Performance improvement in Row

Various performance improvements to Row instanciation
- avoid passing processors if they are all None
- improve processor logic in cython
- improve tuplegetter using slices when contiguous indexes are used

Some timing follow.

In particular [base_]row_new_proc that tests using processors has
a 25% improvement compared to before in cython.
Looking at the [b]row_new_proc_none that test a list of processors
all None, this has 50% improvement in cython when passing the none list,
but in this patch it would usually be disabled by passing None, so the
performance gain is actually 90%, since it would run the case
[base_]row_new.

Tuplegetter is a bit faster in the single item get and when getting
sequential indexes (like indexes 1,2,3,4) at the cost of a bit
longer creation time in python, cython is mostly the same.

Current times
                    | python      | cython      | cy / py     |
base_row_new        | 0.639817400 | 0.118265500 | 0.184842582 |
row_new             | 0.680355100 | 0.129714600 | 0.190657202 |
base_row_new_proc   | 3.076538900 | 1.488428600 | 0.483799701 |
row_new_proc        | 3.119700100 | 1.532197500 | 0.491136151 |
brow_new_proc_none  | 1.917702300 | 0.475511500 | 0.247958977 |
row_new_proc_none   | 1.956253300 | 0.497803100 | 0.254467609 |

tuplegetter_one     | 0.152512600 | 0.148523900 | 0.973846751 |
tuplegetter_many    | 0.184394100 | 0.184511500 | 1.000636680 |
tuplegetter_seq     | 0.154832800 | 0.156270100 | 1.009282917 |
tuplegetter_new_one | 0.523730000 | 0.343402200 | 0.655685563 |
tuplegetter_new_many| 0.738924400 | 0.420961400 | 0.569694816 |
tuplegetter_new_seq | 1.062036900 | 0.495462000 | 0.466520514 |

Parent commit times
                    | python      | cython      | cy / py     |
base_row_new        | 0.643890800 | 0.113548300 | 0.176347138 |
row_new             | 0.674885900 | 0.124391800 | 0.184315304 |
base_row_new_proc   | 3.072020400 | 2.017367000 | 0.656690626 |
row_new_proc        | 3.109943400 | 2.048359400 | 0.658648450 |
brow_new_proc_none  | 1.967133700 | 1.006326000 | 0.511569702 |
row_new_proc_none   | 1.960814900 | 1.025217800 | 0.522852922 |

tuplegetter_one     | 0.197359900 | 0.205999000 | 1.043773330 |
tuplegetter_many    | 0.196575900 | 0.194888500 | 0.991416038 |
tuplegetter_seq     | 0.192723900 | 0.205635000 | 1.066992729 |
tuplegetter_new_one | 0.534644500 | 0.414311700 | 0.774929322 |
tuplegetter_new_many| 0.479376500 | 0.417448100 | 0.870814694 |
tuplegetter_new_seq | 0.481580200 | 0.412697900 | 0.856966088 |

Change-Id: I2ca1f49dca2beff625c283f1363c29c8ccc0c3f7

1 files changed, 14 insertions, 15 deletions

diff --git a/lib/sqlalchemy/engine/result.py b/lib/sqlalchemy/engine/result.py
index cc6d26c88..cf34c195a 100644
--- a/lib/sqlalchemy/engine/result.py
+++ b/lib/sqlalchemy/engine/result.py
@@ -41,6 +41,7 @@ from ..sql.base import _generative
 from ..sql.base import HasMemoized
 from ..sql.base import InPlaceGenerative
 from ..util import HasMemoized_ro_memoized_attribute
+from ..util import NONE_SET
 from ..util._has_cy import HAS_CYEXTENSION
 from ..util.typing import Literal
 from ..util.typing import Self
@@ -84,7 +85,7 @@ across all the result types
 _InterimSupportsScalarsRowType = Union[Row, Any]
 
 _ProcessorsType = Sequence[Optional["_ResultProcessorType[Any]"]]
-_TupleGetterType = Callable[[Sequence[Any]], Tuple[Any, ...]]
+_TupleGetterType = Callable[[Sequence[Any]], Sequence[Any]]
 _UniqueFilterType = Callable[[Any], Any]
 _UniqueFilterStateType = Tuple[Set[Any], Optional[_UniqueFilterType]]
 
@@ -205,6 +206,13 @@ class ResultMetaData:
             else:
                 self._key_fallback(key, None)
 
+    @property
+    def _effective_processors(self) -> Optional[_ProcessorsType]:
+        if not self._processors or NONE_SET.issuperset(self._processors):
+            return None
+        else:
+            return self._processors
+
 
 class RMKeyView(typing.KeysView[Any]):
     __slots__ = ("_parent", "_keys")
@@ -390,7 +398,7 @@ def result_tuple(
 ) -> Callable[[Iterable[Any]], Row[Any]]:
     parent = SimpleResultMetaData(fields, extra)
     return functools.partial(
-        Row, parent, parent._processors, parent._key_to_index
+        Row, parent, parent._effective_processors, parent._key_to_index
     )
 
 
@@ -454,7 +462,7 @@ class ResultInternal(InPlaceGenerative, Generic[_R]):
 
                 def process_row(  # type: ignore
                     metadata: ResultMetaData,
-                    processors: _ProcessorsType,
+                    processors: Optional[_ProcessorsType],
                     key_to_index: Mapping[_KeyType, int],
                     scalar_obj: Any,
                 ) -> Row[Any]:
@@ -468,7 +476,7 @@ class ResultInternal(InPlaceGenerative, Generic[_R]):
         metadata = self._metadata
 
         key_to_index = metadata._key_to_index
-        processors = metadata._processors
+        processors = metadata._effective_processors
         tf = metadata._tuplefilter
 
         if tf and not real_result._source_supports_scalars:
@@ -489,21 +497,12 @@ class ResultInternal(InPlaceGenerative, Generic[_R]):
                 process_row, metadata, processors, key_to_index
             )
 
-        fns: Tuple[Any, ...] = ()
-
         if real_result._row_logging_fn:
-            fns = (real_result._row_logging_fn,)
-        else:
-            fns = ()
-
-        if fns:
+            _log_row = real_result._row_logging_fn
             _make_row = make_row
 
             def make_row(row: _InterimRowType[Row[Any]]) -> _R:
-                interim_row = _make_row(row)
-                for fn in fns:
-                    interim_row = fn(interim_row)
-                return interim_row  # type: ignore
+                return _log_row(_make_row(row))  # type: ignore
 
         return make_row