plastron-repo

Fedora repository resources and operations

Repository

There are several class methods on the Repository class available as shortcuts to quickly create a Repository object.

# simple repo, no authentication
from plastron.repo import Repository

repo = Repository.from_url('http://localhost:8080/fcrepo/rest')

# repo with HTTP Basic authentication
from plastron.repo import Repository
from requests.auth import HTTPBasicAuth

repo = Repository.from_url(
    url='http://localhost:8080/fcrepo/rest',
    # any class that implements requests.auth.AuthBase will work here
    auth=HTTPBasicAuth('username', 'password'),
)

# repo with settings from a configuration dictionary
from plastron.repo import Repository

repo = Repository.from_config({
    'REST_ENDPOINT': 'http://localhost:8080/fcrepo/rest',
    'RELPATH': '/',
    'REPO_EXTERNAL_URL': 'http://fcrepo-local/fcrepo/rest',
    'SERVER_CERT': 'path/to/cert.pem',
    # authentication section
    'FEDORA_USER': 'username',
    'FEDORA_PASSWORD': 'password',
})

# repo with settings from a configuration YAML file;
# the structure of the YAML file is the same as the
# configuration dictionary
from plastron.repo import Repository

repo = Repository.from_config_file('config.yml')

You can always use the full constructor as well.

from requests.auth import HTTPBasicAuth

from plastron.client import Client, Endpoint
from plastron.repo import Repository

repo = Repository(
    client=Client(
        endpoint=Endpoint(
            url='http://localhost:8080/fcrepo/rest',
            default_path='/',
            external_url='http://fcrepo-local/fcrepo/rest',
        ),
        auth=HTTPBasicAuth('username', 'password'),
        server_cert='path/to/cert.pem',
        ua_string='my-client/3.14',
        on_behalf_of='user',
    )
)

Repository Resources

The repository resource classes represent a single URL-addressable HTTP resource that is stored in a repository. The most basic of these is RepositoryResource.

from plastron.repo import Repository, RepositoryResource

repo = Repository.from_url('http://localhost:8080/fcrepo/rest')

# Repository object support indexing with a path
resource = repo['/foo']
assert resource.url == 'http://localhost:8080/fcrepo/rest/foo'
assert isinstance(resource, RepositoryResource)

# this is merely syntactic sugar for the get_resource() method
resource = repo.get_resource('/foo')
assert resource.url == 'http://localhost:8080/fcrepo/rest/foo'
assert isinstance(resource, RepositoryResource)

When getting a resource from a Repository, you can provide a class to instantiate (it defaults to RepositoryResource). So to get an LDP container representation of a resource, use ContainerResource instead.

from plastron.repo import Repository, ContainerResource

repo = Repository.from_url('http://localhost:8080/fcrepo/rest')

container = repo.get_resource('/bar', ContainerResource)
assert container.url == 'http://localhost:8080/fcrepo/rest/bar'
assert isinstance(container, ContainerResource)

# you can also use slice notation to provide the class
container = repo['/bar':ContainerResource]
assert container.url == 'http://localhost:8080/fcrepo/rest/bar'
assert isinstance(container, ContainerResource)

Retrieving Data

None of the methods above actually send any HTTP requests. To connect to the repository and retrieve data, use the read() method on a resource.

RDF Descriptions

Repository resources can be described using RDF resource classes defined using the plastron-rdf library. A resource description is an RDF resource class that shares its underlying RDFLib Graph object with a repository resource. Thus changes to the resource description object will be reflected in the repository resource's graph and will be saved when the repository resource is written back to the repository.

Here is an (extremely simplified) example of retrieving a resource, updating some properties, and writing it back to the repository.

from rdflib import Literal, URIRef

from plastron.rdfmapping.resources import RDFResource
from plastron.repo import Repository, ContainerResource

repo = Repository.from_url('http://localhost:8080/fcrepo/rest')
resource = repo['/obj/123':ContainerResource]
resource.read()

# RDFResource is a very simple RDF description class
# it maps rdf:type to rdf_type and rdfs:label to label
obj = resource.describe(RDFResource)
obj.label = Literal('Digital Object 123')
obj.rdf_type.add(URIRef('http://pcdm.org/models#Object'))

assert obj.has_changes

# sends a PATCH request to the repository with a SPARQL
# update query built from the differences between the
# graph read from the repo and the current graph
resource.update()

Binaries