Q: How would you model one-to-many using @begin/data and DynamoDB? about begin-data HOT 1 CLOSED

So we created @begin/data to make it easier to get started with DynamoDB doing simple persistence tasks. This is a perfect use case!

@begin/data

There is only one database table for all the data defined in an .arc app that uses @begin/data. It is defined as follows:

@app 
myapp

@tables
data
  scopeID *String
  dataID **String
  ttl TTL

The .arc above defines a single DynamoDB table with a partition key of scopeID and a sort key of dataID. This table can optionally have items that expire by setting the ttl attribute on rows

From here on reading and writing is done completely in code. @begin/data methods always have a table attribute but under the hood this is a virtual table of data! I would model the system with separate tables. A good convention is to use plural form of the noun the table represents. I'm not entirely sure what Image represents so I'm just going to focus on the user/org/post nouns.

let data = require('@begin/data')

// create a new org
await data.set({
  table: 'orgs', 
  key: 'engineering'
})

// adds a brand new user (and will automatically create key)
let user = await data.set({
  table: 'users',
  org: 'engineering',
  name: 'brian'
})

let post = await data.set({
  table: `users-${user.key}`,
  title: 'my post',
  body: 'some wisdom here'
})

Let me know if this makes sense so far!

from begin-data.