Hooks for building lightweight, fast and extendable datagrids for React
- Lightweight
- Headless (100% customizable, Bring-your-own-UI)
- Client-side & Server-side pagination support
- Sorting (Multi and Stable)
- Filters
- Pivoting & Aggregation
- Fully controllable
- Extensible via hooks
- "Why I wrote React Table and the problems it has solved for Nozzle.io" by Tanner Linsley
- This documentation is for version 7.
- View the Changelog
- Previous versions:
React Table v7 is being built and maintained by me, @tannerlinsley and I am always in need of more Patreon support to keep this project afloat. If you would like to contribute to my Patreon goal for v7 and beyond, visit my Patreon and help me out!.
Install React Table as a dependency using npm
or yarn
# NPM
$ npm install react-table
# Yarn
$ yarn add react-table
To import React Table:
import {
useTable,
useGroupBy,
useFilters,
useSortBy,
useExpanded,
usePagination,
...
} from 'react-table'
React Table is a headless utility, which means out of the box, it doesn't render or supply any actual UI elements. You are in charge of utilizing the state and callbacks of the hooks provided by this library to render your own table markup. Read this article to understand why React Table is built this way.. If you don't want to, then here's a quick rundown anyway:
- Separation of Concerns - Not that superficial kind you read about all the time. The real kind. React Table as a library honestly has no business being in charge of your UI. The look, feel, and overall experience of your table is what makes your app or product great. The less React Table gets in the way of that, the better!
- Maintenance - By removing the massive (and seemingly endless) API surface area required to support every UI use-case, React Table can remain small, easy-to-use and simple to update/maintain.
- Extensibility - UI presents countless edge cases for a library simply because it's a creative medium, and one where every developer does things differently. By not dictating UI concerns, React Table empowers the developer to design and extend the UI based on their unique use-case.
At the heart of every React Table is a table instance
object. This object contains everything needed to build a table and interact with it's state. This includes, but is not limited to:
- Columns
- Materialized Data
- Sorting
- Filtering
- Grouping
- Pagination
- Expanded State
- Any functionality provided by custom plugin hooks, too!
React Table uses React Hooks both internally and externally for 100% of its configuration and lifecycle management. Naturally, this is what allows React Table to be headless and lightweight while still having a concise and simple API.
React Table is essentially a compatible collection of custom React hooks:
- The primary React Table hook
- Plugin Hooks
- Core Plugin Hooks
- Layout Plugin Hooks
useFlexLayout
useAbsoluteLayout
(coming soon!)
- Custom Plugin Hooks
- Get your custom plugin hook listed here!
useTable
is the primary hook used to build a React Table. It serves as the starting point for every option and every plugin hook that React Table supports. The options passed into useTable
are supplied to every plugin hook after it in the order they are supplied, eventually resulting a final instance
object that you can use to build your table UI and interact with the table's state.
const instance = useTable(
{
data: [...],
columns: [...],
},
useGroupBy,
useFilters,
useSortBy,
useExpanded,
usePagination
)
useTable
is called. A table instance is created.- The
instance.state
is resolved from either a custom user state or an automatically generated one. - A collection of plugin points is created at
instance.hooks
. - Each plugin is given the opportunity to add hooks to
instance.hook
. - As the
useTable
logic proceeds to run, each plugin hook type is used at a specific point in time with each individual hook function being executed the order it was registered. - The final instance object is returned from
useTable
, which the developer then uses to construct their table.
This multi-stage process is the secret sauce that allows React Table plugin hooks to work together and compose nicely, while not stepping on each others toes.
To dive deeper into plugins, see Plugins and the Plugin Guide
The order and usage of plugin hooks must follow The Laws of Hooks, just like any other custom hook. They must always be unconditionally called in the same order.
NOTE: In the event that you want to programmatically enable or disable plugin hooks, most of them provide options to disable their functionality, eg.
options.disableSorting
React Table relies on memoization to determine when state and side effects should update or be calculated. This means that every option you pass to useTable
should be memoized either via React.useMemo
(for objects) or React.useCallback
(for functions).
- Required
useTable
is the root hook for React Table. To use it, pass it with an options object with at least a columns
and rows
value, followed by any React Table compatible hooks you want to use.
The following options are supported via the main options object passed to useTable(options)
columns: Array<Column>
- Required
- Must be memoized
- The core columns configuration object for the entire table.
- Supports nested
columns
arrays via the column'scolumns
key, eg.[{ Header: 'My Group', columns: [...] }]
data: Array<any>
- Required
- Must be memoized
- The data array that you want to display on the table.
state: TableStateTuple[stateObject, stateUpdater]
- Optional
- Must be memoized table state tuple. See
useTableState
for more information. - The state/updater pair for the table instance. You would want to override this if you plan on controlling or hoisting table state into your own code.
- Defaults to using an internal
useTableState()
instance if not defined. - See Controlling and Hoisting Table State
defaultColumn: Object
- Optional
- Defaults to
{}
- The default column object for every column passed to React Table.
- Column-specific properties will override the properties in this object, eg.
{ ...defaultColumn, ...userColumn }
- This is particularly useful for adding global column properties. For instance, when using the
useFilters
plugin hook, add a defaultFilter
renderer for every column, eg.{ Filter: MyDefaultFilterComponent }
debug: Bool
- Optional
- A flag to turn on debug mode.
