Skip to content

fix: Derived where() on non-selected columns derived wrong results #107

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
jsam merged 1 commit into from
Aug 20, 2026
+30 −6
Merged

fix: Derived where() on non-selected columns derived wrong results #107

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
17 changes: 11 additions & 6 deletions packages/inputlayer-py/src/inputlayer/compiler.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -666,6 +666,17 @@ def compile_rule(
else:
head_parts.append(column_to_variable(col))

# Compile filter conditions BEFORE building body atoms, so every column
# the condition references is registered in the env and gets bound to a
# variable in its atom. Compiling them after produced atoms with `_` for
# condition-only columns while the condition referenced an unbound
# variable - which the engine accepts and silently satisfies, deriving
# wrong results (e.g. a tier == "gold" filter matching every row).
cond_parts: list[str] = []
if condition:
cond_parts = compile_bool_expr(condition, env)
cond_parts = [p for p in cond_parts if p]

# Build body atoms
body_atoms: list[str] = []
for rn, cls, alias in body_relations:
Expand All @@ -680,12 +691,6 @@ def compile_rule(
atom_parts.append("_")
body_atoms.append(f"{rn}({', '.join(atom_parts)})")

# Compile filter conditions
cond_parts: list[str] = []
if condition:
cond_parts = compile_bool_expr(condition, env)
cond_parts = [p for p in cond_parts if p]

all_body = body_atoms + cond_parts
prefix = "+" if persistent else ""
head_str = f"{prefix}{head_name}({', '.join(head_parts)})"
Expand Down
19 changes: 19 additions & 0 deletions packages/inputlayer-py/tests/test_compiler.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -494,6 +494,25 @@ def test_with_condition(self):
)
assert "+high_earner(Id, Name)" in result
assert "Salary > 100000" in result
# The condition column must be BOUND in the body atom. Emitting
# employee(Id, Name, _, _, _) alongside `Salary > 100000` leaves
# Salary unbound - the engine accepts that and silently satisfies
# the comparison, deriving wrong results.
assert "employee(Id, Name, _, Salary, _)" in result

def test_condition_only_column_is_bound(self):
# Regression: a column referenced ONLY by the where-condition (not
# selected into the head) must still be bound in its atom.
cond = Comparison("=", AstColumn("employee", "department"), Literal("eng"))
result = compile_rule(
"eng_member",
["name"],
{"name": AstColumn("employee", "name")},
[(Employee._resolve_name(), Employee, None)],
condition=cond,
persistent=True,
)
assert result == '+eng_member(Name) <- employee(_, Name, Department, _, _), Department = "eng"'

def test_recursive(self):
# reachable(Src, Dst) <- reachable(Src, Mid), edge(Mid, Dst)
Expand Down
Loading