child results; see withMerge.
AggregationFnDef AggregationFn Opaque: a fold plus an optional merge.
AggregationFnDescriptor — Only needed to key entries of a multiple result, which has no counterpart.
constructAggregationFn custom Add a merge with withMerge.
BuiltInAggregationFn the values of Table.AggregationFn Names below.

How far down an aggregation reads is withMaxAggregationDepth on the column. See Aggregation.

Column grouping

TanStack elm-table Notes
columnGroupingFeature groupedRowModel and the grouping queries toggleGrouping, setGrouping, resetGrouping, getCanGroup, getIsGrouped, getGroupedIndex, preGroupedRowModel.
GroupingState State.grouping : List String Column ids, outermost first.
GroupingColumnMode GroupedColumnMode The three variants are groupedColumnsReorder, groupedColumnsRemove, groupedColumnsIgnore. It is not in TanStack's nav list, but it is the type behind groupedColumnMode.

See Grouping.

Column ordering

TanStack elm-table Notes
columnOrderingFeature setColumnOrder, resetColumnOrder, orderColumns, orderGroupedColumns, columnIndex, columnIsFirst, columnIsLast Ordering composes in one place, orderColumns.
ColumnOrderState State.columnOrder : List String Column ids. Unlisted columns keep definition order behind the listed ones.

See Column Ordering.

Column pinning

TanStack elm-table Notes
columnPinningFeature pinColumn, setColumnPinning, resetColumnPinning, columnCanPin, columnIsPinned, columnPinnedIndex, and the per-region column, header, and cell lists
ColumnPinningState ColumnPinning { left : List String, right : List String }. TanStack's fields are start and end.
ColumnPinningPosition ColumnPinPosition pinnedLeft, pinnedRight, columnUnpinned replace 'start' | 'end' | false.

Which slice a query is about is a ColumnRegion argument rather than an optional position, with allColumnsRegion standing for TanStack's absent argument. See Column Pinning.

Column resizing

TanStack elm-table Notes
columnResizingFeature — No drag session. A resize is a pointer handler in your code that ends in setColumnSize; the committed width lives in State.columnSizing.
columnResizingState — Nothing to store between pointer events. Keep the drag start and delta in your own model.
ColumnResizeMode — You decide whether your handler writes on every move or only on release.
ColumnResizeDirection — No text-direction handling; the sign of the delta is yours.

See Column Resizing.

Column sizing

TanStack elm-table Notes
columnSizingFeature getColumnSize, getColumnStart, getColumnAfter, getHeaderSize, getHeaderStart, totalSize, leftTotalSize, centerTotalSize, rightTotalSize, setColumnSize, setColumnSizing, resetColumnSize, resetColumnSizing
ColumnSizingState State.columnSizing : Dict String Float Column id to committed width.

Defaults are a SizeDefaults record set with withDefaultColumn, and per column with withSize, withMinSize, withMaxSize. See Column Sizing.

Column visibility

TanStack elm-table Notes
columnVisibilityFeature columnIsVisible, columnCanHide, toggleColumnVisibility, setColumnVisibility, resetColumnVisibility, toggleAllColumnsVisible, isAllColumnsVisible, isSomeColumnsVisible, visibleFlatColumns, visibleLeafColumns
ColumnVisibilityState State.columnVisibility : Dict String Bool A column absent from the dict is visible.

See Column Visibility.

Global filtering

TanStack elm-table Notes
globalFilteringFeature setGlobalFilter, resetGlobalFilter, getCanGlobalFilter, getGlobalFilterFn, globalAutoFilterFn The global filter runs inside filteredRowModel.

The state slice is State.globalFilter : Value, and the function is set with withGlobalFilterFn. TanStack has no separate global-filter state type in this nav section. See Global Filtering.

Row expanding

TanStack elm-table Notes
rowExpandingFeature expandedRowModel and the expansion queries toggleExpanded, toggleAllRowsExpanded, setExpanded, resetExpanded, getCanExpand, getIsExpanded, getIsAllParentsExpanded, getExpandedDepth, preExpandedRowModel.
ExpandedState Expanded expandAll is TanStack's true; expandedIds takes a Set String instead of a record of booleans, and expandedIdsOf reads it back.

Per-row overrides are withRowCanExpand and withIsRowExpanded on the Config. See Expanding.

