Library package that exports several methods for helping with DynamoDB operations. Abstracts most DynamoDB operations and builds query parameters using a unified filter object. Mainly aimed at single table design pattern
Import DynamoHelper
import { DynamoHelper } from '@oolio-group/dynamo-helper';
const { DynamoHelper } = require('@oolio-group/dynamo-helper');
Use constructor to create the DynamoHelper instance
// region and endpoint are optional
const dynamoHelper = new DynamoHelper(tableConf, region, endpoint);
tableConf should be of type TableConfig
type TableIndex = { partitionKeyName: string; sortKeyName: string };
export interface TableConfig {
name: string;
indexes: { default: TableIndex } & Record<string, TableIndex>;
}
import { DynamoHelper } from '@oolio-group/dynamo-helper';
const table = {
name: 'my-ddb-table',
indexes: {
default: {
partitionKeyName: 'pk',
sortKeyName: 'sk',
},
},
};
const client = new DynamoHelper(table, 'ap-south-1');
await client.getItem({ id: 'book-123' });
await client.getItem({ pk: 'library#books', sk: 'book-123' });
await client.query({
where: {
pk: 'library#books',
publishedAt: {
between: [15550000, 15800000],
},
},
});
import { buildQueryTableParams } from '@oolio-group/dynamo-helper';
const { buildQueryTableParams } = require('@oolio-group/dynamo-helper');
This method generates DynamoDB query input params from given filter object of type Filter<T>
buildQueryTableParams<T extends object = AnyObject>(
filter: Filter<T>,
partitionKey = 'pk',
sortKey = 'sk',
): QueryInput
It thoroughly validates input based on following criteria
- Filter must strictly be of type
Filter<T>
- PartitionKey and sortKey if given needs to be string
- PartitionKey cannot be empty and is required for query operation
- Where is required and needs to contain at least partition key condition
- Partition key condition can only be equal operation
- Fields if provided needs to contain string (fields of model)
- Limit if provided needs to be a number greater than zero
export interface Filter<MT extends object = AnyObject> {
/**
* The matching criteria
*/
where?: Where<MT>;
/**
* To include/exclude fields
*/
fields?: Fields<MT>;
/**
* Maximum number of entities
*/
limit?: number;
/**
* Sort order. Only works with sort keys
*/
orderBy?: Direction;
}
You can provide the name of your partition key and sort key, they are defaulted to pk
and sk
It returns a partial QueryInput
which does not contain TableName
or IndexName
import { buildQueryTableParams } from '@oolio-group/dynamo-helper';
// Get all inactive product id's in organization
const queryInput = buildQueryTableParams<ProductModel>({
where: {
pk: 'org_uuid',
sk: {
beginsWith: 'product_',
},
isActive: false,
},
fields: ['id'],
limit: 5,
});
Perform a query operation in DynamoDB. Input parameter Filter
can be customized with various operations
See type Filter
for supported operations
import { query } from '@oolio-group/dynamo-helper';
// Get all inactive product id's in organization
const products = await dynamoHelper.query<ProductModel>({
where: {
pk: 'org_uuid',
sk: {
beginsWith: 'product_',
},
isActive: false,
fields: ['id'],
},
});
- DynamoDB official docs - Paginate table query result and Query
- To use this method, the table must have a
range key
associated to partition key - Using the method from this library (
queryWithCursor
) Example 1:
// Paginate with a custom page size (refer to AWS DynamoDB docs to check the max size / limit)
// Get first `6` organizations (DynamoDB default sort order will be ascending)
const resultForFirstIterationOrPage = await query({
where: {
pk: 'org_uuid',
},
limit: 6,
});
// The next `6` orgs can be retrived based on the `cursor` value (from previous response)
const resultForFirstIterationOrPage = await query({
where: {
pk: 'org_uuid',
},
limit: 6,
prevCursor: result.cursor,
});
// The same step can be repeated until the `cursor` returns `undefined`
Example 2:
// Sort the results in reverse order with a page size `7`- descending
const result = await query({
where: {
pk: 'org_uuid',
},
limit: 7,
orderBy: Direction.DESC,
});
eq
neq
(filter conditions only)gt
gte
lt
lte
inq
between
like
beginsWith
Fetch an item using pk and sk combination. Returns item if found or returns null
getItem<T>(key: DocumentClient.Key, fields: Array<keyof T>)
Required, at least partition key values must be provided.
