identifiers: Also validate the hash is correct

Related T1104

swh/model/identifiers.py

@@ -7,6 +7,8 @@ import binascii
 import datetime
 from functools import lru_cache
 
+from .exceptions import ValidationError
+from .fields.hashes import validate_sha1
 from .hashutil import hash_data, hash_git_data, DEFAULT_ALGORITHMS
 from .hashutil import hash_to_hex
 
@@ -696,6 +698,12 @@ def parse_persistent_identifier(persistent_id):
         raise SWHMalformedIdentifierException(
             'Wrong format: Identifier should be present')
 
+    try:
+        validate_sha1(_id)
+    except ValidationError:
+        raise SWHMalformedIdentifierException(
+           'Wrong format: Identifier should be a valid hash')
+
     persistent_id_metadata = {}
     for part in persistent_id_parts:
         try:

swh/model/tests/test_identifiers.py

@@ -867,7 +867,11 @@ class SnapshotIdentifier(unittest.TestCase):
                      ', '.join(PERSISTENT_IDENTIFIER_TYPES))),
                 ('swh:1:dir:0b6959356d30f1a4e9b7f6bca59b9a336464c03d;invalid;'
                  'malformed',
-                 'Contextual data is badly formatted, form key=val expected')
+                 'Contextual data is badly formatted, form key=val expected'),
+                ('swh:1:snp:gh6959356d30f1a4e9b7f6bca59b9a336464c03d',
+                 'Wrong format: Identifier should be a valid hash'),
+                ('swh:1:snp:foo',
+                 'Wrong format: Identifier should be a valid hash')
         ]:
             with self.assertRaisesRegex(
                     SWHMalformedIdentifierException, _error):