- Filesize: +- 9.1KB.
- Lines: +- 300, not much code, so hopefully easy to understand.
- No dependencies on other libraries like jQuery.
- Sorting on columns, multi-column support with shift-click.
- Filtering values: case-insensitively, tokenized (separated by space).
- Able to add custom filtering, parsing and sorting functions.
- Helper function for delayed (150ms) filtering, so filtering feels more responsive for big datasets.
- Permissive ISC license, see LICENSE file.
- "Lazy scroll" mode:
- fixed column headers and renders only visible rows, this allows you to "lazily" render millions of rows.
- Officially supported browsers are:
- Firefox and Firefox ESR.
- Chrome and most recent webkit-based browsers.
Why? and a comparison
It was created because all the other datatable scripts suck balls.
jQuery adds about 97KB and Bootstrap adds about 100KB to your scripts and CSS as a dependency. This increases the CPU, memory and bandwidth consumption and latency. It also adds complexity to your scripts.
jQuery was mostly used for backwards-compatibility in the Internet Explorer days, but is most often not needed anymore. It contains functionality to query the DOM using CSS-like selectors, but this is now supported with for example document.querySelectorAll. Functionality like a JSON parser is standard available now: JSON.parse().
All sizes are not "minified" or gzipped.
|datatables.net (without plugins)||563.4KB||449.3KB||16KB||0.8KB||97.3KB|
Of course jsdatatable has less features (less is more!), but it does 90% of what's needed. Because it is so small it is also much simpler understand and extend with required features if needed.
See also: The website obesity crisis
git clone git://git.codemadness.org/jscancer
It is in the datatable directory.
Releases are available at: https://codemadness.org/releases/jscancer/.
See example.html for an example. A stylesheet file datatable.css is also included, it contains the icons as embedded images.
A table should have the classname "datatable" set, it must contain a <thead> for the column headers (<td> or <th>) and <tbody> element for the data. The minimal code needed for a working datatable:
The following column attributes are supported:
|data-filterable||If "1" or "true" specifies if the column can be filtered, default: "true".|
|data-parse||Specifies how to parse the values, default: "string", which is datatable_parse_string(). See PARSING section below.|
|data-sort||Specifies how to sort the values: default: "default", which is datatable_sort_default(). See SORTING section below.|
|data-sortable||If "1" or "true" specifies if the column can be sorted, default: "true".|
By default only parsing for the types: date, float, int and string are supported, but other types can be easily added as a function with the name: datatable_parse_<typename>(). The parse functions parse the data-value attribute when set or else the cell content (in order). Because of this behaviour you can set the actual values as the data-value attribute and use the cell content for display. This is useful to display and properly sort locale-aware currency, datetimes etc.
Filtering will be done case-insensitively on the cell content and when set also on the data-value attribute. The filter string is split up as tokens separated by space. Each token must match at least once per row to display it.
Sorting is done on the parsed values by default with the function: datatable_sort_default(). To change this you can set a customname string on the data-sort attribute on the column which translates to the function: datatable_sort_<customname>().
In some applications locale values are used, like for currency, decimal numbers datetimes. Some people also like to use icons or extended HTML elements inside the cell. Because jsdatatable sorts on the parsed value (see section PARSING) it is possible to sort on the data-value attribute values and use the cell content for display.
currency, decimal numbers: use data-value attribute with floating-point number, set data-parse column to "float". date/datetimes: use data-value attribute with UNIX timestamps (type int), set data-parse on column to "int" or set the data-parse attribute on column to "date" which is datatable_parse_date(), then make sure to use Zulu times, like: "2016-01-01T01:02:03Z" or other time strings that are parsable as the data-value attribute. icons: generally use data-value attribute with integer as weight value to sort on, set data-parse column to "int".
Dynamically update data
To update data dynamically see example-ajax.html for an example how to do this.
- A date, integer, float or other values must be able to parse properly, when the parse function returns NaN, null or undefined etc. the sorting behaviour is also undefined. It is recommended to always set a zero value for each type.
- <tfoot> is not supported in datatables in "lazy" mode.