Row pagination

TanStack elm-table Notes
rowPaginationFeature paginatedRowModel and the page queries setPage, setPageSize, setPagination, firstPage, previousPage, nextPage, lastPage, getPageCount, getPageOptions, getRowCount, getCanPreviousPage, getCanNextPage, getCanLastPage, resetPageIndex, resetPageSize, resetPagination.
PaginationState Pagination { pageIndex : Int, pageSize : Int }. TanStack's Infinity page size is unlimitedPageSize.

TanStack renames setPageIndex here to setPage. The counts read the pre-pagination model, prePaginationRowModel. See Pagination.

Row pinning

TanStack elm-table Notes
rowPinningFeature pinRow, pinRowWith, setRowPinning, resetRowPinning, getIsRowPinned, getRowPinnedIndex, getCanPinRow, isSomeRowsPinned, isSomeRowsPinnedTop, isSomeRowsPinnedBottom, topRows, bottomRows, centerRows
RowPinningState RowPinning { top : List String, bottom : List String }, the same shape.
RowPinningPosition RowPinPosition pinnedTop, pinnedBottom, rowUnpinned.

How much of a row's family is pinned along with it is a PinRowOptions record, with defaultPinRowOptions as the starting point. The pinned lists take a PinnedRowsSource, both the pre-pagination model and the page. See Row Pinning.

Row selection

TanStack elm-table Notes
rowSelectionFeature toggleRowSelected, toggleRowSelectedWith, toggleAllRowsSelected, toggleAllPageRowsSelected, deselectAllRows, setRowSelection, resetRowSelection, selectRange, selectRangeWith, canSelectRange, and the getIs* and getCan* queries
RowSelectionState State.rowSelection : Set String TanStack's Record<string, true> is a set of row ids here. Read it with selectedRowIds or selectedRowModel.

Sub-tree state is SubRowSelection with noSubRowsSelected, someSubRowsSelected, allSubRowsSelected, replacing TanStack's false | 'some' | 'all'. TanStack's ToggleSelectedOptions is SelectOptions, with defaultSelectOptions. The shift-click anchor is caller state, so selectRange takes the anchor row id. See Row Selection.

Row sorting

TanStack elm-table Notes
rowSortingFeature sortedRowModel and the sorting queries toggleSort, setSorting, clearSorting, resetSorting, getCanSort, getCanMultiSort, getIsSorted, getSortIndex, getNextSortingOrder.
ColumnSort SortColumn { id : String, desc : Bool }, the same shape under a different name.
SortingState State.sorting : List SortColumn Same list, first entry sorted first.
SortDirection SortDir sortAsc and sortDesc replace 'asc' and 'desc'. getIsSorted returns Maybe SortDir rather than false | 'asc' | 'desc'.
SortFn SortFn A comparison on two Values, plus an optional resolveDataValue normaliser.
SortFns — No registry. Set one per column with withSortFn, or compare whole rows with withCustomSort.
SortFnOption — Same reason. Unset means auto: getAutoSortFn picks it and getSortFn resolves it.
BuiltInSortFn the values of Table.SortFn Names below.

Where Null values land is SortUndefined, with sortNullsFirst, sortNullsLast, sortNullsAsMinusOne, sortNullsAsPlusOne. See Sorting.

Cell selection

TanStack ships a cellSelectionFeature but does not list it in this nav section. It is here for completeness.

TanStack elm-table Notes
cellSelectionFeature selectCell, toggleCellSelection, selectCellRange, selectCellRangeWith, selectAllCells, setCellSelection, clearCellSelection, setFocusedCell, moveCellSelection, extendCellSelection, extendCellSelectionTo
CellSelectionState State.cellSelection : List CellSelectionRange A list of rectangles, each built with cellRange.
CellSelectionRange CellSelectionRange Two corners plus an operation. TanStack's optional operation is required here and defaults to includeCells.
CellSelectionRangeOperation CellSelectionOperation includeCells, excludeCells.
CellSelectionRangeMode CellSelectionMode replaceSelection, includeSelection, excludeSelection.
CellSelectionDirection CellDirection cellUp, cellDown, cellLeft, cellRight.
CellSelectionBounds CellSelectionBounds Inclusive display-order indexes.
CellSelectionEdges CellSelectionEdges Which sides of a cell sit on the outer boundary.

