dbcli/sqlglot
1
0
Fork 0
Code Issues Activity

Merging upstream version 23.13.1.

Browse source 
Signed-off-by: Daniel Baumann <daniel@debian.org>
This commit is contained in:
Daniel Baumann 2025-02-13 21:33:25 +01:00
parent 63a75c51ff
commit 64041d1d66
Signed by: daniel
GPG key ID: FBB4F0E80A80222F
85 changed files with 53899 additions and 50390 deletions
Download patch file Download diff file Expand all files Collapse all files

29
sqlglot/lineage.py
View file

@ -8,6 +8,7 @@ from dataclasses import dataclass, field
from sqlglot import Schema, exp, maybe_parse
from sqlglot.errors import SqlglotError
from sqlglot.optimizer import Scope, build_scope, find_all_in_scope, normalize_identifiers, qualify
from sqlglot.optimizer.scope import ScopeType
if t.TYPE_CHECKING:
from sqlglot.dialects.dialect import DialectType
@ -129,12 +130,6 @@ def to_node(
reference_node_name: t.Optional[str] = None,
trim_selects: bool = True,
) -> Node:
source_names = {
dt.alias: dt.comments[0].split()[1]
for dt in scope.derived_tables
if dt.comments and dt.comments[0].startswith("source: ")
}
# Find the specific select clause that is the source of the column we want.
# This can either be a specific, named select or a generic `*` clause.
select = (
@ -242,13 +237,31 @@ def to_node(
# If the source is a UDTF find columns used in the UTDF to generate the table
if isinstance(source, exp.UDTF):
source_columns |= set(source.find_all(exp.Column))
derived_tables = [
source.expression.parent
for source in scope.sources.values()
if isinstance(source, Scope) and source.is_derived_table
]
else:
derived_tables = scope.derived_tables
source_names = {
dt.alias: dt.comments[0].split()[1]
for dt in derived_tables
if dt.comments and dt.comments[0].startswith("source: ")
}
for c in source_columns:
table = c.table
source = scope.sources.get(table)
if isinstance(source, Scope):
selected_node, _ = scope.selected_sources.get(table, (None, None))
reference_node_name = None
if source.scope_type == ScopeType.DERIVED_TABLE and table not in source_names:
reference_node_name = table
elif source.scope_type == ScopeType.CTE:
selected_node, _ = scope.selected_sources.get(table, (None, None))
reference_node_name = selected_node.name if selected_node else None
# The table itself came from a more specific scope. Recurse into that one using the unaliased column name.
to_node(
c.name,
@ -257,7 +270,7 @@ def to_node(
scope_name=table,
upstream=node,
source_name=source_names.get(table) or source_name,
reference_node_name=selected_node.name if selected_node else None,
reference_node_name=reference_node_name,
trim_selects=trim_selects,
)
else: