Skip to content
/ bigal Public

A fast, lightweight ORM for PostgreSQL and node.js, written in Typescript.

License

Notifications You must be signed in to change notification settings

bigalorm/bigal

Folders and files

NameName
Last commit message
Last commit date

Latest commit

a409fa1 · Sep 27, 2024
Jan 26, 2023
Jul 2, 2024
Feb 7, 2024
Aug 26, 2024
Aug 26, 2024
Nov 22, 2017
May 13, 2021
Mar 18, 2019
Jul 2, 2024
Aug 26, 2024
Jul 19, 2023
Oct 1, 2021
Nov 22, 2017
Jul 2, 2024
Aug 26, 2024
Sep 27, 2024
Aug 26, 2024
Jul 2, 2024
Jul 2, 2024

Repository files navigation

BigAl

NPM version node version Known Vulnerabilities

A fast, lightweight ORM for PostgreSQL and node.js, written in Typescript.

This ORM does not:

  • Create or update db schemas for you
  • Handle associations/joins
  • Do much else than basic queries, inserts, updates, and deletes

Compatibility

  • PostgreSQL 14 or above. Lower versions should work.

Install

npm install pg postgres-pool bigal

Configuring

Defining database models

Model definitions need to extend Entity.

import { column, primaryColumn, table, Entity } from 'bigal';
import { Store } from './Store';
import { Category } from './Category';
import { ProductCategory } from './ProductCategory';

@table({
  name: 'products',
})
export class Product extends Entity {
  @primaryColumn({ type: 'integer' })
  public id!: number;

  @column({
    type: 'string',
    required: true,
  })
  public name!: string;

  @column({
    type: 'string',
  })
  public sku?: string;

  @column({
    type: 'string[]',
    defaultsTo: [],
    name: 'alias_names',
  })
  public aliases?: string[];

  @column({
    model: () => Store.name,
    name: 'store_id',
  })
  public store!: number | Store;

  @column({
    collection: () => Category.name,
    through: () => ProductCategory.name,
    via: 'product',
  })
  public categories?: Category[];
}

Initialize repositories

import {
  initialize,
  Repository,
} from 'bigal';
import { Pool } from 'postgres-pool';
import {
  Category,
  Product,
  ProductCategory,
  Store,
} from './models';

let pool: Pool;
let readonlyPool: Pool;

export function startup({
  connectionString,
  readonlyConnectionString,
}: {
  connectionString: string,
  readonlyConnectionString: string,
}) {
  pool = new Pool(connectionString);
  readonlyPool = new Pool(readonlyConnectionString);

  const repositoriesByName = initialize({
    models: [
      Category,
      Product,
      ProductCategory,
      Store,
    ],
    pool,
    readonlyPool,
  });

  let categoryRepository: Repository<Category>;
  let productRepository: Repository<Category>;
  let storeRepository: Repository<Category>;
  for (const [modelName, repository] = Object.entries(repositoriesByName)) {
    switch (modelName) {
      case 'Category':
        categoryRepository = repository;
        break;
      case 'Product':
        productRepository = repository;
        break;
      case 'Store':
        storeRepository = repository;
        break;
    }
  }

  return {
    categoryRepository,
    productRepository,
    storeRepository,
  }
}

export function shutdown() {
  const shutdownEvents = [];
  if (pool) {
    shutdownEvents.push(pool.end());
  }

  if (readonlyPool) {
    shutdownEvents.push(readonlyPool.end());
  }

  return Promise.all(shutdownEvents);
}

Repository class methods

.findOne() - Fetch a single object

Fetch a single object with where criteria specified as a chained method

const item = await ProductRepository.findOne().where({
  id: context.params.id,
});

Fetch a single object and restrict columns selected from db (query projection)

const item = await ProductRepository.findOne({
  select: ['name'],
}).where({
  id: context.params.id,
});