- Defaults to
false
The following options are supported on any column object you can pass to columns
.
accessor: String | Function
- Required
- This string/function is used to build the data model for your column.
- The data returned by an accessor should be primitive and sortable.
- If a string is passed, the column's value will be looked up on the original row via that key, eg. If your column's accessor is
firstName
then its value would be read fromrow['firstName']
. You can also specify deeply nested values with accessors likeinfo.hobbies
or evenaddress[0].street
- If a function is passed, the column's value will be looked up on the original row using this accessor function, eg. If your column's accessor is
row => row.firstName
, then its value would be determined by passing the row to this function and using the resulting value.
id: String
- Required if
accessor
is a function - This is the unique ID for the column. It is used by reference in things like sorting, grouping, filtering etc.
- If a string accessor is used, it defaults as the column ID, but can be overridden if necessary.
- Required if
columns: Array<Column>
- Optional
- A nested array of columns.
- If defined, the column will act as a header group. Columns can be recursively nested as much as needed.
show: Boolean | Function
- Optional
- Defaults to
true
- If set to
false
, the column will be hidden. - If set to a
function
, it will be called with the current table instance and can then returntrue
orfalse
. - The data model for hidden columns is still calculated including sorting, filters, and grouping.
Header: String | Function | React.Component => JSX
- Optional
- Defaults to
({ id }) => id
- Receives the table instance and column model as props
- Must either be a string or return valid JSX
- If a function/component is passed, it will be used for formatting the header value, eg. You can use a
Header
function to dynamically format the header using any table or column state.
Cell: Function | React.Component => JSX
- Optional
- Defaults to
({ value }) => value
- Receives the table instance and cell model as props
- Must return valid JSX
- This function (or component) is primarily used for formatting the column value, eg. If your column accessor returns a date object, you can use a
Cell
function to format that date to a readable format.
The following properties are available on the table instance returned from useTable
headerGroups: Array<HeaderGroup>
- An array of normalized header groups, each containing a flattened array of final column objects for that row.
- See Header Group Properties for more information
columns: Array<Column>
- A flat array of all final column objects computed from the original columns configuration option.
- See Column Properties for more information
headers[] Array<Column>
- A nested array of final column objects, similar in structure to the original columns configuration option.
- See Column Properties for more information
rows: Array<Row>
- An array of materialized row objects from the original
data
array andcolumns
passed into the table options - See Row Properties for more information
- An array of materialized row objects from the original
getTableProps: Function(?props)
- Required
- This function is used to resolve any props needed for your table wrapper.
- Custom props may be passed. NOTE: Custom props will override built-in table props, so be careful!
prepareRow: Function(Row)
- Required
- This function is responsible for lazily preparing a row for rendering. Any row that you intend to render in your table needs to be passed to this function before every render.
- Why? Since table data could potentially be very large, it can become very expensive to compute all of the necessary state for every row to be rendered regardless if it actually is rendered or not (for example if you are paginating or virtualizing the rows, you may only have a few rows visible at any given moment). This function allows only the rows you intend to display to be computed and prepped with the correct state.
The following additional properties are available on every headerGroup
object returned by the table instance.
headers: Array<Column>
- Required
- The columns in this header group.
getHeaderGroupProps: Function(?props)
- Required
- This function is used to resolve any props needed for this header group's row.
- You can use the
getHeaderGroupProps
hook to extend its functionality. - Custom props may be passed. NOTE: Custom props will override built-in table props, so be careful!
The following properties are available on every Column
object returned by the table instance.
id: String
- The resolved column ID from either the column's
accessor
or the column's hard-codedid
property
- The resolved column ID from either the column's
visible: Boolean
- The resolved visible state for the column, derived from the column's
show
property
- The resolved visible state for the column, derived from the column's
render: Function(type: String | Function | Component, ?props)
- This function is used to render content in context of a column.
- If
type
is a string, will render using thecolumn[type]
renderer. React Table ships with defaultHeader
renderers. Other renderers likeFilter
are available via hooks likeuseFilters
. - If a function or component is passed instead of a string, it will be be passed the table instance and column model as props and is expected to return any valid JSX.
getHeaderProps: Function(?props)
- Required
- This function is used to resolve any props needed for this column's header cell.
- You can use the
getHeaderProps
hook to extend its functionality. - Custom props may be passed. NOTE: Custom props will override built-in table props, so be careful!
The following additional properties are available on every row
object returned by the table instance.
cells: Array<Cell>
- An array of
Cell
objects containing properties and functions specific to the row and column it belongs to. - See Cell Properties for more information
- An array of
values: Object<columnID: any>
- A map of this row's resolved values by columnID, eg.
{ firstName: 'Tanner', lastName: 'Linsley' }
- A map of this row's resolved values by columnID, eg.
getRowProps: Function(?props)
- Required
- This function is used to resolve any props needed for this row.
- You can use the
getRowProps
hook to extend its functionality. - Custom props may be passed. NOTE: Custom props will override built-in table props, so be careful!
The following additional properties are available on every Cell
object returned in an array of cells
on every row object.
column: Column
- The corresponding column object for this cell
row: Row
- The corresponding row object for this cell
value: any
- The resolved value for this cell.
- By default, this value is displayed on the table via the default
Cell
renderer. To override the way a cell displays
getCellProps: Function(?props)
- Required
- This function is used to resolve any props needed for this cell.
