- Fix - Include Redux 5 to peer dep. #9
- Refactor - Move Lodash as a normal dependency. #5
- Performance - Improve performance (90% or better). #7
- Performance - Improve performance (-36%) on large datasets. #6
- Feature - Remove
resolve.nested - Failed to find "${property}" property from warning. #4
- Fix -
resolve.headerRows calculates colSpans correctly now. #3
- Feature - Make
resolve.resolve method optional. If it's not provided, it will inject only _index to the data.
- Breaking - Drop
resolve.index. The functionality has been integrated to resolve.resolve. You can customize it through indexKey option there. #2
- Feature -
resolve.nested accepts custom getters through properties now. Example: property: data => (data.name || {}).first. #2
- Breaking - Drop
resolve.rowKey. That was moved to reactabular-table since it's too specific to fit here.
- Initial re-release under a different name.
- Feature - Add
resolve.columnChildren({ columns, childrenField = 'children' }) => <resolved columns>. Earlier this was in reactabular-utils but it fits this namespace better.
- Breaking - Allow resolvers to be composed more easily. Now the API follows the pattern
(extra) => (rowData) => <resolved row>. This means the functions fit within compose` like this:
const resolver = resolve.resolve({
columns,
method: (extra) => compose(
resolve.byFunction('cell.resolve')(extra),
resolve.nested(extra)
)
});
- Bug fix - Make sure
resolve does not crash if rows aren't provided. It will return an empty array in that case.
- Breaking - Rework
resolve interface to be object based and pass row index through it.
- Feature - Implement
resolve.index. This attached the row indices to _index. That can be handy data to have for optimization.
- Feature - Make sure
undefined keys aren't included in the resolved result.