Fetch a single object using explicit DB pool

const poolOverride = new Pool(connectionString);

const item = await ProductRepository.findOne({
  pool: poolOverride,
}).where({
  id: context.params.id,
});

Populate relation - Relations can be one-to-many (ether direction) or many-to-many

const item = await ProductRepository.findOne()
  .where({
    id: context.params.id,
  })
  .populate('store', {
    select: ['name'],
  });

Populate relation using explicit DB pool

const poolOverride = new Pool(connectionString);

const item = await ProductRepository.findOne()
  .where({
    id: context.params.id,
  })
  .populate('store', {
    select: ['name'],
    pool: poolOverride,
  });

Fetch a single object and perform a db sort before returning result

const item = await ProductRepository.findOne()
  .where({
    foo: context.params.foo,
  })
  .sort('name asc');

.find() - Fetch a multiple objects

Fetch multiple objects with where criteria specified as a chained method

const items = await ProductRepository.find().where({
  foo: context.params.foo,
});

Fetch multiple objects and restrict columns selected from db (query projection)

const items = await ProductRepository.find({
  select: ['name'],
}).where({
  foo: context.params.foo,
});

Fetch a multiple objects using explicit DB pool

const poolOverride = new Pool(connectionString);

const item = await ProductRepository.find({
  pool: poolOverride,
}).where({
  foo: context.params.foo,
});

Example of an OR statement

const items = await PersonRepository.find().where({
  firstName: {
    like: ['walter', 'Jess%'],
  },
});

Example of an AND statement

const items = await PersonRepository.find().where({
  lastName: {
    '!': {
      lastName: [null, '', 'Whi%'],
    },
  },
});

Fetch multiple objects and perform a db sort before returning result

const items = await PersonRepository.find()
  .where({
    firstName: {
      like: 'walter',
    },
    lastName: {
      like: 'white',
    },
  })
  .sort({
    age: 1,
    occupation: -1,
  });

Limit number results returned

const items = await PersonRepository.find()
  .where({
    age: [22, 23, 24],
  })
  .limit(42);

Skip x results

const items = await FooRepository.find()
  .where({
    or: [
      {
        foo: context.params.foo,
      },
      {
        bar: context.params.foo,
      },
    ],
  })
  .skip(42);

Page results using skip() & limit()

const items = await FooRepository.find()
  .where({
    foo: context.params.foo,
  })
  .skip(84)
  .limit(42);

Page results using paginate

const page = 2;
const pageSize = 42;
const items = await FooRepository.find()
  .where({
    foo: context.params.foo,
  })
  .paginate(page, pageSize);

.count() - Get the number of records matching the where criteria

const count = await PersonRepository.count().where({
  name: {
    like: 'Karl',
  },
});
// count = 3

.create() - Insert one or multiple objects

Insert a single object

const item = await PersonRepository.create({
  name: 'Karl',
});
// item = { id: 42, name: 'Karl', createdAt: ... }

Insert a single object without returning results from the db

await PersonRepository.create(
  {
    name: 'Karl',
  },
  {
    returnRecords: false,
  },
);

Insert a single object but limit columns returned from db for inserted records (query projection)

const item = await PersonRepository.create(
  {
    name: 'Karl',
  },
  {
    returnSelect: ['name'],
  },
);
// item = { id: 42, name: 'Karl' }

Note: The primary key will always be included. To only return the primary key value, pass an empty array

Insert multiple objects

const items = await PersonRepository.create([
  {
    name: 'LX',
  },
  {
    name: 'Big Al',
  },
]);
// items = [{ id: 24, name: 'LX', createdAt: ... }, { id: 25, name: 'Big Al', createdAt: ... }]

Insert multiple objects without returning results from the db

await PersonRepository.create(
  [
    {
      name: 'LX',
    },
    {
      name: 'Big Al',
    },
  ],
  {
    returnRecords: false,
  },
);