- You can use the
getCellProps
hook to extend its functionality. - Custom props may be passed. NOTE: Custom props will override built-in table props, so be careful!
render: Function(type: String | Function | Component, ?props)
- This function is used to render content in context of a cell.
- If
type
is a string, will render using thecolumn[type]
renderer. React Table ships with a defaultCell
renderer. Other renderers likeAggregated
are available via hooks likeuseFilters
. - If a function or component is passed instead of a string, it will be be passed the table instance and cell model as props and is expected to return any valid JSX.
function App() {
const columns = React.useMemo(
() => [
{
Header: 'Name',
columns: [
{
Header: 'First Name',
accessor: 'firstName',
},
{
Header: 'Last Name',
accessor: 'lastName',
},
],
},
],
[]
)
const data = [
{
firstName: 'Tanner',
lastName: 'Linsley',
},
{
firstName: 'Shawn',
lastName: 'Wang',
},
{
firstName: 'Kent C.',
lastName: 'Dodds',
},
{
firstName: 'Ryan',
lastName: 'Florence',
},
]
return <MyTable columns={columns} data={data} />
}
function MyTable({ columns, data }) {
const { getTableProps, headerGroups, rows, prepareRow } = useTable({
columns,
data,
})
return (
<table {...getTableProps()}>
<thead>
{headerGroups.map(headerGroup => (
<tr {...headerGroup.getHeaderGroupProps()}>
{headerGroup.headers.map(column => (
<th {...column.getHeaderProps()}>{column.render('Header')}</th>
))}
</tr>
))}
</thead>
<tbody>
{rows.map(
(row, i) =>
prepareRow(row) || (
<tr {...row.getRowProps()}>
{row.cells.map(cell => {
return <td {...cell.getCellProps()}>{cell.render('Cell')}</td>
})}
</tr>
)
)}
</tbody>
</table>
)
}
- Plugin Hook
- Optional
useSortBy
is the hook that implements row sorting. It also support multi-sort (keyboard required).
- Multi-sort is enabled by default
- To sort the table via UI, attach the props generated from each column's
getSortByToggleProps()
, then click any of those elements. - To multi-sort the table via UI, hold
shift
while clicking on any of those same elements that have the props fromgetSortByToggleProps()
attached. - To programmatically sort (or multi-sort) any column, use the
toggleSortBy
method located on the instance or each individual column.
The following options are supported via the main options object passed to useTable(options)
state[0].sortBy: Array<Object<id: columnID, desc: Bool>>
- Must be memoized
- An array of sorting objects. If there is more than one object in the array, multi-sorting will be enabled. Each sorting object should contain an
id
key with the corresponding column ID to sort by. An optionaldesc
key may be set to true or false to indicated ascending or descending sorting for that column. This information is stored in state since the table is allowed to manipulate the filter through user interaction.
manualSorting: Bool
- Enables sorting detection functionality, but does not automatically perform row sorting. Turn this on if you wish to implement your own sorting outside of the table (eg. server-side or manual row grouping/nesting)
disableSorting: Bool
- Disables sorting for every column in the entire table.
disableMultiSort: Bool
- Disables multi-sorting for the entire table.
disableSortRemove: Bool
- If true, the un-sorted state will not be available to columns once they have been sorted.
disableMultiRemove: Bool
- If true, the un-sorted state will not be available to multi-sorted columns.
orderByFn: Function
- Must be memoizd
- Defaults to the built-in default orderBy function
- This function is responsible for composing multiple sorting functions together for multi-sorting, and also handles both the directional sorting and stable-sorting tie breaking. Rarely would you want to override this function unless you have a very advanced use-case that requires it.
sortTypes: Object<sortKey: sortType>
- Must be memoized
- Allows overriding or adding additional sort types for columns to use. If a column's sort type isn't found on this object, it will default to using the built-in sort types.
- For mor information on sort types, see Sorting
The following options are supported on any Column
object passed to the columns
options in useTable()
disableSorting: Bool
- Optional
- Defualts to
false
- If set to
true
, the sorting for this column will be disabled
sortDescFirst: Bool
- Optional
- Defaults to
false
- If set to
true
, the first sort direction for this column will be descending instead of ascending
sortInverted: Bool
- Optional
- Defaults to
false
- If set to
true
, the underlying sorting direction will be inverted, but the UI will not. - This may be useful in situations where positive and negative connotation is inverted, eg. a Golfing score where a lower score is considered more positive than a higher one.
sortType: String | Function
- If a function is passed, it must be memoized
- Defaults to
alphanumeric
- The resolved function from the this string/function will be used to sort the this column's data.
- If a
string
is passed, the function with that name located on either the customsortTypes
option or the built-in sorting types object will be used. If - If a
function
is passed, it will be used.
- If a
- For mor information on sort types, see Sorting
The following values are provided to the table instance
:
rows: Array<Row>
- An array of sorted rows.
preSortedRows: Array<Row>
- The array of rows that were originally sorted.
toggleSortBy: Function(ColumnID: String, descending: Bool, isMulti: Bool) => void
- This function can be used to programmatically toggle the sorting for any specific column
The following properties are available on every Column
object returned by the table instance.
canSort: Bool
- Denotes whether a column is sortable or not depending on if it has a valid accessor/data model or is manually disabled via an option.
toggleSortBy: Function(descending, multi) => void
- This function can be used to programmatically toggle the sorting for this column.
