Horton
Menu

Re: [PATCH] Add tests for Bitmapset

Michael Paquier <michael@paquier.xyz>

View thread
From: Michael Paquier <michael@paquier.xyz>
To: Greg Burd <greg@burd.me>
Cc: PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>, Robert Haas <robertmhaas@gmail.com>, Nathan Bossart <nathandbossart@gmail.com>, Masahiko Sawada <sawada.mshk@gmail.com>
Date: 2025-09-22T07:57:05Z
Lists: pgsql-hackers 
On Fri, Sep 19, 2025 at 02:48:43PM -0400, Greg Burd wrote:
> I think I've incorporated your feedback which did indeed make the patch
> more readable/crisp and maintainable, thank you.
> 
> coverage:    HEAD    v7    v8
> lines......:  87.1  90.5  92.2

bitmap_match() had some duplicated tests.

bms_replace_members() was not covered, and a function was added to the
module.  I get up to 93.4% for the lines, once added.

+PG_FUNCTION_INFO_V1(test_bms_from_array);
+PG_FUNCTION_INFO_V1(test_bms_to_array);
These two were still declared, are not required.

There were a couple of functions that acted as dead code in the
module, covered in the tree.  As we are trying to cover the gap I
think that we should just do the whole thing.  Here are the functions: 
- is_empty
- subset_compare
- get_singleton_member
- nonempty_difference
- member_index
- join.  The test C function had a mistake, because either input may
be modified or updated by bms_join().  The code was freeing the right
input, leading to errors in the node output function when tested.

Mixing the CTEs and the UNION ALL would have been nice if the queries
in the UNIONs relied on more than one input value, but none of them
did that, so I have removed this part and made the tests leaner.  That
feels easier for the eye.

Testing on equality for the hash functions should be OK, so I have
kept these, including the NULL/0 cases.

A few incorrect things related to the style, fixed on the way, like an
incorrect copyright notice in meson.build, the top of test_bitmapset.c
was largely incorrect.  It seems like these were copy-pasted from
elsewhere.

Another thing was testing with -DREALLOCATE_BITMAPSETS, which I've
done.  A couple of code paths related to the test C functions don't
test for NULL input, which is still OK as these don't impact the
internal tests or the backend coverage, so I've left these out.

The result I had was good enough, so applied.  The CI was OK, the
buildfarm may have a different opinion.
--
Michael

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. Minor fixes for test_bitmapset.c

  2. test_bitmapset: Improve random function

  3. Add mem_exceeded_count column to pg_stat_replication_slots.

  4. Tidy-up unneeded NULL parameter checks from SQL function

  5. Remove check for NULL in STRICT function

  6. Fixes for comments in test_bitmapset

  7. Minor fixups of test_bitmapset.c

  8. test_bitmapset: Simplify code of the module

  9. test_bitmapset: Expand more the test coverage

  10. Fix compiler warnings in test_bitmapset

  11. Add a test module for Bitmapset