Ra-data-postgraphile NPM

ra-data-postgraphile

Postgraphile data provider for react-admin

This data provider heavily extends https://github.com/BowlingX/ra-postgraphile with a more generic approach.

Install

$ yarn add ra-data-postgraphile / npm install ra-data-postgraphile --save

Usage

The ra-data-postgraphile data provider accepts these arguments:

client - The ApolloClient instance to use.
cache - optional apollo cache default: return new InMemoryCache({dataIdFromObject: (object: any) => object.nodeId || null})
config - optional configuration
pgDataProvider({client, cache, config})

The following examples shows the basic usage:

import React, { useEffect, useState } from 'react'
import { Admin, Resource, Loading } from 'react-admin'
import pgDataProvider from 'ra-data-postgraphile'
import { CompanyList, CompanyEdit, CompanyCreate } from './posts'

const App = () => {
  const [dataProvider, setDataProvider] = useState(null)

  useEffect(() => {
    ;(async () => {
      const dataProvider = await pgDataProvider({
        uri: '/graphql',
        options: {
          resources: {
            Company: {
              pluralizedName: 'Companies',
            }
          }
        }
      })
      setDataProvider(dataProvider)
    })()
  }, [])

  if (!dataProvider) {
    return <Loading />
  }
  return (
    <Admin dataProvider={dataProvider}>
      <Resource
        name="Company"
        list={CompanyList}
        edit={CompanyEdit}
        create={CompanyCreate}
      />
    </Admin>
  )
}

export default App

Features

Filter Operations

It is possible to use different filter operations for the list filters. The operation can be provided in the filter name:

name = <propertyName> <operation>

The possible operations depend on the field type of the propertyName.

Example:

Filter field refId is null: refId null

See src/filters.ts for the integrated filters:

Compound Primary Keys

Because react admin expect that a record has a single id field ra-data-postgraphile detects compound primary keys and switch to use nodeId as the internal id for the record.

The internal representation of a record is different for compound keys. The original id field is copied to __rawId and nodeId is copied into the id field. This makes sure react admin receives a unique id value.

When your project contains compound primary keys it is also necessary to change the apollo cache to not use the id field. Use this code to create an apollo instance which uses the nodeId as primary key:

  const cache = new InMemoryCache({
    dataIdFromObject: object => object.nodeId || null
  })
  const client = new ApolloClient({
    uri: '/graphql',
    cache,
  })

Limitations

ra-data-postgraphile expects these plugins to be installed on the postgraphile server:

const PgSimplifyInflectorPlugin = require('@graphile-contrib/pg-simplify-inflector')
const PgConnectionFilterPlugin = require('postgraphile-plugin-connection-filter')

TODO

implement DELETE_MANY

Release

To release the package:

update changes (version should match the one you chose in the next step)
git add CHANGELOG.md to make sure changelog update is added to the next commit done by yarn version
update the version: yarn version [--major|minor|patch]
push changes: git push && git push --tags
publish to npmjs.com by running yarn release (might require to run yarn login beforehand)