python/sqlglot
1
0
Fork 0
Code Issues Activity

Merging upstream version 10.0.1.

Browse source 
Signed-off-by: Daniel Baumann <daniel@debian.org>
This commit is contained in:
Daniel Baumann 2025-02-13 14:53:05 +01:00
parent 528822bfd4
commit b7d21c45b7
Signed by: daniel
GPG key ID: FBB4F0E80A80222F
98 changed files with 4080 additions and 1666 deletions
Download patch file Download diff file Expand all files Collapse all files

23
sqlglot/diff.py
View file

@ -4,7 +4,7 @@ from heapq import heappop, heappush
from sqlglot import Dialect
from sqlglot import expressions as exp
from sqlglot.helper import ensure_list
from sqlglot.helper import ensure_collection
@dataclass(frozen=True)
@ -116,7 +116,9 @@ class ChangeDistiller:
source_node = self._source_index[kept_source_node_id]
target_node = self._target_index[kept_target_node_id]
if not isinstance(source_node, LEAF_EXPRESSION_TYPES) or source_node == target_node:
edit_script.extend(self._generate_move_edits(source_node, target_node, matching_set))
edit_script.extend(
self._generate_move_edits(source_node, target_node, matching_set)
)
edit_script.append(Keep(source_node, target_node))
else:
edit_script.append(Update(source_node, target_node))
@ -158,13 +160,16 @@ class ChangeDistiller:
max_leaves_num = max(len(source_leaf_ids), len(target_leaf_ids))
if max_leaves_num:
common_leaves_num = sum(
1 if s in source_leaf_ids and t in target_leaf_ids else 0 for s, t in leaves_matching_set
1 if s in source_leaf_ids and t in target_leaf_ids else 0
for s, t in leaves_matching_set
)
leaf_similarity_score = common_leaves_num / max_leaves_num
else:
leaf_similarity_score = 0.0
adjusted_t = self.t if min(len(source_leaf_ids), len(target_leaf_ids)) > 4 else 0.4
adjusted_t = (
self.t if min(len(source_leaf_ids), len(target_leaf_ids)) > 4 else 0.4
)
if leaf_similarity_score >= 0.8 or (
leaf_similarity_score >= adjusted_t
@ -201,7 +206,10 @@ class ChangeDistiller:
matching_set = set()
while candidate_matchings:
_, _, source_leaf, target_leaf = heappop(candidate_matchings)
if id(source_leaf) in self._unmatched_source_nodes and id(target_leaf) in self._unmatched_target_nodes:
if (
id(source_leaf) in self._unmatched_source_nodes
and id(target_leaf) in self._unmatched_target_nodes
):
matching_set.add((id(source_leaf), id(target_leaf)))
self._unmatched_source_nodes.remove(id(source_leaf))
self._unmatched_target_nodes.remove(id(target_leaf))
@ -241,8 +249,7 @@ def _get_leaves(expression):
has_child_exprs = False
for a in expression.args.values():
nodes = ensure_list(a)
for node in nodes:
for node in ensure_collection(a):
if isinstance(node, exp.Expression):
has_child_exprs = True
yield from _get_leaves(node)
@ -268,7 +275,7 @@ def _expression_only_args(expression):
args = []
if expression:
for a in expression.args.values():
args.extend(ensure_list(a))
args.extend(ensure_collection(a))
return [a for a in args if isinstance(a, exp.Expression)]