pacifica-metadata 0.4.0

Pacifica Metadata

Pacifica Metadata Services

This is the Pacifica Metadata Services API.

This service provides metadata for objects and relationships between objects. See the metadata model page for a complete listing of objects and relationships between objects.

Installing the Service

Use the Docker Container

docker pull pacifica/metadata

The Docker Compose Way

docker-compose up

The Manual Way

Install the dependencies using the pip command:

pip install pacifica-metadata

Setup the database the default is PostgreSQL.

pacifica-metadata-cmd dbsync

Then, run the code:

pacifica-metadata

The API

The Pacifica Metadata Services API covers the complete object life-cycle: create, read, update, and delete.

The examples in this section demonstrate the life-cycle of the User object using a Pacifica Metadata Services deployment at http://localhost:8121/ (see "Installing the Service" section).

Creating an Object

To create a new User object, start by generating a new file create.json to store the JSON data:

{
  "_id": 127,
  "email_address": "[e-mail]",
  "first_name": "John",
  "last_name": "Doe",
  "middle_initial": "",
  "network_id": "guest"
}

Then, provide the contents of the new file as the body for a HTTP PUT request using the curl command:

curl -X PUT -T create.json 'http://localhost:8121/users'

Reading an Object

To retrieve the JSON data for the User object that was just created, send a HTTP GET request (with the _id attribute as a query parameter) using the curl command:

curl -X GET 'http://localhost:8121/users?_id=127'

The response body is a JSON array of JSON objects. Because the query uses the _id attribute, a primary key, the JSON array contains either zero (no match) or one (match) JSON objects:

[
  {
    "_id": 127,
    "email_address": "[e-mail]",
    "encoding": "UTF8",
    "first_name": "John",
    "last_name": "Doe",
    "middle_initial": "",
    "network_id": "guest",
    "created": 1459204793,
    "deleted": null,
    "updated": 1459204793
  }
]

Optionally, query on any other parts of an object by using its attributes as query parameters, e.g., to query on both the first_name and last_name attributes using the curl command:

curl -X GET 'http://localhost:8121/users?last_name=Doe&first_name=John'

Response bodies for queries on other parts may contain JSON data for more than one match:

[
  {
    "_id": 127,
    "email_address": "[e-mail]",
    "encoding": "UTF8",
    "first_name": "John",
    "last_name": "Doe",
    "middle_initial": "",
    "network_id": "guest",
    "created": 1459204793,
    "deleted": null,
    "updated": 1459204793
  },
  ...
]

Pagination

By default, Pacifica Metadata Services API endpoints respond with JSON data for all matching objects. Hence, the response size, and therefore, the response time is proportional to the number of matches.

To reduce the response time, but still enable access to all matching objects, all Pacifica Metadata Services API endpoints offer a "pagination" capability, where large responses are subdivided into "pages" with a fixed number of matching objects "per page".

For example, 100 matching objects are divided into 4 pages of 25 matching objects per page:

• Page 1 = Matching objects 1 to 25
• Page 2 = Matching objects 26 to 50
• Page 3 = Matching objects 51 to 75
• Page 4 = Matching objects 76 to 100
• Page >4 = No matching objects

Use the page_number and items_per_page query parameters to specify the number of pages and the number of matching objects per page, respectively, e.g., to retrieve the 7th page of 10 Users per page using the curl command:

curl -X GET 'http://localhost:8121/users?items_per_page=10&page_number=7'

Updating an Object

To modify a preexisting object, use the query parameters to identify the object (or objects) and then send a HTTP POST request with the JSON data for the modified attributes as the request body, e.g., to modify the network_id attribute, start by generating a new file update.json to store the JSON data:

{
  "network_id": "example"
}

Then, provide the contents of the new file as the body for a HTTP POST request using the curl command:

curl -X POST -T update.json 'http://localhost:8121/users?last_name=Doe&first_name=John'

Finally, verify the modifications by retrieving the most recent version of the object (see "Reading an Object" section), e.g., using the curl command:

curl -X GET 'http://localhost:8121/users?_id=127'

The updated attribute is automatically set to the current time when an object is modified:

[
  {
    "_id": 127,
    "email_address": "[e-mail]",
    "encoding": "UTF8",
    "first_name": "John",
    "last_name": "Doe",
    "middle_initial": "",
    "network_id": "example",
    "created": 1459204793,
    "deleted": null,
    "updated": 1459205143
  }
]

(Soft) Deleting an Object

To mark an object as deleted, i.e., to "soft delete" an object, send a HTTP DELETE request using the curl command:

curl -X DELETE 'http://localhost:8121/users?_id=127'

NOTE Don't worry! The object isn't really deleted.

Finally, verify the "soft delete" by retrieving the most recent version of the object (see "Reading an Object" section), e.g., using the curl command:

curl -X GET 'http://localhost:8121/users?_id=127'

The deleted attribute is automatically set to the current time when an object is "soft deleted":

[
  {
    "_id": 127,
    "email_address": "[e-mail]",
    "encoding": "UTF8",
    "first_name": "John",
    "last_name": "Doe",
    "middle_initial": "",
    "network_id": "example",
    "created": 1459204793,
    "deleted": 1459205341,
    "updated": 1459205143
  }
]

Contributions

Contributions are accepted on GitHub via the fork and pull request workflow. GitHub has a good help article if you are unfamiliar with this method of contributing.