Skip to content

fix(gfql): preserve quoted ORDER BY aliases #2208

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
lmeyerov merged 1 commit into from
Oct 21, 2026
Merged

fix(gfql): preserve quoted ORDER BY aliases #2208

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 3 additions & 3 deletions .github/workflows/ci.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1760,9 +1760,9 @@ jobs:
needs: [changes, test-minimal-python, generate-lockfiles]
if: ${{ ((needs.changes.outputs.python == 'true' && needs.changes.outputs.narrow_python_only != 'true') || needs.changes.outputs.gfql == 'true' || needs.changes.outputs.pandas_compat == 'true' || needs.changes.outputs.core == 'true' || needs.changes.outputs.infra == 'true' || github.event_name == 'workflow_dispatch' || github.event_name == 'schedule') && !(needs.changes.outputs.docs_only_latest == 'true' && (github.event_name == 'push' || github.event_name == 'pull_request')) }}
runs-on: ubuntu-latest
# The py3.12 coverage cell can exceed 15 minutes including setup and
# artifact work; retain enough margin for a completed test verdict.
timeout-minutes: 20
# The py3.12 coverage cell can exceed 20 minutes including setup,
# two test phases, and coverage artifact work.
timeout-minutes: 30

strategy:
matrix:
Expand Down
1 change: 1 addition & 0 deletions CHANGELOG.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,6 +12,7 @@ This project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.htm

- **Chained `compute_igraph` calls preserve custom-named edge attributes**: edge attributes named `source` or `target` are restored to their original column names after a round trip when endpoint bindings use other names, so repeated igraph calls no longer accumulate collision renames or fail on duplicate columns. Attributes that collide with the final endpoint bindings are moved to unique generated names rather than merging with endpoint columns (#2204).

- **GFQL Cypher quoted ORDER BY aliases**: quoted aliases containing spaces or escaped backticks can now order values in pandas, Polars, and cuDF lowering, including aggregate outputs. Quoted output names keep their own binding identity when their text matches another aggregate expression, and unresolved quoted aliases retain structured validation errors (#2205).

## [0.61.0] - 2026-10-18

Expand Down
67 changes: 57 additions & 10 deletions graphistry/compute/gfql/cypher/lowering.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -169,7 +169,7 @@
from graphistry.compute.gfql.cypher.reentry import naming as _reentry_naming, scope as _reentry_scope
from graphistry.compute.gfql.cypher.ast import CypherParams
from graphistry.compute.gfql.cypher.where_membership import peel_literal_membership_where
from graphistry.compute.gfql.identifiers import render_cypher_identifier, shortest_path_hops_column
from graphistry.compute.gfql.identifiers import parse_cypher_backtick_identifier, render_cypher_identifier, shortest_path_hops_column


@dataclass(frozen=True)
Expand Down Expand Up @@ -1245,6 +1245,7 @@ def _rewrite(node_in: ExprNode) -> ExprNode:
def _rewrite_expr_to_output_names(
expr_text: str,
*,
available_columns: AbstractSet[str],
replacements: Mapping[str, str],
params: Optional[Mapping[str, Any]],
field: str,
Expand All @@ -1260,12 +1261,16 @@ def _rewrite_expr_to_output_names(
)

def _rewrite(node_in: ExprNode) -> ExprNode:
if isinstance(node_in, Identifier):
if node_in.name in available_columns:
return Identifier(node_in.name, quoted=node_in.quoted)
replacement = replacements.get(node_in.name)
if replacement is not None:
return Identifier(replacement, quoted=node_in.quoted)
rendered = _render_expr_node(node_in)
replacement = replacements.get(rendered)
if replacement is not None:
return Identifier(replacement)
if isinstance(node_in, Identifier):
return Identifier(replacements.get(node_in.name, node_in.name))
return Identifier(replacement, quoted=True)
return _rebuild_expr_node(node_in, rewrite=_rewrite, error_context="output rewrite")

