Apply ManGO metadata schemas in Python

Small application to apply metadata from a metadata schema using iRODS PRC.

Requirements

• python-irodsclient
• validators

Installation

You can install the package and its dependencies with pip:

pip install mango-mdschema

Usage

# authentication to iRODS
import os, os.path
from irods.session import iRODSSession
from mango_mdschema import Schema

env_file = os.getenv('IRODS_ENVIRONMENT_FILE', os.path.expanduser('~/.irods/irods_environment.json'))

# load a schema from file
my_schema = Schema('/path/to/schema') # this also validates it as a schema
with iRODSSession(irods_env_file=env_file) as session:
    obj = session.data_objects.get('path/to/my/object')
    my_metadata = {
        'text' : 'Some text',
        'a_date' : '2023-04-26',
        'author' : {
            'name' : 'Mariana',
            'email' : '[email protected]'
        }
    }
    my_schema.apply(obj, my_metadata)

The apply function will first validate the provided metadata dictionary before applying it on the iRODS object. See the tutorial for more details.

To only validate your metadata, without applying it, you can use the validate

function of the schema.

try:
    validated = my_schema.validate(metadata)
except (ConversionError, ValidationError) as err:
    print("Oops my metadata was not valid: {err}")

You can check the fields in a schema with:

print(my_schema)

The list of required fields and default values is found as the required_fields attribute. You can also check all the characteristics of a specific field such as 'name' with:

my_schema.print_requirements('name') # same as print(my_schema.fields['name'])

You can test the list of AVUs that would be sent by providing a given metadata dictonary like so:

avus = my_schema.to_avus(metadata)

The schema class can also read the metadata from ManGO and return it as nested dictionary with all values converted to their Python representation ( i.e. integer fields result in int values, datetime fields in datetime objects, etc.).

with iRODSSession(irods_env_file=env_file) as session:
    obj = session.data_objects.get('path/to/my/object')
    my_metadata = my_schema.extract(obj)

Check the tutorial notebook for more details.