reactpy-table

Headless UI for building powerful tables with ReactPy. The project takes its design ideas from the hugely popular, ReactJS based, TanStack Table.

The initial release supports the following features:

• Headless UI, CSS agnostic
• Freeform text search
• Forward/Reverse column based sort
• Pagination (offset, limit)
• Integration with SQLModel/SQLAlchemy (fts5 based full text search, sort and pagination)
• Integration with Pandas (in progress)
• Remote large dataset support (in progress)

Usage

pip install reactpy-table

basic_example.py

from reactpy import component, html, use_memo
from reactpy_table import Options, Table, use_reactpy_table

from .data.sp500 import COLS, CompanyModel, get_sp500

@component
def THead(table: Table[CompanyModel]):
    cols = table.data.cols
    return html.thead(
        html.th(cols[0].label),
        html.th(cols[1].label),
        html.th(cols[2].label),
        html.th(cols[3].label)
    )


def TRow(index: int, row: CompanyModel):
    return  html.tr(
        html.td(str(row.index)),
        html.td(row.symbol),
        html.td(row.name),
        html.td(row.sector),
    )

@component
def TBody(table: Table[CompanyModel]):
    rows = table.data.rows
    return  html.tbody(
        For(TRow, iterator=enumerate(rows))
    )

@component
def AppMain():
    table_data = use_memo(lambda:get_sp500(rows=50))
    table = use_reactpy_table(Options(
        rows=table_data,
        cols = COLS,
    ))

    return html.div(
        html.br(),
        html.h2('ReactPy Table Example'),
        html.table({"role": "grid"},
            THead(table),
            TBody(table)
        ),
    )

More complex examples

examples/table_example.py demonstrates search, sort and pagination on SP500 companies table. The built-in feature data pipeline is used.

examples/books/books_example.py demonstrates search, sort and pagination on a SQLite book database containing 100k books. Search, sort and pagination is handled by SQL queries.

• FTS5 full text search on database < 20ms
• Next/Prev page traversal < 2ms

Feature Set

The row data presented to the user for display is pre-processed by a data-pipeline of table features:

1. Search, filters the initial usr data
2. Sort, the search result by column (up/down)
3. Paginate, the sort result (if enabled)
4. RowOps, CRUD operations on the data

Each feature presents an API to the user that directs its operation. Any change will, if required, recalculate the table data.

A default set of features is applied by default. One or more custom features that will replace the default can be supplied as options when the table is created.

A callback mechanism is available for each feature that provides support for accessing external data sources (databases, pandas data-frames etc).

Custom Features

All features are instantiated using a strict pattern. As an example a custom paginator that accepts page _size and start_page would be created as follows:

def getCustomPaginator(page_size, start_page) 

    def _feature_factory(table, upstream_data):
        return CustomPaginator(table, upstream_data, page_size, start_page)
    
    return _feature_factory

The custom feature is passed as an option when the table is created:

table = use_reactpy_table(Options(
    ...
    paginator = getCustomPaginator(page_size=100, start_page=25)
))

The table initialization logic calls _feature_factory(table, updater) to supply reactpy's internal table and updater to the custom factory.