Avoid string types when using ClassVar

As of Python 3.9.8, this does not work with Pydantic BaseModel, and
does not work in a way that causes the whole module to fail to
import.  See pydantic/pydantic#3401.

In general, Pydantic does not work well with string types as enabled
by from __future__ import annotations, which is why making that
feature the default has been delayed.  See
https://lwn.net/Articles/858576/ for more information.

python/lsst/daf/butler/core/logging.py

@@ -19,8 +19,6 @@
 # You should have received a copy of the GNU General Public License
 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
 
-from __future__ import annotations
-
 __all__ = ("ButlerMDC", "ButlerLogRecords", "ButlerLogRecordHandler",
            "ButlerLogRecord", "JsonLogFormatter")
 
@@ -192,7 +190,7 @@ class Config:
         allow_mutation = False
 
     @classmethod
-    def from_record(cls, record: LogRecord) -> ButlerLogRecord:
+    def from_record(cls, record: LogRecord) -> "ButlerLogRecord":
         """Create a new instance from a `~logging.LogRecord`.
 
         Parameters

python/lsst/daf/butler/core/serverModels.py

@@ -19,8 +19,6 @@
 # You should have received a copy of the GNU General Public License
 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
 
-from __future__ import annotations
-
 __all__ = (
     "QueryDatasetsModel",
     "QueryDataIdsModel",