check xml catalog

check_xml_catalog.py

"""
Validates against internal mods xml-schema, indicating whether xmlcatalog is used.
"""

import os, unittest
from lxml import etree
import requests


class StrictCatalogResolver(etree.Resolver):
    def resolve(self, system_url, public_id, context):
        # Try to resolve using the catalog only
        resolved_url = self.resolve_filename(system_url, context)
        if resolved_url:
            return self.resolve_filename(resolved_url, context)
        else:
            raise ValueError(f"Failed to resolve {system_url} using the catalog")


class TestXMLCatalogProcessing(unittest.TestCase):

    def setUp(self):	
        # Set up the environment variable to point to the XML catalog if needed	
        os.environ['XML_CATALOG_FILES'] = '/foo/catalog.xml'	

    def test_schema_validation_from_root(self):
        ## Super-minimal MODS with schema location in the root element
        ## bad MODS; <blah> is not a valid element
        # xml_content = '''
        # <mods xmlns="http://www.loc.gov/mods/v3"
        #       xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
        #       xsi:schemaLocation="http://www.loc.gov/mods/v3 http://www.loc.gov/standards/mods/v3/mods-3-7.xsd">
        #   <titleInfo>
        #     <blah>Example Title</blah>
        #   </titleInfo>
        # </mods>
        # '''
        ## good MODS
        xml_content = '''
        <mods xmlns="http://www.loc.gov/mods/v3"
              xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
              xsi:schemaLocation="http://www.loc.gov/mods/v3 http://www.loc.gov/standards/mods/v3/mods-3-7.xsd">
          <titleInfo>
            <!-- <blah>Example Title</blah> -->
            <title>Example Title</title>
          </titleInfo>
        </mods>
        '''

        ## Parse the XML --------------------------------------------
        parser = etree.XMLParser( no_network=True )  # no_network=False allows the parser to fetch the schema from the network
        parser.resolvers.add(StrictCatalogResolver())
        xml_doc = etree.fromstring( xml_content.encode('utf-8'), parser )

        ## Get the schema url from the root element -----------------
        schema_locations = xml_doc.attrib['{http://www.w3.org/2001/XMLSchema-instance}schemaLocation'].split()  # the url in the brackets represents the xsi namespace
        schema_url = schema_locations[1]  # the second item is the actual schema URL

        ## load the schema ------------------------------------------
        try:
            # Attempt to parse the schema using the strict resolver
            schema_doc = etree.parse( schema_url, parser )
            print(etree.tostring(schema_doc, pretty_print=True).decode('utf-8'))
            print('Schema loaded from the catalog')
        except ValueError as e:
            print(f'Error: {e}')
            print('Schema could not be resolved from the catalog, and no fallback was allowed.')

        ## validate -------------------------------------------------
        validity = xmlschema.validate( xml_doc )
        if validity == True:
            print( 'XML passed validation against the schema' )
        else:
            print( f'XML failed validation against the schema; errors, ``{xmlschema.error_log}``' )
        self.assertEqual( True, validity )


if __name__ == '__main__':
    unittest.main()