Skip to content

A reactive notebook for Python — run reproducible experiments, query with SQL, execute as a script, deploy as an app, and version with git. Stored as pure Python. All in a modern, AI-native editor.

fix sql cells referencing itself (#7000)

Shahmir Varqha345602095ace
diff --git a/marimo/_runtime/dataflow.py b/marimo/_runtime/dataflow.py
index bd80e06979..769fd5c7b0 100644
--- a/marimo/_runtime/dataflow.py
+++ b/marimo/_runtime/dataflow.py
@@ -331,11 +331,19 @@ class DirectedGraph:
if sql_ref:
sql_matches = self._find_sql_hierarchical_matches(sql_ref)
for matching_cell_ids, _ in sql_matches:
+ if cell_id in matching_cell_ids:
+ LOGGER.debug(
+ "Cell %s is referencing itself", cell_id
+ )
+ continue
other_ids_defining_name.update(matching_cell_ids)
# If other_ids_defining_name is empty, the user will get a
# NameError at runtime (unless the symbol is a builtin).
for other_id in other_ids_defining_name:
+ if other_id == cell_id:
+ LOGGER.error("Cell %s is referencing itself", cell_id)
+ continue
if not self._is_valid_cell_reference(
other_id, variable_name
):
diff --git a/tests/_runtime/test_dataflow_cases.py b/tests/_runtime/test_dataflow_cases.py
index 5e8a85a4f2..997fda0a70 100644
--- a/tests/_runtime/test_dataflow_cases.py
+++ b/tests/_runtime/test_dataflow_cases.py
@@ -484,6 +484,104 @@ SQL_CASES = [
},
expected_defs={"0": ["my_db"], "1": ["my_table"], "2": [], "3": []},
),
+ GraphTestCase(
+ name="create table with the same name from a different schema",
+ enabled=HAS_DUCKDB,
+ code={
+ "0": "_ = mo.sql(f'CREATE TABLE my_table AS SELECT * FROM schema_one.my_table')",
+ "1": "_ = mo.sql(f'SELECT * FROM my_table')",
+ },
+ expected_parents={"0": [], "1": ["0"]},
+ expected_children={"0": ["1"], "1": []},
+ expected_refs={
+ "0": ["mo", "schema_one.my_table"],
+ "1": ["mo", "my_table"],
+ },
+ expected_defs={"0": ["my_table"], "1": []},
+ ),
+ GraphTestCase(
+ name="create table with the same name from catalog.schema hierarchy",
+ enabled=HAS_DUCKDB,
+ code={
+ "0": "_ = mo.sql(f'CREATE TABLE my_table AS SELECT * FROM catalog_one.schema_one.my_table')",
+ "1": "_ = mo.sql(f'SELECT * FROM my_table')",
+ },
+ expected_parents={"0": [], "1": ["0"]},
+ expected_children={"0": ["1"], "1": []},
+ expected_refs={
+ "0": ["mo", "catalog_one.schema_one.my_table"],
+ "1": ["mo", "my_table"],
+ },
+ expected_defs={"0": ["my_table"], "1": []},
+ ),
+ GraphTestCase(
+ name="create schema with the same name from a different catalog",
+ enabled=HAS_DUCKDB,
+ code={
+ "0": "_ = mo.sql(f'CREATE SCHEMA my_schema')",
+ "1": "_ = mo.sql(f'CREATE TABLE my_schema.my_table AS SELECT * FROM catalog_one.my_schema.my_table')",
+ "2": "_ = mo.sql(f'SELECT * FROM my_schema.my_table')",
+ },
+ expected_parents={"0": [], "1": [], "2": ["1"]},
+ expected_children={"0": [], "1": ["2"], "2": []},
+ expected_refs={
+ "0": ["mo"],
+ "1": ["mo", "catalog_one.my_schema.my_table"],
+ "2": ["mo", "my_schema.my_table"],
+ },
+ expected_defs={"0": ["my_schema"], "1": ["my_table"], "2": []},
+ ),
+ GraphTestCase(
+ name="create table that references itself in join prevents self-loop",
+ enabled=HAS_DUCKDB,
+ code={
+ "0": "_ = mo.sql(f'CREATE TABLE users AS SELECT 1 as id')",
+ "1": "_ = mo.sql(f'CREATE TABLE orders AS SELECT u.id FROM schema_one.users u JOIN schema_two.orders o ON u.id = o.user_id')",
+ "2": "_ = mo.sql(f'SELECT * FROM orders')",
+ },
+ expected_parents={"0": [], "1": ["0"], "2": ["1"]},
+ expected_children={"0": ["1"], "1": ["2"], "2": []},
+ expected_refs={
+ "0": ["mo"],
+ "1": ["mo", "schema_one.users", "schema_two.orders"],
+ "2": ["mo", "orders"],
+ },
+ expected_defs={"0": ["users"], "1": ["orders"], "2": []},
+ ),
+ GraphTestCase(
+ name="multiple tables with hierarchical self-reference patterns",
+ enabled=HAS_DUCKDB,
+ code={
+ "0": "_ = mo.sql(f'CREATE TABLE table_a AS SELECT * FROM db1.table_a')",
+ "1": "_ = mo.sql(f'CREATE TABLE table_b AS SELECT * FROM db2.table_b')",
+ "2": "_ = mo.sql(f'SELECT * FROM table_a UNION ALL SELECT * FROM table_b')",
+ },
+ expected_parents={"0": [], "1": [], "2": ["0", "1"]},
+ expected_children={"0": ["2"], "1": ["2"], "2": []},
+ expected_refs={
+ "0": ["mo", "db1.table_a"],
+ "1": ["mo", "db2.table_b"],
+ "2": ["mo", "table_a", "table_b"],
+ },
+ expected_defs={"0": ["table_a"], "1": ["table_b"], "2": []},
+ ),
+ GraphTestCase(
+ name="create table from hierarchical ref then reference it hierarchically",
+ enabled=HAS_DUCKDB,
+ code={
+ "0": "_ = mo.sql(f'CREATE SCHEMA my_schema')",
+ "1": "_ = mo.sql(f'CREATE TABLE my_schema.data AS SELECT * FROM external.my_schema.data')",
+ "2": "_ = mo.sql(f'SELECT * FROM my_schema.data')",
+ },
+ expected_parents={"0": [], "1": [], "2": ["1"]},
+ expected_children={"0": [], "1": ["2"], "2": []},
+ expected_refs={
+ "0": ["mo"],
+ "1": ["mo", "external.my_schema.data"],
+ "2": ["mo", "my_schema.data"],
+ },
+ expected_defs={"0": ["my_schema"], "1": ["data"], "2": []},
+ ),
GraphTestCase(
name="sql table multiple definitions, different order",
enabled=HAS_DUCKDB,