- This function is similar to the
instance
-leveltoggleSortBy
, however, passing a columnID is not required since it is located on aColumn
object already.
getSortByToggleProps: Function(props) => props
- Required
- This function is used to resolve any props needed for this column's UI that is responsible for toggling the sort direction when the user clicks it.
- You can use the
getSortByToggleProps
hook to extend its functionality. - Custom props may be passed. NOTE: Custom props may override built-in sortBy props, so be careful!
sorted: Boolean
- Denotes whether this column is currently being sorted
sortedIndex: Int
- If the column is currently sorted, this integer will be the index in the
sortBy
array from state that corresponds to this column. - If this column is not sorted, the index will always be
-1
- If the column is currently sorted, this integer will be the index in the
sortedDesc: Bool
- If the column is currently sorted, this denotes whether the column's sort direction is descending or not.
- If
true
, the column is sorteddescending
- If
false
, the column is sortedascending
- If
undefined
, the column is not currently being sorted.
function Table({ columns, data }) {
// Set some default sorting state
const state = useTableState({ sortBy: [{ id: 'firstName', desc: true }] })
const { getTableProps, headerGroups, rows, prepareRow } = useTable(
{
columns,
data,
},
useSortBy // Use the sortBy hook
)
return (
<table {...getTableProps()}>
<thead>
{headerGroups.map(headerGroup => (
<tr {...headerGroup.getHeaderGroupProps()}>
{headerGroup.headers.map(column => (
// Add the sorting props to control sorting. For this example
// we can add them into the header props
<th {...column.getHeaderProps(column.getSortByToggleProps())}>
{column.render('Header')}
<span>
{/* Add a sort direction indicator */}
<span>
{column.sorted ? (column.sortedDesc ? ' 🔽' : ' 🔼') : ''}
</span>
{/* Add a sort index indicator */}
<span>({column.sorted ? column.sortedIndex + 1 : ''})</span>
</span>
</th>
))}
</tr>
))}
</thead>
<tbody>
{rows.map(
(row, i) =>
prepareRow(row) || (
<tr {...row.getRowProps()}>
{row.cells.map(cell => {
return <td {...cell.getCellProps()}>{cell.render('Cell')}</td>
})}
</tr>
)
)}
</tbody>
</table>
)
}
- Plugin Hook
- Optional
useFilters
is the hook that implements row filtering.
The following options are supported via the main options object passed to useTable(options)
state[0].filters: Object<columnID: filterValue>
- Must be memoized
- An object of columnID's and their corresponding filter values. This information is stored in state since the table is allowed to manipulate the filter through user interaction.
defaultFilter: String | Function
- If a function is passed, it must be memoized
- Defaults to
text
- The function (or resolved function from the string) will be used as the default/fallback filter method for every column that has filtering enabled.
- If a
string
is passed, the function with that name located on thefilterTypes
option object will be used. - If a
function
is passed, it will be used.
- If a
- For mor information on filter types, see Filtering
manualFilters: Bool
- Enables filter detection functionality, but does not automatically perform row filtering.
- Turn this on if you wish to implement your own row filter outside of the table (eg. server-side or manual row grouping/nesting)
disableFilters: Bool
- Disables filtering for every column in the entire table.
filterTypes: Object<filterKey: filterType>
- Must be memoized
- Allows overriding or adding additional filter types for columns to use. If a column's filter type isn't found on this object, it will default to using the built-in filter types.
- For mor information on filter types, see Filtering
The following options are supported on any Column
object passed to the columns
options in useTable()
Filter: Function | React.Component => JSX
- Required
- Receives the table instance and column model as props
- Must return valid JSX
- This function (or component) is used to render this column's filter UI, eg.
disableFilters: Bool
- Optional
- If set to
true
, will disable filtering for this column
filter: String | Function
- Optional
- Defaults to
text
- The resolved function from the this string/function will be used to filter the this column's data.
- If a
string
is passed, the function with that name located on either the customfilterTypes
option or the built-in filtering types object will be used. If - If a
function
is passed, it will be used directly.
- If a
- For mor information on filter types, see Filtering
- If a function is passed, it must be memoized
The following values are provided to the table instance
:
rows: Array<Row>
- An array of filtered rows.
preFilteredRows: Array<Row>
- The array of rows used right before filtering.
- Among many other use-cases, these rows are directly useful for building option lists in filters, since the resulting filtered
rows
do not contain every possible option.
setFilter: Function(columnID, filterValue) => void
- An instance-level function used to update the filter value for a specific column.
setAllFilters: Function(filtersObject) => void
- An instance-level function used to update the values for all filters on the table, all at once.
The following properties are available on every Column
object returned by the table instance.
canFilter: Bool
- Denotes whether a column is filterable or not depending on if it has a valid accessor/data model or is manually disabled via an option.
setFilter: Function(filterValue) => void
- An column-level function used to update the filter value for this column
filterValue: any
- The current filter value for this column, resolved from the table state's
filters
object
- The current filter value for this column, resolved from the table state's
preFilteredRows: Array<row>
- The array of rows that were originally passed to this columns filter before they were filtered.
- This array of rows can be useful if building faceted filter options.
