Creating and dropping a macro in same txn triggers assertion #744
Description
What happens?
When creating and dropping a macro within a transaction, an assertion error is thrown (on a debug build). I didn't see any tests that tested that exact combination of events.
I am not sure how common this would be. One scenario that I could see it being useful is if you wanted to add or customize a macro just for a specific operation and not interfere with other ongoing operations. (For example, add a macro just for sorting a specific way during an inline flush, then removing that macro before committing).
To Reproduce
This test case will trigger the error in a debug build.
require ducklake
require parquet
test-env DUCKLAKE_CONNECTION __TEST_DIR__/{UUID}.db
statement ok
ATTACH 'ducklake:${DUCKLAKE_CONNECTION}' AS ducklake (DATA_PATH '__TEST_DIR__/create_then_drop_macro/')
statement ok
USE ducklake;
statement ok
BEGIN
statement ok
CREATE MACRO zip_varchar(i, j, num_chars := 6) AS (
-- By default using 6 characters from each string so that
-- if data is ASCII, we can fit it all in 12 bytes so that it is stored inline
-- rather than requiring a pointer
[
list_value(z[1], z[2])
FOR z
IN list_zip(
substr(i, 1, num_chars).rpad(num_chars, ' ').string_split(''),
substr(j, 1, num_chars).rpad(num_chars, ' ').string_split('')
)
].flatten().array_to_string('')
);
# Test that the macro is working as intended independently of sorting
query III
SELECT
'ABC' AS i,
'XYZ' AS j,
zip_varchar(i, j, num_chars := 3) AS zipped_varchar;
----
ABC XYZ AXBYCZ
statement ok
DROP MACRO zip_varchar
# Error happens here
statement ok
COMMIT
OS:
MacOS
DuckDB Version:
1.5
DuckLake Version:
DuckDB Client:
CLI
Hardware:
No response
Full Name:
Alex Monahan
Affiliation:
MotherDuck
What is the latest build you tested with? If possible, we recommend testing with the latest nightly build.
I have tested with a source build
Did you include all relevant data sets for reproducing the issue?
Yes
Did you include all code required to reproduce the issue?
- Yes, I have
Did you include all relevant configuration (e.g., CPU architecture, Python version, Linux distribution) to reproduce the issue?
- Yes, I have