Merge pull request #1412 from datajoint/fix/index-attribute-regex

fix: allow attribute names starting with 'index' in declarations

Author: dimitri-yatsenko

src/datajoint/builtin_codecs/attach.py

@@ -106,7 +106,8 @@ def decode(self, stored: bytes, *, key: dict | None = None) -> str:
         # Write to download path
         config = (key or {}).get("_config")
         if config is None:
-            from ..settings import config
+            from ..settings import config  # type: ignore[assignment]
+        assert config is not None
         download_path = Path(config.get("download_path", "."))
         download_path.mkdir(parents=True, exist_ok=True)
         local_path = download_path / filename

src/datajoint/builtin_codecs/filepath.py

@@ -102,7 +102,8 @@ def encode(self, value: Any, *, key: dict | None = None, store_name: str | None
 
         config = (key or {}).get("_config")
         if config is None:
-            from ..settings import config
+            from ..settings import config  # type: ignore[assignment]
+        assert config is not None
 
         path = str(value)
 

src/datajoint/declare.py

@@ -370,7 +370,7 @@ def prepare_declare(
                 adapter,
                 fk_attribute_map,
             )
-        elif re.match(r"^(unique\s+)?index\s*.*$", line, re.I):  # index
+        elif re.match(r"^(unique\s+)?index\s*\(.*\)$", line, re.I):  # index
             compile_index(line, index_sql, adapter)
         else:
             name, sql, store, comment = compile_attribute(line, in_key, foreign_key_sql, context, adapter)

src/datajoint/hash_registry.py

@@ -32,14 +32,19 @@
 datajoint.gc : Garbage collection for orphaned storage items.
 """
 
+from __future__ import annotations
+
 import base64
 import hashlib
 import logging
-from typing import Any
+from typing import TYPE_CHECKING, Any
 
 from .errors import DataJointError
 from .storage import StorageBackend
 
+if TYPE_CHECKING:
+    from .settings import Config
+
 logger = logging.getLogger(__name__.split(".")[0])
 
 
@@ -130,7 +135,7 @@ def build_hash_path(
         return f"_hash/{schema_name}/{content_hash}"
 
 
-def get_store_backend(store_name: str | None = None, config=None) -> StorageBackend:
+def get_store_backend(store_name: str | None = None, config: Config | None = None) -> StorageBackend:
     """
     Get a StorageBackend for hash-addressed storage.
 
@@ -147,13 +152,14 @@ def get_store_backend(store_name: str | None = None, config=None) -> StorageBack
         StorageBackend instance.
     """
     if config is None:
-        from .settings import config
+        from .settings import config  # type: ignore[assignment]
+    assert config is not None
     # get_store_spec handles None by using stores.default
     spec = config.get_store_spec(store_name)
     return StorageBackend(spec)
 
 
-def get_store_subfolding(store_name: str | None = None, config=None) -> tuple[int, ...] | None:
+def get_store_subfolding(store_name: str | None = None, config: Config | None = None) -> tuple[int, ...] | None:
     """
     Get the subfolding configuration for a store.
 
@@ -170,7 +176,8 @@ def get_store_subfolding(store_name: str | None = None, config=None) -> tuple[in
         Subfolding pattern (e.g., (2, 2)) or None for flat storage.
     """
     if config is None:
-        from .settings import config
+        from .settings import config  # type: ignore[assignment]
+    assert config is not None
     spec = config.get_store_spec(store_name)
     subfolding = spec.get("subfolding")
     if subfolding is not None:
@@ -182,7 +189,7 @@ def put_hash(
     data: bytes,
     schema_name: str,
     store_name: str | None = None,
-    config=None,
+    config: Config | None = None,
 ) -> dict[str, Any]:
     """
     Store content using hash-addressed storage.
@@ -231,7 +238,7 @@ def put_hash(
     }
 
 
-def get_hash(metadata: dict[str, Any], config=None) -> bytes:
+def get_hash(metadata: dict[str, Any], config: Config | None = None) -> bytes:
     """
     Retrieve content using stored metadata.
 
@@ -275,7 +282,7 @@ def get_hash(metadata: dict[str, Any], config=None) -> bytes:
 def delete_path(
     path: str,
     store_name: str | None = None,
-    config=None,
+    config: Config | None = None,
 ) -> bool:
     """
     Delete content at the specified path from storage.

tests/integration/test_declare.py

@@ -339,14 +339,20 @@ class WithSuchALongPartNameThatItCrashesMySQL(dj.Part):
         schema_any(WhyWouldAnyoneCreateATableNameThisLong)
 
 
-def test_regex_mismatch(schema_any):
+def test_index_attribute_name(schema_any):
+    """Attributes named 'index' should not be misclassified as index declarations (#1411)."""
+
     class IndexAttribute(dj.Manual):
         definition = """
-        index: int
+        index : int
+        ---
+        index_value : float
         """
 
-    with pytest.raises(dj.DataJointError):
-        schema_any(IndexAttribute)
+    schema_any(IndexAttribute)
+    assert "index" in IndexAttribute.heading.attributes
+    assert "index_value" in IndexAttribute.heading.attributes
+    IndexAttribute.drop()
 
 
 def test_table_name_with_underscores(schema_any):