// A great library for fuzzy filtering/sorting items
import matchSorter from 'match-sorter'
const state = useTableState({ filters: { firstName: 'tanner' } })
const filterTypes = React.useMemo(() => ({
// Add a new fuzzyText filter type.
fuzzyText: (rows, id, filterValue) => {
return matchSorter(rows, filterValue, { keys: [row => row[id] })
},
// Or, override the default text filter to use
// "startWith"
text: (rows, id, filterValue) => {
return rows.filter(row => {
const rowValue = row.values[id]
return rowValue !== undefined
? String(rowValue)
.toLowerCase()
.startsWith(String(filterValue).toLowerCase())
: true
})
}
}), [matchSorter])
// Override the default column filter to be our new `fuzzyText` filter type
const defaultColumn = React.useMemo(() => ({
filter: 'fuzzyText'
}))
const { rows } = useTable(
{
// state[0].groupBy === ['firstName']
state,
manualFilters: false,
disableFilters: false,
// Pass our custom filter types
filterTypes,
defaultColumn
},
useFilters
)
- Plugin Hook
- Optional
useGroupBy
is the hook that implements row grouping and aggregation.
- Each column's
getGroupByToggleProps()
function can be used to generate the props needed to make a clickable UI element that will toggle the grouping on or off for a specific column. - Instance and column-level
toggleGroupBy
functions are also made available for programmatic grouping.
The following options are supported via the main options object passed to useTable(options)
state[0].groupBy: Array<String>
- Must be memoized
- An array of groupBy ID strings, controlling which columns are used to calculate row grouping and aggregation. This information is stored in state since the table is allowed to manipulate the groupBy through user interaction.
manualGroupBy: Bool
- Enables groupBy detection and functionality, but does not automatically perform row grouping.
- Turn this on if you wish to implement your own row grouping outside of the table (eg. server-side or manual row grouping/nesting)
disableGrouping: Bool
- Disables groupBy for the entire table.
aggregations: Object<aggregationKey: aggregationFn>
- Must be memoized
- Allows overriding or adding additional aggregation functions for use when grouping/aggregating row values. If an aggregation key isn't found on this object, it will default to using the built-in aggregation functions
groupByFn: Function
- Must be memoized
- Defaults to
defaultGroupByFn
- This function is responsible for grouping rows based on the
state.groupBy
keys provided. It's very rare you would need to customize this function.
The following options are supported on any Column
object passed to the columns
options in useTable()
Aggregated: Function | React.Component => JSX
- Optional
- Defaults to this column's
Cell
formatter - Receives the table instance and cell model as props
- Must return valid JSX
- This function (or component) formats this column's value when it is being grouped and aggregated, eg. If this column was showing the number of visits for a user to a website and it was currently being grouped to show an average of the values, the
Aggregated
function for this column could format that value to1,000 Avg. Visits
disableGrouping: Boolean
- Defaults to
true
- If
true
, this column is able to be grouped.
- Defaults to
The following values are provided to the table instance
:
rows: Array<Row>
- An array of grouped and aggregated rows.
preGroupedRows: Array<Row>
- The array of rows originally used to create the grouped rows.
toggleGroupBy: Function(columnID: String, ?set: Bool) => void
- This function can be used to programmatically set or toggle the groupBy state for a specific column.
The following properties are available on every Column
object returned by the table instance.
canGroupBy: Boolean
- If
true
, this column is able to be grouped. - This is resolved from the column having a valid accessor / data model, and not being manually disabled via other
useGroupBy
related options
- If
grouped: Boolean
- If
true
, this column is currently being grouped
- If
groupedIndex: Int
- If this column is currently being grouped, this integer is the index of this column's ID in the table state's
groupBy
array.
- If this column is currently being grouped, this integer is the index of this column's ID in the table state's
toggleGroupBy: Function(?set: Bool) => void
- This function can be used to programmatically set or toggle the groupBy state fo this column.
getGroupByToggleProps: Function(props) => props
- Required
- This function is used to resolve any props needed for this column's UI that is responsible for toggling grouping when the user clicks it.
- You can use the
getGroupByToggleProps
hook to extend its functionality. - Custom props may be passed. NOTE: Custom props may override built-in sortBy props, so be careful!
The following properties are available on every Row
object returned by the table instance.
groupByID: String
- The column ID for which this row is being grouped.
- Will be
undefined
if the row is an original row fromdata
and not a materialized one from the grouping.
groupByVal: any
- If the row is a materialized group row, this will be the grouping value that was used to create it.
values: Object
- Similar to a regular row, a materialized grouping row also has a
values
object - This object contains the aggregated values for this row's sub rows
- Similar to a regular row, a materialized grouping row also has a
subRows: Array<Row>
- If the row is a materialized group row, this property is the array of materialized subRows that were grouped inside of this row.
depth: Int
- If the row is a materialized group row, this is the grouping depth at which this row was created.
path: Array<String|Int>
- Similar to normal
Row
objects, materialized grouping rows also have a path array. The keys inside it though are not integers like nested normal rows though. Since they are not rows that can be traced back to an original data row, they are given a unique path based on theirgroupByVal
- If a row is a grouping row, it will have a path like
['Single']
or['Complicated', 'Anderson']
, whereSingle
,Complicated
, andAnderson
would all be derived from their row'sgroupByVal
.
