AmiDb offers insert, delete, list, search and file services over http/s. The following explains the key building blocks that need to be understood before trying out other APIs.
This is very similar to a database table. The developer defines a model, say "Store", identifying the following:
- Persistence scope - Whether the data is stored by your customer ([email protected]), a section of your application (sports) or yourself ([email protected])
- Pscope: O - Username scope
- Pscope: C - Container scope
- Pscope: T - Tenant scope
- Engine - What type of engine you want to use. Supported are Cassandra and Elasticsearch.
- Names - Similar to column names in a database table.
- Types - The type of the data. One of string, integer, long, float, double, map and list.
AmiDb does not introduce another query language
AmiDb acts as a pass through allowing developers to use the query language of the underlying engine. This frees developers from having to learn something new or locking them into a new query language.
Restful - IDLS operations
Rows can be inserted, deleted, listed and searched (IDLS) using REST APIs.
Attachment Key (_ak)
Every row has a pointer that acts as its alias and is called the attachment key (_ak)
Attachments (files) can be added to any row. A row can have one more attachments.
Expiry and notifications
Every row can be marked to expire after a certain time. An expired row is auto-deleted and notified to the URL defined in the model.