Skip to content

diogoko/dynamo-facade-v3

Repository files navigation

dynamo-facade-v3

QA npm bundle size install size downloads Known Vulnerabilities Coverage Status

dynamo-facade-v3 is a library that helps calling DynamoDBClient of @aws-sdk/lib-dynamodb - specially functions that have expression parameters (like FilterExpression and UpdateExpression).

This project is focused on version 3 of @aws-sdk/lib-dynamodb.

Usage

You can see the complete reference to the API at https://diogoko.github.io/dynamo-facade-v3.

Install

npm install dynamo-facade-v3

Examples

In the following examples, the movies table has a compound primary key (actor is the hash key, movie is the range key).

import { fromIni } from "@aws-sdk/credential-providers";
import df, { between, gt, inList, transactWrite as tr } from 'dynamo-facade-v3';

// Use credentials from ~/.aws/credentials
df.defaults.credentials = fromIni({ profile: 'my_project' });
df.defaults.region = 'us-east-1';

// The returned value is the same one returned by DynamoDBDocumentClient.send()
const response = await df.get('movies', { actor: 'Tom Hanks' });
console.log(response.Item);

// query(tableName, keyCondition, options)
await df.query('movies', { actor: 'Tom Hanks' }, { filter: { year: gt(2000) } })

// scan(tableName, filter, options)
await df.scan('movies', { genre: inList('Drama', 'Action'), year: between(1990, 1999) })

// put(tableName, item, options)
await df.put('movies', { actor: 'Tom Hanks', movie: 'Finch', year: 2021 });

// update(tableName, key, updatedValues, options)
await df.update(
  'movies',
  { actor: 'Tom Hanks', movie: 'Forrest Gump' },
  { year: 1994, tomatometer: 71 }
)

// delete(tableName, key, options)
await df.delete('movies', { actor: 'Tom Hanks', movie: 'The Bonfire of the Vanities' })

// transactWrite(transactItems, options)
await df.transactWrite([
  tr.put('movies', { actor: 'Tom Hanks', movie: 'Toy Story 2', year: 1999 }),
  tr.update('movies', { actor: 'Tom Hanks', movie: 'Big'}, { year: 1988 }),
])

Expressions

Every command option that is an expression (KeyConditionExpression, FilterExpression - and soon ProjectionExpression too) can be described by an object whose keys are the table attributes, and whose values indicate the comparison to be made. If none of the operator helpers is used, the equals (=) operator is assumed.

df.scan('movies', { actor: 'Tom Hanks' })
// FilterExpression: '#actor = :actor'
// ExpressionAttributeValues: { ':actor': 'Tom Hanks' }

There are several operator helpers that can be imported as functions from this module. To use them, call the helper passing the compared value as its parameter.

import { gt } from 'dynamo-facade-v3';
df.scan('movies', { year: gt(2000) })
// FilterExpression: '#year > :year'
// ExpressionAttributeValues: { ':year': 2000 }

Notice that the ExpressionAttributeValues and ExpressionAttributeNames options are built automatically for you.

Queries with duplicate attributes

Sometimes there's a need to use the same attribute name more than once. It may happen when using the same attribute both in the key condition and in the filter expressions of a query. Other times, the same attribute needs to appear in multiple comparisons in the same filter, update condition, or condition check.

As objects keys must be unique in JavaScript, you can also create expressions using arrays of [key, value] pairs (just like the result of Object.entries).

This means you can safely do this:

df.query(
  'movies',
  { actor: 'Sylvester Stallone', movie: begins_with('Rambo') },
  {
    filter: [
      ['year', gt(1980)],
      ['year', lt(1990)],
      ['movie', contains('Blood')],
    ],
  }
)

Extra options

All commands accept an optional parameters object as their last argument. This way, you can use options from the original *Command classes.

df.scan(
  'movies',
  { year: gt(2000) },
  { Limit: 30 }
)

If you specify parameters that are automatically managed by the library functions, you override what was built automatically.

df.scan(
  'movies',
  { year: gt(2000) },
  { FilterExpression: 'and actor = :actor' } // oops!
)
// FilterExpression: 'and actor = :actor'    // ouch!

Responses

All functions that call original the *Command classes return the promise created by the send() method from DynamoDBDocumentClient.

const item = await df.get('movies', { actor, movie });          // ok
const item = await df.get('movies', { actor, movie }).promise() // wrong!

Development

Build

Run (this project is using Yarn 1.x):

yarn build

This will create the compiled files under ./dist folder.

Test

Run to execute tests with Jest:

yarn test

Generating docs

This project uses TypeDoc.

Run yarn make:docs and a folder named docs will be created in your root directory.

Thanks

This project was created from the alioguzhan/typescript-library-template project template.