return _render_expr_node(_rewrite(node))
Expand Down Expand Up @@ -4115,6 +4120,10 @@ def _lower_order_by_clause(
)
else:
order_key = alias_name
if order_key == alias_name and not (
plan.whole_row_output_names and alias_name in plan.output_to_expr_source
):
order_key = render_cypher_identifier(order_key)
else:
source_alias = (
plan.whole_row_sources.get(alias_name, plan.source_alias)
Expand Down Expand Up @@ -4223,17 +4232,27 @@ def _lower_order_by_outputs(
) -> ASTObject:
keys: List[Tuple[str, str]] = []
for item in clause.items:
order_key = expr_to_output.get(item.expression.text)
order_node = _parse_row_expr(item.expression.text, params=params, field="order_by", line=item.span.line, column=item.span.column)
quoted_alias = order_node.name if isinstance(order_node, Identifier) and order_node.quoted else None
if isinstance(order_node, Identifier) and order_node.name in available_columns:
keys.append((render_cypher_identifier(order_node.name), item.direction))
continue
order_key: Optional[str] = expr_to_output.get(item.expression.text)
if order_key is None and quoted_alias is not None and quoted_alias in available_columns:
order_key = render_cypher_identifier(quoted_alias)
if order_key is None:
order_key = _rewrite_expr_to_output_names(
item.expression.text,
available_columns=available_columns,
replacements=expr_to_output,
params=params,
field="order_by",
line=item.span.line,
column=item.span.column,
)
_reject_order_by_aggregate(item, expr_text=order_key, params=params)
if order_key != item.expression.text and order_key in available_columns:
order_key = render_cypher_identifier(order_key)
if order_key not in available_columns:
node = _parse_row_expr(
order_key,
Expand Down Expand Up @@ -4913,6 +4932,35 @@ def _lower_match_alias_stage(
)
row_steps: List[ASTObject] = []
defer_return_projection = False
order_projection_items = list(plan.projection_items)
order_plan: _ProjectionPlan = plan
if stage.order_by is not None:
order_plan = _projection._plan_with_visible_projected_columns(plan, scope.projected_columns)
order_output_names = _projection._projection_output_names(plan)
for order_item in stage.order_by.items:
try:
order_alias, order_prop = _projection._projection_ref_from_expr(
order_item.expression.text,
alias_targets=scope.alias_targets,
params=params,
field="order_by",
line=order_item.span.line,
column=order_item.span.column,
)
except GFQLValidationError:
continue
if (
order_prop is None
and order_alias is not None
and order_alias not in order_output_names
and order_alias in order_plan.available_columns
and order_alias in scope.projected_columns
):
order_binding = scope.projected_columns[order_alias]
if order_binding.source_name is not None:
order_projection_items.append((order_alias, order_binding.source_name))
order_output_names.add(order_alias)
order_plan = replace(order_plan, projection_items=order_projection_items)
if not plan.whole_row_output_names:
projection_fn = with_ if stage.clause.kind == "with" else return_
if (
Expand All @@ -4922,7 +4970,7 @@ def _lower_match_alias_stage(
):
# Keep pre-existing columns alive for ORDER BY references that are
# intentionally not part of the final RETURN projection.
row_steps.append(with_(plan.projection_items, extend=True))
row_steps.append(with_(order_projection_items, extend=True))
defer_return_projection = True
else:
row_steps.append(projection_fn(plan.projection_items))
Expand Down Expand Up @@ -4971,7 +5019,6 @@ def _lower_match_alias_stage(
)
)
if stage.order_by is not None:
order_plan = _projection._plan_with_visible_projected_columns(plan, scope.projected_columns)
row_steps.append(
_lower_order_by_clause(
stage.order_by,
Expand Down Expand Up @@ -5245,7 +5292,7 @@ def _lower_match_alias_aggregate_stage(
expr_to_output,
source_expr=agg_spec.source_text,
output_name=agg_spec.output_name,
alias_name=agg_spec.output_name,
alias_name=None,
)

bindings_row_path = bool(scope.allowed_match_aliases)
Expand Down Expand Up @@ -5817,7 +5864,7 @@ def _lower_row_column_aggregate_stage(
expr_to_output,
source_expr=agg_spec.source_text,
output_name=agg_spec.output_name,
alias_name=agg_spec.output_name,
alias_name=None,
)

row_steps: List[ASTObject] = []
Expand Down Expand Up @@ -7470,7 +7517,7 @@ def _lower_general_row_projection(
expr_to_output,
source_expr=agg_spec.source_text,
output_name=agg_spec.output_name,
alias_name=agg_spec.output_name,
alias_name=None,
)

if not binding_row_aliases:
Expand Down
Loading
Loading