Simplifed with default value for parse via parse_taxonomy_url & fixed…

… the set hashing bug.

tests/test_local_taxonomy.py

@@ -18,10 +18,10 @@ def test_parse_taxonomy(self):
         cache_dir: str = './cache/'
         cache: HttpCache = HttpCache(cache_dir)
         print(f"Saving to {cache_dir}")
-
+        imported_schema_uris = set()
         extension_schema_path: str = './tests/data/example.xsd'
         # extension_schema_path: str = './data/example.xsd'
-        tax: TaxonomySchema = parse_taxonomy(extension_schema_path, cache)
+        tax: TaxonomySchema = parse_taxonomy(extension_schema_path, cache, imported_schema_uris = set())
         print(tax)
         srt_tax: TaxonomySchema = tax.get_taxonomy('http://fasb.org/srt/2020-01-31')
         self.assertTrue(srt_tax)

xbrl/instance.py

@@ -349,18 +349,19 @@ def parse_xbrl(instance_path: str, cache: HttpCache, instance_url: str or None =
     # submissions from SEC normally have their own schema files, whereas submissions from the uk have absolute schemas
 
     # initalise a set that will store cached taxonomy schemas uris to avoid recursive loops
-    imported_schema_uris = set()
+
 
     if is_url(schema_uri):
         # fetch the taxonomy extension schema from remote
-        taxonomy: TaxonomySchema = parse_taxonomy_url(schema_uri, cache, imported_schema_uris)
+        taxonomy: TaxonomySchema = parse_taxonomy_url(schema_uri, cache)
     elif instance_url:
         # fetch the taxonomy extension schema from remote by reconstructing the url
         schema_url = resolve_uri(instance_url, schema_uri)
-        taxonomy: TaxonomySchema = parse_taxonomy_url(schema_url, cache, imported_schema_uris)
+        taxonomy: TaxonomySchema = parse_taxonomy_url(schema_url, cache)
     else:
         # try to find the taxonomy extension schema file locally because no full url can be constructed
         schema_path = resolve_uri(instance_path, schema_uri)
+        imported_schema_uris = set()
         taxonomy: TaxonomySchema = parse_taxonomy(schema_path, cache, imported_schema_uris)
 
     # parse contexts and units
@@ -457,21 +458,26 @@ def parse_ixbrl(instance_path: str, cache: HttpCache, instance_url: str or None
     schema_uri: str = schema_ref.attrib[XLINK_NS + 'href']
     # check if the schema uri is relative or absolute
     # submissions from SEC normally have their own schema files, whereas submissions from the uk have absolute schemas
+
+    # initalise a set that will store cached taxonomy schemas uris to avoid recursive loops
+    imported_schema_uris = set()
+
+
     if is_url(schema_uri):
         # fetch the taxonomy extension schema from remote
-        taxonomy: TaxonomySchema = parse_taxonomy_url(schema_uri, cache)
+        taxonomy: TaxonomySchema = parse_taxonomy_url(schema_uri, cache, imported_schema_uris)
     elif schema_root:
         # take the given schema_root path as directory for searching for the taxonomy schema
         schema_path = str(next(Path(schema_root).glob(f'**/{schema_uri}')))
-        taxonomy: TaxonomySchema = parse_taxonomy(schema_path, cache)
+        taxonomy: TaxonomySchema = parse_taxonomy(schema_path, cache, imported_schema_uris)
     elif instance_url:
         # fetch the taxonomy extension schema from remote by reconstructing the url
         schema_url = resolve_uri(instance_url, schema_uri)
-        taxonomy: TaxonomySchema = parse_taxonomy_url(schema_url, cache)
+        taxonomy: TaxonomySchema = parse_taxonomy_url(schema_url, cache, imported_schema_uris)
     else:
         # try to find the taxonomy extension schema file locally because no full url can be constructed
         schema_path = resolve_uri(instance_path, schema_uri)
-        taxonomy: TaxonomySchema = parse_taxonomy(schema_path, cache)
+        taxonomy: TaxonomySchema = parse_taxonomy(schema_path, cache, imported_schema_uris)
 
     # get all contexts and units
     xbrl_resources: ET.Element = root.find('.//ix:resources', ns_map)

xbrl/taxonomy.py

@@ -588,7 +588,7 @@ def parse_common_taxonomy(cache: HttpCache, namespace: str) -> TaxonomySchema or
 
 
 @lru_cache(maxsize=60)
-def parse_taxonomy_url(schema_url: str, cache: HttpCache, imported_schema_uris : set) -> TaxonomySchema:
+def parse_taxonomy_url(schema_url: str, cache: HttpCache, imported_schema_uris: set = set()) -> TaxonomySchema:
     """
     Parses a taxonomy schema file from the internet
 
@@ -643,17 +643,16 @@ def parse_taxonomy(schema_path: str, cache: HttpCache, imported_schema_uris : se
         if is_url(import_uri):
             # fetch the schema file from remote
             taxonomy.imports.append(parse_taxonomy_url(import_uri, cache))
-            imported_schema_uris.add(import_uri)
         elif schema_url:
             # fetch the schema file from remote by reconstructing the full url
             import_url = resolve_uri(schema_url, import_uri)
-            taxonomy.imports.append(parse_taxonomy_url(import_url, cache))
             imported_schema_uris.add(import_uri)
+            taxonomy.imports.append(parse_taxonomy_url(import_url, cache))
         else:
             # We have to try to fetch the linkbase locally because no full url can be constructed
             import_path = resolve_uri(schema_path, import_uri)
-            taxonomy.imports.append(parse_taxonomy(import_path, cache))
-            imported_schema_uris.add(import_uri)
+            taxonomy.imports.append(parse_taxonomy(import_path, cache, imported_schema_uris))
+
 
     role_type_elements: List[ET.Element] = root.findall('xsd:annotation/xsd:appinfo/link:roleType', NAME_SPACES)
     # parse ELR's