MongoDB Schema to Dart Schema

This repo provides a method to generate Dart Schema from MongoDb

How to generate schema

• Install pnpm
• Run pnpm i to install dependencies
• Run cp .env.template .env to create a .env file
• Populate .env file with your MongoDB URL and DB_NAME
• Copy your sync rules to sync-rules.yaml
• Run pnpm generateSchema

You should see output similar to the following:

Schema([
  Table('products', [
    Column.text('_id'),
    Column.text('name'),
    Column.text('category'),
    Column.real('price'),
    Column.integer('inStock'),
    Column.text('features'),
    Column.text('dimensions'),
    Column.text('releaseDate'),
    Column.text('ratings')
  ]),
...
]);

Known Limitations

• This has only been minimally tested and therefore the output should be reviewed to confirm correctness.
• This only works for Dart Schema at present.
• None of the code has been optimized for performance.
• NumberDecimals are converted to text and need to be manually adjusted to real.

How to test

This assumes you have Docker installed and running:

• Run pnpm i
• cp .env.template .env
• Run ./setup_mongodb.sh
• Run pnpm generateSchema

You should see the following output:

Schema([
  Table('products', [
    Column.text('_id'),
    Column.text('name'),
    Column.text('category'),
    Column.real('price'),
    Column.integer('inStock'),
    Column.text('features'),
    Column.text('dimensions'),
    Column.text('releaseDate'),
    Column.text('ratings')
  ]),
  Table('users', [
    Column.text('_id'),
    Column.text('username'),
    Column.text('email'),
    Column.integer('age'),
    Column.integer('isActive'),
    Column.text('registrationDate'),
    Column.text('lastLogin'),
    Column.text('preferences'),
    Column.text('tags'),
    Column.text('scores')
  ]),
  Table('orders', [
    Column.text('_id'),
    Column.text('userId'),
    Column.text('orderDate'),
    Column.text('status'),
    Column.text('totalAmount'),
    Column.text('items'),
    Column.text('shippingAddress'),
    Column.text('paymentMethod'),
    Column.text('paypalTransactionId'),
    Column.text('bankTransactionId'),
    Column.text('lastFourDigits')
  ])
]);

How it works

• Collections are used to generate table names.
• Collection documents are sampled to generate the table columns. Number of documents sampled is based on SAMPLE_SIZE_INT set in .env file.
• Sync rules are used to filter out tables generated from collections that are not required.