Skip to content

Commit

Permalink
libcove2/common.py: Handle new jsonschema
Browse files Browse the repository at this point in the history
  • Loading branch information
Ed (ODSC) committed Sep 22, 2024
1 parent b909b0f commit 12b3bdf
Show file tree
Hide file tree
Showing 7 changed files with 6,504 additions and 11 deletions.
63 changes: 52 additions & 11 deletions libcove2/common.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -11,34 +11,75 @@
"^(.*_(((([A-Za-z]{2,3}(-([A-Za-z]{3}(-[A-Za-z]{3}){0,2}))?)|[A-Za-z]{4}|[A-Za-z]{5,8})(-([A-Za-z]{4}))?(-([A-Za-z]{2}|[0-9]{3}))?(-([A-Za-z0-9]{5,8}|[0-9][A-Za-z0-9]{3}))*(-([0-9A-WY-Za-wy-z](-[A-Za-z0-9]{2,8})+))*(-(x(-[A-Za-z0-9]{1,8})+))?)|(x(-[A-Za-z0-9]{1,8})+)))$" # noqa
)

def process_items(schema_dict, registry=None, defs=None):
items_schema_dicts = []
if "oneOf" in schema_dict["items"] and isinstance(
schema_dict["items"]["oneOf"], list
):
for oneOf in schema_dict["items"]["oneOf"]:
items_schema_dicts.append(oneOf)
elif '$ref' in schema_dict["items"]:
if schema_dict["items"]['$ref'].startswith('urn:'):
subschema = registry.contents(schema_dict["items"]['$ref'].split("#")[0])
if '#/$defs/' in schema_dict["items"]['$ref']:
name = schema_dict["items"]['$ref'].split('$defs/')[-1]
defs = subschema['$defs']
subschema = defs[name]
items_schema_dicts.append(subschema)
elif schema_dict["items"]['$ref'].startswith('#/$defs/'):
name = schema_dict["items"]['$ref'].split('$defs/')[-1]
items_schema_dicts.append(defs[name])
elif "properties" in schema_dict["items"] and isinstance(
schema_dict["items"]["properties"], dict
):
items_schema_dicts.append(schema_dict["items"])
return items_schema_dicts

def schema_dict_fields_generator(schema_dict):
def schema_dict_fields_generator(schema_dict, registry=None, defs=None):
if '$defs' in schema_dict:
defs = schema_dict['$defs']
if "properties" in schema_dict and isinstance(schema_dict["properties"], dict):
for property_name, value in schema_dict["properties"].items():
if "oneOf" in value:
if '$ref' in value:
if value['$ref'].startswith('urn:'):
property_schema_dicts = registry.contents(value['$ref'].split('#')[0])
if '$defs/' in value['$ref']:
name = value['$ref'].split('$defs/')[-1]
property_schema_dicts = property_schema_dicts['$defs'][name]
else:
name = value['$ref'].split('$defs/')[-1]
property_schema_dicts = defs[name]
property_schema_dicts = [property_schema_dicts]
elif ("items" in value and isinstance(value["items"], dict) and
(not "type" in value["items"] or value["items"]["type"] == "object")):
property_schema_dicts = process_items(value, registry=registry, defs=defs)
elif "oneOf" in value:
property_schema_dicts = value["oneOf"]
else:
property_schema_dicts = [value]
for property_schema_dict in property_schema_dicts:
if not isinstance(property_schema_dict, dict):
continue
if "properties" in property_schema_dict:
for field in schema_dict_fields_generator(property_schema_dict):
for field in schema_dict_fields_generator(property_schema_dict, registry=registry, defs=defs):
yield f"/{property_name}{field}"
elif "items" in property_schema_dict:
for field in schema_dict_fields_generator(
property_schema_dict["items"]
property_schema_dict["items"], registry=registry, defs=defs
):
yield f"/{property_name}{field}"
yield f"/{property_name}"
if "items" in schema_dict and isinstance(schema_dict["items"], dict):
if "oneOf" in schema_dict["items"] and isinstance(
schema_dict["items"]["oneOf"], list
):
for oneOf in schema_dict["items"]["oneOf"]:
for field in schema_dict_fields_generator(oneOf):
if "allOf" in schema_dict and isinstance(schema_dict["allOf"], list):
for clause in schema_dict["allOf"]:
if "then" in clause and isinstance(clause["then"], dict):
for field in schema_dict_fields_generator(clause["then"], registry=registry, defs=defs):
yield field

if ("items" in schema_dict and isinstance(schema_dict["items"], dict) and
(not "type" in schema_dict["items"] or schema_dict["items"]["type"] == "object")):
items_schema_dicts = process_items(schema_dict, registry=registry, defs=defs)
for items_schema_dict in items_schema_dicts:
for field in schema_dict_fields_generator(items_schema_dict, registry=registry, defs=defs):
yield field

