dtulibrary/shohisha

Provider-Consumer metadata and access control

shohisha

Provider-Consumer metadata and access control.

This ruby project creates a concept of providers which deliver some sort of data. And consumers which are granted access to this data.

Resource structure

The following models are available providers consumers packages provider_types deliver_periods supplies fetchers transports ipaddresses retain_types fulltexts statuses

The 2 key resources are providers and consumers.

Providers have the following resources attached to it. This means that the following resources have a provider_id in them. packages fetchers ipaddresses

Providers references the following structures provider_types

There is a link between consumer and provider, and between consumer and packages. This link contains a reference to a fulltext resource.

Fetchers contains a reference to the following resources: transports statuses retain_types deliver_periods

Fetchers are used to get information about how to get data from a provider.

Ipaddresses contain an IP address, IP address range or a IP address/bitmask.

Supplies is used to identify what kind of data the fetcher returns.

Statutes is used to give the status for a given fetcher.

REST interface

Data can be extracted through a REST based interface.

For each resource available you can do the follwing requests GET /rest/.json GET /rest//.json

The returned format will be JSON. Any request not having .json added will return an empty response.

The following special lookup exists. GET /rest/providercode/(.json or .text) .text will return the id only as plain/text.

The nested resources can be used to limit what is extracted. GET /rest/providers//consumers_providers.json GET /rest/consumers//consumers_providers.json GET /rest/providers//consumers_packages.json GET /rest/consumers//consumers_packages.json GET /rest/providers//fetchers.json GET /rest/supplies//fetchers.json

ipaddresses is intended for storing IP addresses which should be able to access some kind of service related to this. E.g. an internal ftp server where data is dropped of by providers.

consumers_providers and consumers_packages will return information about links between consumer - provider and consumer - packages respectivily

How to use the REST interface

To get a list of providers do:

GET /rest/providers.json

This will return an array of providers in the JSON format. This could look like this: [ {"code":"acm", "created_at":"2012-08-17T09:50:08Z", "description":"American Computer Manufactoring", "id":1, "provider_type_id":1, "updated_at":"2012-08-17T09:50:08Z" }, {"code":"ebsco", "created_at":"2012-08-21T06:48:09Z", "description":"EBSCO", "id":2, "provider_type_id":1, "updated_at":"2012-08-21T06:48:09Z" }, {"code":"alephcash", "created_at":"2012-09-11T11:57:12Z", "description":"Aleph Cash", "id":3, "provider_type_id":1, "updated_at":"2012-09-11T11:57:12Z" } ]

Any value in the returned information from a REST request which ends in "_id" is a mapping to another view. You can find the reference name by removing "_id" and add "s" to the name. If the name ends in a "s" you must add "es".

So in the returned provider data you can lookup the provider type by doing GET /rest/provider_types/

If you find you need to do a lot of lookup on the type. You can also get the full list of provider_types by doing GET /rest/provider_types And then do the mapping yourself.

You can lookup a single provider by doing: GET /rest/providers/.json

E.g. GET /rest/providers/1.json

could return: { "code":"acm", "created_at":"2012-08-17T09:50:08Z", "description":"American Computer Manufactoring", "id":1, "provider_type_id":1, "updated_at":"2012-08-17T09:50:08Z" }

For a single provider yuo can also get a list of attached consumers, packages and fetchers.

E.g. to get all consumers for a given provider: GET /rest/providers/1/consumers.json

Which return an array of consumers linked to the given provider. Like this: [ {"code":"dtu", "created_at":"2012-08-20T09:53:46Z", "description":"DTU employees and staff", "id":1, "updated_at":"2012-08-20T09:53:46Z" }, {"code":"dtu_pub", "created_at":"2012-08-20T13:21:42Z", "description":"Public DTU", "id":2, "updated_at":"2012-08-20T13:21:42Z" } ]

Getting provider packages or consumer packages works the same way GET /rest/providers/2/packages.json

Which would return something like this: [ {"code":"ebsco_afh", "created_at":"2012-09-26T15:01:48Z", "description":"Ebsco AFH", "id":3, "provider_id":2, "updated_at":"2012-09-26T15:01:48Z" }, {"code":"buh", "created_at":"2012-08-21T06:54:19Z", "description":"EBSCO Buh", "id":1, "provider_id":2, "updated_at":"2012-08-21T06:54:19Z" } ]

Or if no packages are available (needed) it will return [].

GET /rest/consumers/1/packages.json

Would return all packages for all providers like this: [ {"code":"buh", "created_at":"2012-08-21T06:54:19Z", "description":"EBSCO Buh", "id":1, "provider_id":2, "updated_at":"2012-08-21T06:54:19Z" }, {"code":"acm_package", "created_at":"2012-09-26T10:50:13Z", "description":"Acm Package", "id":2, "provider_id":1, "updated_at":"2012-09-26T10:50:13Z" } ]

To get the actual information about a link between consumer and provider you use: GET /rest/consumers/1/consumers_providers.json

That will get all links from consumer 1 to a provider. It looks like this: [ {"consumer_id":1, "created_at":"2012-08-20T14:55:52Z", "fulltext_id":1, "id":1, "provider_id":1, "updated_at":"2012-08-20T14:55:52Z" } ]

That same is possible for packages GET /rest/consumers/1/consumers_packages.json

Or from the provider perspective GET /rest/providers/1/consumers_providers.json GET /rest/providers/1/consumers_packages.json

To get all fetchers do GET /rest/fetchers.json

Or to get all fetchers for a given provider. GET /rest/providers/1/fetchers.json