mabel-dev · joocer · Nov 20, 2023 · Nov 20, 2023 · Nov 20, 2023
diff --git a/.github/workflows/regression_suite.yaml b/.github/workflows/regression_suite.yaml
@@ -7,8 +7,9 @@ jobs:
   regression_matrix:
     strategy:
       max-parallel: 4
+      fail-fast: false
       matrix:
-        python-version: ['3.9', '3.10', '3.11']
+        python-version: ['3.9', '3.10', '3.11', '3.12']
         os: [ubuntu-latest]
     runs-on: ${{ matrix.os }}
     steps:

diff --git a/Makefile b/Makefile
@@ -1,5 +1,6 @@
 lint:
-	python -m pip install --quiet --upgrade pycln isort black
+	python -m pip install --quiet --upgrade pycln isort black yamllint
+#	python -m yamllint .
 	python -m pycln .
 	python -m isort .
 	python -m black .

diff --git a/opteryx/components/heuristic_optimizer.py b/opteryx/components/heuristic_optimizer.py
@@ -119,6 +119,8 @@ def visit(self, parent: str, nid: str, context: HeuristicOptimizerContext):
             nodes = heuristic_optimizer.rule_split_conjunctive_predicates(node)
             # deduplicate the nodes - note this 'randomizes' the order
             nodes = _unique_nodes(nodes)
+            # order the predicates
+            nodes = heuristic_optimizer.rule_order_predicates(nodes)
 
             previous = parent
             for predicate_node in nodes:

diff --git a/opteryx/components/rules/heuristic_optimizer/__init__.py b/opteryx/components/rules/heuristic_optimizer/__init__.py
@@ -1 +1,2 @@
+from .order_predicates import rule_order_predicates
 from .split_conjuctive_predicates import rule_split_conjunctive_predicates
diff --git a/...euristic_optimizer/apply_demorgans_law.py → ...ic_optimizer/bench/apply_demorgans_law.py b/...euristic_optimizer/apply_demorgans_law.py → ...ic_optimizer/bench/apply_demorgans_law.py
diff --git a/...uristic_optimizer/constant_evaluations.py → ...c_optimizer/bench/constant_evaluations.py b/...uristic_optimizer/constant_evaluations.py → ...c_optimizer/bench/constant_evaluations.py
diff --git a/...optimizer/move_literal_join_conditions.py → ...zer/bench/move_literal_join_conditions.py b/...optimizer/move_literal_join_conditions.py → ...zer/bench/move_literal_join_conditions.py
diff --git a/...es/heuristic_optimizer/push_predicates.py → ...ristic_optimizer/bench/push_predicates.py b/...es/heuristic_optimizer/push_predicates.py → ...ristic_optimizer/bench/push_predicates.py
diff --git a/...s/heuristic_optimizer/reduce_negations.py → ...istic_optimizer/bench/reduce_negations.py b/...s/heuristic_optimizer/reduce_negations.py → ...istic_optimizer/bench/reduce_negations.py
diff --git a/...ules/heuristic_optimizer/use_heap_sort.py → ...euristic_optimizer/bench/use_heap_sort.py b/...ules/heuristic_optimizer/use_heap_sort.py → ...euristic_optimizer/bench/use_heap_sort.py
diff --git a/opteryx/components/rules/heuristic_optimizer/order_predicates.py b/opteryx/components/rules/heuristic_optimizer/order_predicates.py
@@ -0,0 +1,63 @@
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+"""
+Optimization Rule - Order Predicates
+
+Type: Heuristic
+Goal: Reduce rows
+"""
+
+from opteryx.managers.expression import NodeType
+
+NODE_ORDER = {
+    "Eq": 1,
+    "NotEq": 1,
+    "Gt": 2,
+    "GtEq": 2,
+    "Lt": 2,
+    "LtEq": 2,
+    "Like": 4,
+    "ILike": 4,
+    "NotLike": 4,
+    "NotILike": 4,
+}
+
+
+def rule_order_predicates(nodes):
+    """
+    Ordering of predicates based on rules, this is mostly useful for situations where
+    we do not have statistics to make cost-based decisions.
+
+    We're going to start with arbitrary numbers, we need to find a way to refine these
+    over time.
+    """
+
+    for node in nodes:
+        if not node.node_type == NodeType.COMPARISON_OPERATOR:
+            node.score = 25
+            continue
+        node.score = NODE_ORDER.get(node.value, 12)
+        if node.left.node_type == NodeType.LITERAL:
+            node.score += 1
+        elif node.left.node_type == NodeType.IDENTIFIER:
+            node.score += 3
+        else:
+            node.score += 10
+        if node.right.node_type == NodeType.LITERAL:
+            node.score += 1
+        elif node.right.node_type == NodeType.IDENTIFIER:
+            node.score += 3
+        else:
+            node.score += 10
+
+    return sorted(nodes, key=lambda node: node.score)
Original file line number	Diff line number	Diff line change
		@@ -1 +1,2 @@
		from .order_predicates import rule_order_predicates
		from .split_conjuctive_predicates import rule_split_conjunctive_predicates