def get_additional_fields_info(json_data, schema_fields, fields_regex=False):
fields_present = get_fields_present_with_examples(json_data)
Expand Down
102 changes: 102 additions & 0 deletions tests/fixtures/bods-data-0-3-0.json
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,102 @@
[
{
"statementID": "1dc0e987-5c57-4a1c-b3ad-61353b66a9b7",
"statementType": "entityStatement",
"isComponent": false,
"statementDate": "2017-11-18",
"entityType": "registeredEntity",
"name": "CHRINON LTD",
"foundingDate": "2010-11-18",
"identifiers": [
{
"scheme": "GB-COH",
"id": "07444723"
}
],
"publicListing": {
"hasPublicListing": true,
"companyFilingsURLs": ["http://example.com/"],
"securitiesListings": []
},
"publicationDetails": {
"publicationDate": "2018-02-13",
"bodsVersion": "0.3",
"publisher": {
"name": "CHRINON LTD"
}
}
},
{
"statementID": "019a93f1-e470-42e9-957b-03559861b2e2",
"statementType": "personStatement",
"isComponent": false,
"statementDate": "2017-11-18",
"personType": "knownPerson",
"nationalities": [
{
"code": "GB",
"name": "United Kingdom of Great Britain and Northern Ireland (the)"
}
],
"names": [
{
"type": "individual",
"fullName": "Christopher Taggart",
"givenName": "Christopher",
"familyName": "Taggart"
},
{
"type": "alternative",
"fullName": "Chris Taggart"
}
],
"birthDate": "1964-04",
"addresses": [
{
"type": "service",
"address": "Aston House, Cornwall Avenue, London",
"country": "GB",
"postCode": "N3 1LF"
}
],
"publicationDetails": {
"publicationDate": "2018-02-13",
"bodsVersion": "0.3",
"publisher": {
"name": "CHRINON LTD"
}
}
},
{
"statementID": "fbfd0547-d0c6-4a00-b559-5c5e91c34f5c",
"statementType": "ownershipOrControlStatement",
"isComponent": false,
"statementDate": "2017-11-18",
"subject": {
"describedByEntityStatement": "1dc0e987-5c57-4a1c-b3ad-61353b66a9b7"
},
"interestedParty": {
"describedByPersonStatement": "019a93f1-e470-42e9-957b-03559861b2e2"
},
"interests": [
{
"type": "shareholding",
"directOrIndirect": "direct",
"beneficialOwnershipOrControl": true,
"startDate": "2016-04-06",
"share": {
"exact": 100,
"minimum": 100,
"maximum": 100
}
}
],
"publicationDetails": {
"publicationDate": "2018-02-13",
"bodsVersion": "0.3",
"publisher": {
"name": "CHRINON LTD"
}
}
}
]
111 changes: 111 additions & 0 deletions tests/fixtures/bods-data-0-4-0.json
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,111 @@
[
{
"statementId": "1dc0e987-5c57-4a1c-b3ad-61353b66a9b7",
"declarationSubject": "c359f58d2977",
"statementDate": "2020-03-04",
"publicationDetails": {
"publicationDate": "2020-03-04",
"bodsVersion": "0.4",
"publisher": {
"name": "Profitech Ltd"
}
},
"recordId": "c359f58d2977",
"recordStatus": "new",
"recordType": "entity",
"recordDetails": {
"isComponent": false,
"entityType": {
"type": "registeredEntity"
},
"name": "Profitech Ltd",
"foundingDate": "2019-09-03",
"identifiers": [
{
"scheme": "GB-COH",
"id": "2063384560"
}
]
}
},
{
"statementId": "019a93f1-e470-42e9-957b-03559861b2e2",
"declarationSubject": "c359f58d2977",
"statementDate": "2020-03-04",
"publicationDetails": {
"publicationDate": "2020-03-04",
"bodsVersion": "0.4",
"publisher": {
"name": "Profitech Ltd"
}
},
"recordId": "10478c6cf6de",
"recordStatus": "new",
"recordType": "person",
"recordDetails": {
"isComponent": false,
"personType": "knownPerson",
"nationalities": [
{
"code": "GB",
"name": "United Kingdom of Great Britain and Northern Ireland (the)"
}
],
"names": [
{
"type": "legal",
"fullName": "Jennifer Hewitson-Smith",
"givenName": "Jennifer",
"familyName": "Hewitson-Smith"
},
{
"type": "alternative",
"fullName": "Jenny Hewitson-Smith"
}
],
"birthDate": "1978-07",
"addresses": [
{
"type": "service",
"address": "76 York Road Bournemouth",
"postCode": "BH81 3LO",
"country": {
"name": "United Kingdom",
"code": "GB"
}
}
]
}
},
{
"statementId": "fbfd0547-d0c6-4a00-b559-5c5e91c34f5c",
"declarationSubject": "c359f58d2977",
"statementDate": "2020-03-04",
"publicationDetails": {
"publicationDate": "2020-03-04",
"bodsVersion": "0.4",
"publisher": {
"name": "Profitech Ltd"
}
},
"recordId": "93b53022ae6a",
"recordStatus": "new",
"recordType": "relationship",
"recordDetails": {
"isComponent": false,
"subject": "c359f58d2977",
"interestedParty": "10478c6cf6de",
"interests": [
{
"type": "shareholding",
"beneficialOwnershipOrControl": true,
"directOrIndirect": "direct",
"startDate": "2016-04-06",
"share": {
"exact": 100
}
}
]
}
}
]
Loading

0 comments on commit 12b3bdf

Please sign in to comment.