csf_utilities/reparsing_TO_REVIEW.py

# %%
import json
import csv

with open('./etc/man_draft.json', 'r') as in_file:
    data = json.load(in_file)
# %%
entity_rels = {ent for rel in data['Relazioni'] for ent in [rel['Entità 1'], rel['Entità 2']]}
# %%
entities = set(data['Entità'].keys())
entities.add('#any') # For compatibility
# %%
# Consistency check
if not entity_rels.issubset(entities):
    print(entity_rels.difference(entities))
# %%
with open('./dat/template.rdf', 'r') as in_file:
    raw_rdf = in_file.read()

entity_template = '''
    <!-- http://www.h2iosc.it/onto##NAME# -->

    <owl:Class rdf:about="&h2iosc;#NAME#">
        <rdfs:label>#LABEL#</rdfs:label>
        <rdfs:subClassOf>#PARENT#</rdfs:subClassOf>
    </owl:Class>
'''
subclass_string = "        <rdfs:subClassOf>#PARENT#</rdfs:subClassOf>\n"

object_property_template = '''
    <!-- http://www.h2iosc.it/onto##NAME# -->

    <owl:ObjectProperty rdf:about="&h2iosc;#NAME#">
        <rdfs:label>#LABEL#</rdfs:label>
        <rdfs:range rdf:resource="&h2iosc;#RANGE#"/>
        <rdfs:domain rdf:resource="&h2iosc;#DOMAIN#"/>
    </owl:ObjectProperty>
'''

object_property_inverse_template = '''
    <!-- http://www.h2iosc.it/onto##NAME# -->

    <owl:ObjectProperty rdf:about="&h2iosc;#NAME#">
        <rdfs:label>#LABEL#</rdfs:label>
        <owl:inverseOf rdf:resource="&h2iosc;#INV#"/>
        <rdfs:range rdf:resource="&h2iosc;#RANGE#"/>
        <rdfs:domain rdf:resource="&h2iosc;#DOMAIN#"/>
    </owl:ObjectProperty>
'''

datatype_property_template = '''
    <!-- http://www.h2iosc.it/onto##NAME# -->

    <owl:DatatypeProperty rdf:about="&h2iosc;#NAME#">
        <rdfs:label>#LABEL#</rdfs:label>
        <rdfs:domain rdf:resource="&h2iosc;#DOMAIN#"/>
    </owl:DatatypeProperty>
'''


def label_to_name(label):
    return label.replace(' ', '_').replace('à', 'a').replace('è', 'e').replace('é', 'e').replace('ì', 'i').replace('ò', 'o').replace('ù', 'u')


datatype_xsd = {
    "#string": 'string',
    '#uri': '#uri',
    '#number': 'decimal',
    '#date': 'date',
    '#coordinates': '#coordinates'
}

entities_rdf_list = []
entities_csv = []
datatype_properties_rdf_list = []
same_as = list(data['Same_as'].keys())
for label, ent in data['Entità'].items():
    
    entity_name = label_to_name(label)
    entity_rdf = entity_template.replace('#LABEL#', label).replace('#NAME#', entity_name)

    # Subclasses
    if 'Sottoclasse di' in ent.keys():
        parent = ent['Sottoclasse di']
        data['Relazioni'].append({"Entità 1": label,
            "Entità 2": parent,
            "Etichetta": "is_subclass_of", "Inversa": "is_superclass_of"})
        entity_rdf = entity_rdf.replace('#PARENT#', label_to_name(parent))
    else:
        entity_rdf = entity_rdf.replace(subclass_string, '')

    entities_rdf_list.append(entity_rdf)
    #
    if label in same_as:
        entities_csv.append( [label, "", ', '.join(data['Same_as'][label])] )
    else:
        entities_csv.append([label, "", ""])
    for datatype_label, datatype_val in ent.items():
        if not isinstance(datatype_val, str) or not datatype_val.startswith('#'):
            continue
        entities_csv.append(["", datatype_label, ""])
        datatype_name = label_to_name(datatype_label)
        datatype_properties_rdf_list.append(
            datatype_property_template.replace('#LABEL#', datatype_label).replace(
                '#NAME#', datatype_name
            ).replace('#DOMAIN#', entity_name)
        )

relations_rdf_list = []
relations_csv = []
for rel in data['Relazioni']:
    label = rel['Etichetta']
    inverse_label = rel['Inversa']
    domain = label_to_name(rel['Entità 1'])
    range1 = label_to_name(rel['Entità 2'])
    relations_csv.append([rel['Entità 1'], rel['Entità 2'], rel['Etichetta'], rel['Inversa']])
    name = domain + '_' + label_to_name(label) + '_' + range1
    inverse_name = range1 + '_' + label_to_name(inverse_label) + '_' + domain
    #
    relation_rdf = object_property_template.replace('#NAME#', name).replace('#LABEL#', label).replace('#DOMAIN#', domain).replace('#RANGE#', range1)
    #
    relation_inverse_rdf = object_property_inverse_template.replace('#NAME#', inverse_name).replace('#LABEL#', inverse_label).replace('#DOMAIN#', range1).replace('#RANGE#', domain).replace('#INV#', name)
    #
    relation_full_rdf = relation_rdf + '\n\n\n' + relation_inverse_rdf
    relations_rdf_list.append(relation_full_rdf)


with open('./etc/draft.rdf', 'w') as out_file:
    to_out = raw_rdf.replace(entity_template, '\n\n\n'.join(entities_rdf_list)).replace(
        datatype_property_template, '\n\n\n'.join(datatype_properties_rdf_list)
    ).replace(object_property_inverse_template, '\n\n\n'.join(relations_rdf_list))
    out_file.write(to_out)
# %%
with open('./etc/entities.csv', 'w') as out_csv:
    writer = csv.writer(out_csv)
    writer.writerow(['ENTITÀ', 'ATTRIBUTO (LITERAL)', 'SAME AS'])
    writer.writerows(entities_csv)
with open('./etc/relations.csv', 'w') as out_csv:
    writer = csv.writer(out_csv)
    writer.writerow(['ENTITÀ 1', 'ENTITÀ 2', 'NOME RELAZIONE', 'NOME RELAZIONE INVERSA'])
    writer.writerows(relations_csv)
# %%