Optional, specify fields to project
import { getItem } from '@oolio-group/dynamo-helper';
// Get a single product matching the key
await dynamoHelper.getItem<ProductModel>({ pk: 'org_uuid', sk: 'product_xxx' });
await dynamoHelper.getItem<ProductModel>({ id: 'product_xxx' }, [
'id',
'isActive',
]);
Fetch many items using pk and sk combination
import { batchGetItems } from '@oolio-group/dynamo-helper';
// Get all products matching the keys
const products = await dynamoHelper.batchGetItems<ProductModel>([
{ pk: 'x', sk: '1' },
{ pk: 'y', sk: '2' },
]);
Check if an item exists in the database with the keys provided. Returns a boolean value
import { exists } from '@oolio-group/dynamo-helper';
// Check if product already exists
if (await dynamoHelper.exists({ id: 'x' })) {
console.log('exists');
}
Create a new item or replace an existing item in the database
import { putItem } from '@oolio-group/dynamo-helper';
await dynamoHelper.putItem({
pk: 'x',
sk: '1',
name: 'Product A',
});
Remove an item from database matching the key provided if it exists
import { deleteItem } from '@oolio-group/dynamo-helper';
// delete product
await dynamoHelper.deleteItem({ id: '1' });
Create or replace multiple items in the database as a transaction If any of the many operations fails then whole transaction is considered as failed. This is useful when multiple entries needs to be created or replaced for an API operation
import { transactPutItems } from '@oolio-group/dynamo-helper';
// initiate a transaction
await dynamoHelper.transactPutItems([
{
pk: 'product_1',
sk: 'product_1',
name: 'Product A',
},
{
pk: 'product_2',
sk: 'product_2',
name: 'Product B',
},
{
pk: 'org_1',
sk: 'product_1',
isActive: true,
},
]);
Create or replace multiple items in the database at the same time. This method will chunk your data into batches of 25 items.
The BatchWriteItem operation puts or deletes multiple items in one or more tables. A single call to BatchWriteItem can write up to 16 MB of data, which can comprise as many as 25 put or delete requests. Individual items to be written can be as large as 400 KB.
import { batchPutItems } from '@oolio-group/dynamo-helper';
// create multiple items
await dynamoHelper.batchPutItems([
{
pk: 'product_1',
sk: 'product_1',
name: 'Product A',
},
{
pk: 'product_2',
sk: 'product_2',
name: 'Product B',
},
{
pk: 'org_1',
sk: 'product_1',
isActive: true,
},
]);
Delete multiple items in the database at the same time. This method will chunk your data into batches of 25 items.
The BatchWriteItem operation puts or deletes multiple items in one or more tables. A single call to BatchWriteItem can write up to 16 MB of data, which can comprise as many as 25 put or delete requests. Individual items to be written can be as large as 400 KB.
import { batchDeleteItems } from '@oolio-group/dynamo-helper';
// delete multiple items
await dynamoHelper.batchDeleteItems([
{
pk: 'product_1',
sk: 'product_1',
},
{
pk: 'product_2',
sk: 'product_2',
},
{
pk: 'org_1',
sk: 'product_1',
},
]);
Checks if keys provided exists in database or not. Returns empty list if all items are found in DB. Returns list of items not found in DB if there are non existent items.
import { batchExists } from '@oolio-group/dynamo-helper';
const result = await dynamoHelper.batchExists([
{
pk: 'product_1',
sk: 'product_1',
},
{
pk: 'product_2',
sk: 'product_2',
},
{
pk: 'org_1',
sk: 'product_1',
},
]);
if (result.length === 0) {
// All items exists
} else {
// Items that does not exist
console.log(result);
}
Supports updating item conditionally using Conditional writes
import {
updateItem,
ConditionExpressionInput,
} from '@oolio-group/dynamo-helper';
const where = {
pk: 'product_123',
};
const conditions: ConditionExpressionInput[] = [
{
kind: ConditionExpressionKind.Comparison,
key: 'Price',
comparator: 'gt',
value: 50,
},
{
kind: ConditionExpressionKind.AndOr,
value: 'AND',
},
{
kind: ConditionExpressionKind.Comparison,
key: 'Quantity',
comparator: 'lt',
value: 20,
},
{
kind: ConditionExpressionKind.AndOr,
value: 'OR',
},
{
kind: ConditionExpressionKind.Comparison,
key: 'IsPromotionalProduct',
comparator: 'eq',
value: true,
},
{
kind: ConditionExpressionKind.AndOr,
value: 'OR',
},
{
kind: ConditionExpressionKind.Comparison,
key: 'color',
comparator: 'like',
value: 'red',
},
{
kind: ConditionExpressionKind.AndOr,
value: 'AND',
},
{
kind: ConditionExpressionKind.Comparison,
key: 'createdAt',
comparator: 'between',
value: [1670803200000, 1674586168676],
},
];
const prevPrice = 3000;
const prevSales = 2;
const updates = {
price: prevPrice - 100,
salesCount: prevSales + 1,
};
updateItem(where, conditions, updates);
- Scan operation support
- Support tables without sort key
Build all projects yarn build
Clean build output yarn clean
Follow these steps to publish package to npm whenever a new version is available:
- Step 1: Create a new publish branch
- Step 2: Change package version in
package.json
to whatever version that fits (Eg: 0.4.14) - Step 3: Commit the change with commit message as
Release <version number>
(Eg:Release 0.4.14
) - Step 4: Push all the changes and create a PR for the release