Insert multiple objects with limited return properties

const items = await PersonRepository.create(
  [
    {
      name: 'LX',
    },
    {
      name: 'Big Al',
    },
  ],
  {
    returnSelect: ['id'],
  },
);
// items = [{ id: 24 }, { id: 25 }]

Note: The primary key will always be included. To only return the primary key value, pass an empty array


Insert a single object with onConflict ignore (ON CONFLICT DO NOTHING)

const item = await PersonRepository.create(
  {
    name: 'Karl',
  },
  {
    onConflict: {
      action: 'ignore',
      targets: ['name'],
    },
  },
);
// item = { id: 42, name: 'Karl', createdAt: ... }

Insert a single object with onConflict merge (ON CONFLICT DO UPDATE) - Update all data

const item = await PersonRepository.create(
  {
    name: 'Karl',
  },
  {
    onConflict: {
      action: 'merge',
      targets: ['ssn'],
    },
  },
);
// item = { id: 42, name: 'Karl', createdAt: ... }

Insert a single object with onConflict merge (ON CONFLICT DO UPDATE) - Update specific data

const item = await PersonRepository.create(
  {
    name: 'Karl',
  },
  {
    onConflict: {
      action: 'merge',
      targets: ['ssn'],
      merge: ['name', 'age'],
    },
  },
);
// item = { id: 42, name: 'Karl', createdAt: ... }

.update() - Update objects

Update single record

const items = await PersonRepository.update(
  {
    id: 42,
  },
  {
    name: 'Big Al',
  },
);
// items = [{ id: 42, name: 'Big Al', createdAt: ... }]

Note: This method will return an array, regardless of how many records were affected

Update record without returning results from the db

await PersonRepository.update(
  {
    id: 42,
  },
  {
    name: 'Big Al',
  },
  {
    returnRecords: false,
  },
);

Update records and limit columns returned from db for affected records (query projection)

const items = await PersonRepository.update(
  {
    id: [42, 43],
  },
  {
    occupation: 'Water Purification Engineer',
  },
  {
    returnSelect: ['id'],
  },
);
// items = [{ id: 42 }, { id: 43 }]

.destroy() - Delete objects from the db

Delete single record

const items = await PersonRepository.destroy({
  id: 42,
});
// items = [{ id: 42, name: 'Big Al', createdAt: ... }]

Note: This method will return an array, regardless of how many records were affected

Delete record without returning row data from the db

await PersonRepository.destroy(
  {
    id: 42,
  },
  {
    returnRecords: false,
  },
);

Delete records and limit columns returned from db for affected records (query projection)

const items = await PersonRepository.destroy(
  {
    id: [24, 25],
  },
  {
    returnSelect: ['name'],
  },
);
// items = [{ id: 24, name: 'LX' }, { id: 25, name: 'Big Al' }]

Note: The primary key will always be included. To only return the primary key value, pass an empty array

Known issues

Entity collections must be optional

BigAl expects that all entity collection properties must be optional. There will be some type issues with QueryResult if you make a collection non-optional.

For example:

export class Store extends Entity {
  @primaryColumn({ type: 'integer' })
  public id!: number;

  @column({
    type: 'string',
    required: true,
  })
  public name!: string;

  // This property MUST be optional
  @column({
    collection: () => Product.name,
    via: 'store',
  })
  public products?: Product[];
}

Non-entity object arrays

If you have a json property, with an id field, on an entity model, TypeScript will probably think it is a BigAl entity due to how the type system works. In that case, you'll want to wrap the type with NotEntity<>. For example:

export interface IMyJsonType {
  id: string;
  foo: string;
}

export class Product extends Entity {
  @primaryColumn({ type: 'integer' })
  public id!: number;

  @column({
    type: 'string',
    required: true,
  })
  public name!: string;

  @column({
    type: 'json',
  })
  public myJson?: NotEntity<IMyJsonType>;
}

Debugging

Debugging can be enabled by passing the DEBUG_BIGAL environment flag with a value of true.

Debugging will print the generated SQL code in the console.

License

MIT