- Similar to normal
The following additional properties are available on every Cell
object returned in an array of cells
on every row object.
grouped: Bool
- If
true
, this cell is a grouped cell, meaning it contains a grouping value and should usually display and expander.
- If
repeatedValue: Bool
- If
true
, this cell is a repeated value cell, meaning it contains a value that is already being displayed elsewhere (usually by a parent row's cell). - Most of the time, this cell is not required to be displayed and can safely be hidden during rendering
- If
aggregated: Bool
- If
true
, this cell's value has been aggregated and should probably be rendered with theAggregated
cell renderer.
- If
function Table({ columns, data }) {
const {
getTableProps,
headerGroups,
rows,
prepareRow,
state: [{ groupBy, expanded }],
} = useTable(
{
columns,
data,
},
useGroupBy,
useExpanded // useGroupBy would be pretty useless without useExpanded ;)
)
// We don't want to render all 2000 rows for this example, so cap
// it at 20 for this use case
const firstPageRows = rows.slice()
return (
<table {...getTableProps()}>
<thead>
{headerGroups.map(headerGroup => (
<tr {...headerGroup.getHeaderGroupProps()}>
{headerGroup.headers.map(column => (
<th {...column.getHeaderProps()}>
{column.canGroupBy ? (
// If the column can be grouped, let's add a toggle
<span {...column.getGroupByToggleProps()}>
{column.grouped ? '🛑' : '👊'}
</span>
) : null}
{column.render('Header')}
</th>
))}
</tr>
))}
</thead>
<tbody>
{firstPageRows.map(
(row, i) =>
prepareRow(row) || (
<tr {...row.getRowProps()}>
{row.cells.map(cell => {
return (
<td {...cell.getCellProps()}>
{cell.grouped ? (
// If it's a grouped cell, add an expander and row count
<>
<span
style={{
cursor: 'pointer',
}}
onClick={() => row.toggleExpanded()}
>
{row.isExpanded ? '👇' : '👉'}
</span>
{cell.render('Cell')} ({row.subRows.length})
</>
) : cell.aggregated ? (
// If the cell is aggregated, use the Aggregated
// renderer for cell
cell.render('Aggregated')
) : cell.repeatedValue ? null : ( // For cells with repeated values, render null
// Otherwise, just render the regular cell
cell.render('Cell')
)}
</td>
)
})}
</tr>
)
)}
</tbody>
</table>
)
}
// This is a custom aggregator that
// takes in an array of values and
// returns the rounded median
function roundedMedian(values) {
let min = values[0] || ''
let max = values[0] || ''
values.forEach(value => {
min = Math.min(min, value)
max = Math.max(max, value)
})
return Math.round((min + max) / 2)
}
function App() {
const columns = React.useMemo(
() => [
{
Header: 'Name',
columns: [
{
Header: 'First Name',
accessor: 'firstName',
// Use a two-stage aggregator here to first
// count the total rows being aggregated,
// then sum any of those counts if they are
// aggregated further
aggregate: ['sum', 'count'],
Aggregated: ({ value }) => `${value} Names`,
},
{
Header: 'Last Name',
accessor: 'lastName',
// Use another two-stage aggregator here to
// first count the UNIQUE values from the rows
// being aggregated, then sum those counts if
// they are aggregated further
aggregate: ['sum', 'uniqueCount'],
Aggregated: ({ value }) => `${value} Unique Names`,
},
],
},
{
Header: 'Info',
columns: [
{
Header: 'Age',
accessor: 'age',
// Aggregate the average age of visitors
aggregate: 'average',
Aggregated: ({ value }) => `${value} (avg)`,
},
{
Header: 'Visits',
accessor: 'visits',
// Aggregate the sum of all visits
aggregate: 'sum',
Aggregated: ({ value }) => `${value} (total)`,
},
{
Header: 'Status',
accessor: 'status',
},
{
Header: 'Profile Progress',
accessor: 'progress',
// Use our custom roundedMedian aggregator
aggregate: roundedMedian,
Aggregated: ({ value }) => `${value} (med)`,
},
],
},
],
[]
)
const data = React.useMemo(() => makeData(10000), [])
return <Table columns={columns} data={data} />
}
- Plugin Hook
- Optional
useExpanded
is the hook that implements row expanding. It is most often used with useGroupBy
to expand grouped rows, but is not limited to that use-case. It supports expanding rows both via internal table state and also via a hard-coded key on the raw row model.
The following options are supported via the main options object passed to useTable(options)
state[0].expanded: Object<[pathIndex]: Boolean | ExpandedStateObject>
- Must be memoized
- An nested object of expanded paths.
- A
pathIndex
can be set as the key and its value set totrue
to expand that row's subRows into view. For example, if{ '3': true }
was passed as theexpanded
state, the 4th row in the original data array would be expanded. - For nested expansion, you may use another object instead of a Boolean to expand sub rows. For example, if
{ '3': { '5' : true }}
was passed as theexpanded
state, then the 6th subRow of the 4th row and the 4th row of the original data array would be expanded. - This information is stored in state since the table is allowed to manipulate the filter through user interaction.
subRowsKey: String
- Required
- Defaults to
subRows
- React Table will use this key when materializing the final row object. It also uses this key to infer sub-rows from the raw data.
- See Grouping and Aggregation for more information
paginateSubRows: Bool
- Defaults to
true
- If set to
false
, expanded rows will not be paginated. Thus, any expanded subrows would potentially increase the size of any given page by the amount of total expanded subrows on the page.