The cell queries are on the Cell API. The reads that resolve a selection take a SelectionRows.

Cell selection reads and geometry

Nothing in TanStack's nav list corresponds to these, so they are listed here.

Value What it does
focusedCell The active cell, when there is one.
selectedCellIds The ids of every selected cell.
selectedCellCount How many cells are selected.
selectedCellRangesData The values of each selected rectangle, row by row, ready for a clipboard copy.
cellSelectionRowIds The row ids the selection touches.
cellSelectionColumnIds The column ids the selection touches.
cellSelectionColumnIndexes Column id to display-order index.
cellSelectionBounds The selection resolved into rectangles of indexes.
cellSelectionMergeBounds The same rectangles with spanning cells folded in.
intersectCellSelectionBounds The overlap of two rectangles.
subtractCellSelectionBounds One rectangle minus another, as a list.
addCellSelectionBounds Add a rectangle to a list without overlapping the others.
mergeAdjacentCellSelectionBounds Join touching rectangles.
expandCellSelectionBounds Grow a rectangle until it covers every span it clips.
applyCellSelectionBoundsOperations Fold a list of include and exclude rectangles into the final positive area.
spanAllColumns The span value that means "to the end of the region".
cellSpanIndexRowIds The row ids a span index was built over.
cellSpanIndexRowSpans The row spans of a span index, keyed by column id.

See Cell Selection.

Built-in sort functions

TanStack's sortFns object. Every name has the same name here, as a value of Table.SortFn. Nothing is missing in either direction.

TanStack sortFns elm-table
alphanumeric SortFn.alphanumeric
alphanumericCaseSensitive SortFn.alphanumericCaseSensitive
basic SortFn.basic
datetime SortFn.datetime
text SortFn.text
textCaseSensitive SortFn.textCaseSensitive

The module also exposes what TanStack keeps outside the registry: custom for constructSortFn, withResolveDataValue for overriding a built-in's normaliser, compare and resolveDataValue for applying one, and splitAlphaNumeric, which is TanStack's reSplitAlphaNumeric.

Built-in filter functions

TanStack's filterFns object. Every name has the same name here, as a value of Table.FilterFn.

TanStack filterFns elm-table
arrIncludes FilterFn.arrIncludes
arrIncludesAll FilterFn.arrIncludesAll
arrHas FilterFn.arrHas
arrIncludesSome FilterFn.arrIncludesSome
between FilterFn.between
betweenInclusive FilterFn.betweenInclusive
empty FilterFn.empty
endsWith FilterFn.endsWith
equals FilterFn.equals
equalsString FilterFn.equalsString
equalsStringSensitive FilterFn.equalsStringSensitive
inDateRange FilterFn.inDateRange
inNumberRange FilterFn.inNumberRange
includesString FilterFn.includesString
includesStringSensitive FilterFn.includesStringSensitive
notEmpty FilterFn.notEmpty
startsWith FilterFn.startsWith
weakEquals FilterFn.weakEquals

Four more filter functions are exported by TanStack but left out of its filterFns object, so they are not part of BuiltInFilterFn. All four exist here as ordinary values.

TanStack export elm-table
filterFn_greaterThan FilterFn.greaterThan
filterFn_greaterThanOrEqualTo FilterFn.greaterThanOrEqualTo
filterFn_lessThan FilterFn.lessThan
filterFn_lessThanOrEqualTo FilterFn.lessThanOrEqualTo

The rest of the module matches TanStack's constructFilterFn and the four parts every filter function carries: custom, withResolveFilterValue, withResolveDataValue, withAutoRemove, and the appliers filter, resolveFilterValue, resolveDataValue, autoRemove, plus toDateTimestamp.

There is no fuzzy filter in either package. TanStack's fuzzy example builds one with an external ranking library; see Fuzzy Filtering.

Built-in aggregation functions

TanStack's aggregationFns object. Every name has the same name here, as a value of Table.AggregationFn. Nothing is missing in either direction.

TanStack aggregationFns elm-table
sum AggregationFn.sum
min AggregationFn.min
max AggregationFn.max
extent AggregationFn.extent
mean AggregationFn.mean
median AggregationFn.median
unique AggregationFn.unique
uniqueCount AggregationFn.uniqueCount
count AggregationFn.count
first AggregationFn.first
last AggregationFn.last

Building and applying one: custom, withMerge, aggregate, merge.