Fixed Issue #3325

[Hydrocharged]

Fixes #3325.

Stacked on #3348.

[itoqa]

Commit: 6ad7bee: 13 test cases ran, 13 passed ✅.

Summary

Coverage spans fixed-width character handling across comparisons, ordering, casting, formatting, joins, grouping, indexing, uniqueness, concurrency, and null safety, along with contrasting variable-length text behavior. It exercises normal flows, boundary cases involving padding and meaningful spaces, and adversarial cases such as concurrent inserts and null inputs, with healthy results throughout.

Safe to merge — the run found no regressions, new failures, or previously flagged failures attributable to this PR, and the broad behavior coverage passed. No merge-blocking risk was identified.

Tests run by Ito

Result	Severity	Type	Description
✅	—	General	Empty, single-space, and repeated-space endings compare as equal, while values with different non-space content remain distinct and ordered correctly.
✅	—	General	Rows that differ only by trailing spaces are treated as one key in equality checks, DISTINCT, grouping, window partitions, and indexed lookups, while the different B key stays separate.
✅	—	General	Rows filtered through the CHAR column and rows grouped through cast-derived text keys stayed aligned. Leading spaces and the meaningful X suffix were preserved.
✅	—	General	Two sessions tried to insert keys that differ only by trailing spaces. One insert committed, the other was rejected, and later lookups found one matching row in both completion orders.
✅	—	Cast	Casting a fixed-width CHAR value to text, varchar, and name returned L with no trailing space. Each result had length 1.
✅	—	Cast	Casting a value with a leading space kept that space in TEXT, VARCHAR, and NAME results while removing only the trailing padding spaces.
✅	—	Compare	CHAR values with the same letters compare equal even when one has trailing spaces, while different letters remain unequal.
✅	—	Compare	The value L sorts before R, and the comparison function returns a negative result as expected.
✅	—	Output	A CHAR value keeps its fixed-width display, reports its logical length, and becomes an unpadded text value when converted.
✅	—	Output	CHAR(2) returns AB at its full width, pads the shorter value A to two characters, and leaves unbounded CHAR output as A.
✅	—	Rev	Casting a NULL CHAR value to text, varchar, and name returned NULL. Comparing it with another CHAR value and calling the comparison function also returned NULL without an error.
✅	—	Rev	CHAR values that differ only by trailing spaces joined as the same key. The equality filter kept only A, excluded C, and the ordered results returned A before B.
✅	—	String	Text and varchar values that differ only by a trailing space remain different, and the unpadded value sorts first.

Tip

Reply with @itoqa to send us feedback on this test run.

[github-actions]

	Main	PR
Total	42090	42090
Successful	19854	19856
Failures	22236	22234
Partial Successes1	5440	5440

	Main	PR
Successful	47.1703%	47.1751%
Failures	52.8297%	52.8249%

${\color{lightgreen}Progressions (2)}$

constraints

QUERY: SELECT * FROM DEFAULTEXPR_TBL;

subselect

QUERY: select count(*) from tenk1 t
where (exists(select 1 from tenk1 k where k.unique1 = t.unique2) or ten < 0);

Footnotes

1. These are tests that we're marking as Successful, however they do not match the expected output in some way. This is due to small differences, such as different wording on the error messages, or the column names being incorrect while the data itself is correct. ↩

[coffeegoddd]

@Hydrocharged DOLT

read_tests	from_latency	to_latency	percent_change
covering_index_scan_postgres	2.43	2.43	0.0
groupby_scan_postgres	80.03	78.6	-1.79
index_join_postgres	2.22	2.22	0.0
index_join_scan_postgres	1.61	1.61	0.0
index_scan_postgres	467.3	467.3	0.0
oltp_point_select	0.37	0.37	0.0
oltp_read_only	6.32	6.43	1.74
select_random_points	0.72	0.72	0.0
select_random_ranges	1.03	1.04	0.97
table_scan_postgres	467.3	475.79	1.82
types_table_scan_postgres	1170.65	1170.65	0.0

write_tests	from_latency	to_latency	percent_change
oltp_delete_insert_postgres	6.67	6.67	0.0
oltp_insert	3.36	3.36	0.0
oltp_read_write	13.46	13.46	0.0
oltp_update_index	3.55	3.55	0.0
oltp_update_non_index	3.25	3.25	0.0
oltp_write_only	7.04	7.04	0.0
types_delete_insert_postgres	7.17	7.17	0.0

[zachmu]

Mostly looks good, but confirm how postgres handles other non-space whitespace characters for these cases and add tests for them.

[zachmu]

Is this the right character set to trim? What about other whitespace chars like tabs and newlines?

[zachmu]

This should be an assertion, not part of the setup

[zachmu]

This is a good test, but we should also have a test case which is an exact match of the queries in the original issue. This is close but misses some, e.g. it doesn't directly select 'L '::character(2), it only tests it for equality.

[zachmu]

Same comment about other whitespace chars. What does postgres do?