- Defaults to
manualExpandedKey: String
- Defaults to
expanded
- This string is used as the key to detect manual expanded state on any given row. For example, if a raw data row like
{ name: 'Tanner Linsley', friends: [...], expanded: true}
was detected, it would be forcibly expanded, regardless of state.
- Defaults to
The following values are provided to the table instance
:
rows: Array<Row>
- An array of sorted rows.
const state = useTableState({ expanded: { '3': true, '5': { '2': true } } })
const { rows } = useTable(
{
// state[0].sortBy === { '3': true, '5': { '2': true } }
state,
},
useExpanded
)
- Plugin Hook
- Optional
usePagination
is the hook that implements row pagination. It can be used for both client-side pagination or server-side pagination. For more information on pagination, see Pagination
NOTE Some server-side pagination implementations do not use page index and instead use token based pagination! If that's the case, please use the
useTokenPagination
plugin instead.
The following options are supported via the main options object passed to useTable(options)
state[0].pageSize: Int
- Required
- Defaults to
10
- Determines the amount of rows on any given page
state[0].pageIndex: Int
- Required
- Defaults to
0
- The index of the page that should be displayed via the
page
instance value
pageCount: Int
- Required if
manualPagination
is set totrue
- If
manualPagination
istrue
, then this value used to determine the amount of pages available. This amount is then used to materialize thepageOptions
and also compute thecanNextPage
values on the table instance.
- Required if
manualPagination: Bool
- Enables pagination functionality, but does not automatically perform row pagination.
- Turn this on if you wish to implement your own pagination outside of the table (eg. server-side pagination or any other manual pagination technique)
disablePageResetOnDataChange
- Defaults to
false
- Normally, any changes detected to
rows
,state.filters
,state.groupBy
, orstate.sortBy
will trigger thepageIndex
to be reset to0
- If set to
true
, thepageIndex
will not be automatically set to0
when these dependencies change.
- Defaults to
The following values are provided to the table instance
:
pages: Array<page>
- An array of every generated
page
, each containing its respective rows.
- An array of every generated
page: Array<row>
- An array of rows for the current page, determined by the current
pageIndex
value.
- An array of rows for the current page, determined by the current
pageCount: Int
- If
manualPagination
is set tofalse
, this is the total amount of pages available in the table based on the currentpageSize
value - if
manualPagination
is set totrue
, this is merely the samepageCount
option that was passed in the table options.
- If
pageOptions: Array<Int>
- An array of zero-based index integers corresponding to available pages in the table.
- This can be useful for generating things like select interfaces for the user to select a page from a list, instead of manually paginating to the desired page.
canPreviousPage: Bool
- If there are pages and the current
pageIndex
is greater than0
, this will betrue
- If there are pages and the current
canNextPage:
- If there are pages and the current
pageIndex
is less thanpageCount
, this will betrue
- If there are pages and the current
gotoPage: Function(pageIndex)
- This function, when called with a valid
pageIndex
, will setpageIndex
to that value. - If the passed index is outside of the valid
pageIndex
range, then this function will do nothing.
- This function, when called with a valid
previousPage: Function
- This function decreases
state.pageIndex
by one. - If there are no pages or
canPreviousPage
is false, this function will do nothing.
- This function decreases
nextPage: Function
- This function increases
state.pageIndex
by one. - If there are no pages or
canNextPage
is false, this function will do nothing.
- This function increases
setPageSize: Function(pageSize)
- This function sets
state.pageSize
to the new value. - As a result of a pageSize change, a new
state.pageIndex
is also calculated. It is calculated viaMath.floor(currentTopRowIndex / newPageSize)
- This function sets
pageIndex: Int
- This is the resolved
state.pageIndex
value.
- This is the resolved
pageSize: Int
- This is the resolved
state.pageSize
value.
- This is the resolved
const state = useTableState({ pageSize: 20, pageIndex: 1 })
const { rows } = useTable(
{
// state[0] === { pageSize: 20, pageIndex: 1 }
state,
},
usePagination
)
- Plugin Hook
- Optional
useTokenPagination
is the hook that aids in implementing row pagination using tokens. It is useful for server-side pagination implementations that use tokens instead of page index. For more information on pagination, see Pagination
Documentation Coming Soon...
- Optional
useTableState
is a hook that allows you to hoist the table state out of the table into your own code. You should use this hook if you need to:
- Know about the internal table state
- React to changes to the internal table state
- Manually control or override the internal table state
Some common use cases for this hook are:
- Reacting to
pageIndex
andpageSize
changes for server-side pagination to fetch new data - Disallowing specific states via a custom state reducer
- Enabling parent/unrelated components to manipulate the table state
The following options are supported via the main options object passed to useTable(options)
initialState: Object
- Optional
- The initial state object for the table.
- This object is merged over the
defaultState
object (eg.{...defaultState, ...initialState}
) that React Table and its hooks use to register default state to produce the final initial state object passed to the resolveduseState
hook.
overrides: Object
- Optional
- Must be memoized
- This object is merged over the current table state (eg.
{...state, ...overrides}
) to produce the final state object that is then passed to theuseTable
options
options: Object
reducer: Function(oldState, newState) => finalState
- Optional
- Inspired by Kent C. Dodd's State Reducer Pattern
- With every
setState
call to a table state (even internally), this reducer is called and is allowed to modify the final state object for updating. - It is passed the
oldState
, thenewState
, and an actiontype
.
useState
- Optional
- Defaults to
React.useState
- This function, if defined will be used as the state hook internally instead of the default
React.useState
. This can be useful for implementing custom state storage hooks like useLocalStorage, etc.
tableStateTuple: [tableState, setTableState]
- Similar in structure to the result of
React.useState
- Memoized. This tuple array will not change between renders unless state or
useTableState
options change. tableState: Object
- This is the final state object of the table, which is the product of the
initialState
,overrides
and thereducer
options (if applicable)
- This is the final state object of the table, which is the product of the
setTableState: Function(updater, type) => void
- This function is used both internally by React Table, and optionally by you (the developer) to update the table state programmatically.
updater: Function
- This function signature is almost (see next point) identical to the functional API exposed by
React.setState
. It is passed the previous state and is expected to return a new version of the state. - NOTE:
updater
must be a function. Passing a replacement object is not supported as it is with React.useState
- This function signature is almost (see next point) identical to the functional API exposed by
type: String
- The action type corresponding to what action being taken against the state.
- Similar in structure to the result of
export default function MyTable({ manualPageIndex }) {
// This is the initial state for our table
const initialState = { pageSize: 10, pageIndex: 0 }
// Here, we can override the pageIndex
// regardless of the internal table state
const overrides = React.useMemo(() => ({
pageIndex: manualPageIndex,
}))
const state = useTableState(initialState, overrides)
// You can use effects to observe changes to the state
React.useEffect(
() => {
console.log('Page Size Changed!', initialState.pageSize)
},
[initialState.pageSize]
)
const { rows } = useTable({
state,
})
}
Client-side sorting can be accomplished by using the useSortBy
plugin hook. Start by importing the hook from react-table
:
-import { useTable } from 'react-table'
+import { useTable, useSortBy } from 'react-table'
Next, add the useSortBy
hook to your useTable
hook and add the necessary UI pieces we need to make sorting work:
function MyTable() {
const { getTableProps, headerGroups, rows, prepareRow } = useTable(
{
data,
columns,
},
- useSortBy
+ useSortBy
)
return (
<table {...getTableProps()}>
<thead>
{headerGroups.map(headerGroup => (
<tr {...headerGroup.getHeaderGroupProps()}>
{headerGroup.headers.map(column => (
- <th {...column.getHeaderProps()}>
+ <th {...column.getHeaderProps(column.getSortByToggleProps())}>
{column.render('Header')}
+ <span>
+ {column.sorted ? (column.sortedDesc ? ' 🔽' : ' 🔼') : ''}
+ </span>
</th>
))}
</tr>
))}
</thead>
<tbody>
{rows.map(
(row, i) =>
prepareRow(row) || (
<tr {...row.getRowProps()}>
{row.cells.map(cell => {
return <td {...cell.getCellProps()}>{cell.render('Cell')}</td>
})}
</tr>
)
)}
</tbody>
</table>
)
}
Server-side sorting can be accomplished by using the useSortBy
plugin hook in controlled mode along with the useTableState
hook. Start by importing these hooks from react-table
:
-import { useTable } from 'react-table'
+import { useTable, useSortBy, useTableState } from 'react-table'
Next, add the useSortBy
and useTableState
hooks to your useTable
hook, configure the table state, then add the necessary UI pieces we need to make sorting work:
function MyTable(data, columns, fetchData) {
+ const state = useTableState()
+ const [{ sortBy }] = state
+ React.useEffect(() => {
+ // When sorting changes, trigger your parent component
+ // or hook to fetch new data with the table state
+ fetchData(state[0])
+ }, [sortBy])
const { getTableProps, headerGroups, rows, prepareRow } = useTable(
{
data,
columns,
+ state,
+ manualSorting: true
},
- useSortBy
+ useSortBy
)
return (
<table {...getTableProps()}>
<thead>
{headerGroups.map(headerGroup => (
<tr {...headerGroup.getHeaderGroupProps()}>
{headerGroup.headers.map(column => (
- <th {...column.getHeaderProps()}>
+ <th {...column.getHeaderProps(column.getSortByToggleProps())}>
{column.render('Header')}
+ <span>
+ {column.sorted ? (column.sortedDesc ? ' 🔽' : ' 🔼') : ''}
+ </span>
</th>
))}
</tr>
))}
</thead>
<tbody>
{rows.map(
(row, i) =>
prepareRow(row) || (
<tr {...row.getRowProps()}>
{row.cells.map(cell => {
return <td {...cell.getCellProps()}>{cell.render('Cell')}</td>
})}
</tr>
)
)}
</tbody>
</table>
)
}
To suggest a feature, create an issue if it does not already exist. If you would like to help develop a suggested feature follow these steps:
- Fork this repo
- Install dependencies with
$ yarn
- Link
react-table
locally with$ yarn link
- Auto-build files as you edit with
$ yarn start
- Implement your changes and tests to files in the
src/
directory - In any example directory, link to the local
react-table
with$ yarn link react-table
- Follow example directions for running. Usually just
$ yarn && yarn start
- Document your changes in the root
README.md
- To stage a commit, run
yarn commit
- Submit PR for review