Horton

Re: POC, WIP: OR-clause support for indexes

Nikolay Shaplov <dhyan@nataraj.su>

View thread
From: Nikolay Shaplov <dhyan@nataraj.su>
To: Alexander Korotkov <aekorotkov@gmail.com>, pgsql-hackers@lists.postgresql.org
Cc: Andrei Lepikhov <a.lepikhov@postgrespro.ru>, jian he <jian.universality@gmail.com>, Robert Haas <robertmhaas@gmail.com>, pgsql-hackers@postgresql.org, Peter Geoghegan <pg@bowt.ie>, "Finnerty, Jim" <jfinnert@amazon.com>, Marcos Pegoraro <marcos@f10.com.br>, teodor@sigaev.ru, Tomas Vondra <tomas.vondra@enterprisedb.com>, Peter Eisentraut <peter@eisentraut.org>, Ranier Vilela <ranier.vf@gmail.com>, Alena Rybakina <a.rybakina@postgrespro.ru>
Date: 2024-06-26T17:41:13Z
Lists: pgsql-hackers

Commits

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

  1. Make group_similar_or_args() reorder clause list as little as possible

  2. Allow usage of match_orclause_to_indexcol() for joins

  3. Skip not SOAP-supported indexes while transforming an OR clause into SAOP

  4. Remove the wrong assertion from match_orclause_to_indexcol()

  5. Teach bitmap path generation about transforming OR-clauses to SAOP's

  6. Transform OR-clauses to SAOP's during index matching

  7. Fix the value of or_to_any_transform_limit in postgresql.conf.sample

  8. Transform OR clauses to ANY expression

  9. MergeAttributes code deduplication

  10. SEARCH and CYCLE clauses

  11. Improve estimation of OR clauses using extended statistics.

  12. Teach btree to handle ScalarArrayOpExpr quals natively.

  13. Revise collation derivation method and expression-tree representation.

  14. Instead of trying to force WHERE clauses into CNF or DNF normal form,

В письме от понедельник, 24 июня 2024 г. 23:51:56 MSK пользователь Nikolay 
Shaplov написал:

So,  I continue reading the patch.

I see there is `entries` variable in the code, that is the list of 
`RestrictInfo` objects and `entry` that is `OrClauseGroup` object.

This naming is quite misguiding (at least for me).

 `entries` variable name can be used, as we deal only with RestrictInfo 
entries here. It is kind of "generic" type. Though naming it 
`restric_info_entry` might make te code more readable.

But when we come to an `entry` variable, it is very specific entry, it should 
be `OrClauseGroup` entry, not just any entry. So I would suggest to name this 
variable `or_clause_group_entry`, or even `or_clause_group` , so when we meet 
this variable in the middle of the code, we can get the idea what we are 
dealing with, without scrolling code up.

Variable naming is very important thing. It can drastically improve (or ruin) 
code readability.

========

Also I see some changes in the tests int this patch. There are should be tests 
that check that this new feature works well. And there are test whose behavior 
have been just accidentally affected.

I whould suggest to split these tests into two patches, as they should be 
reviewed in different ways. Functionality tests should be thoroughly checked 
that all stuff we added is properly tested, and affected tests should be checked 
that nothing important is not broken. It would be more easy to check if these 
are two different patches.

I would also suggest to add to the commit message of affected tests changes 
some explanation why this changes does not really breaks anything. This will 
do the checking more simple.

To be continued.

-- 
Nikolay Shaplov aka Nataraj
Fuzzing Engineer at Postgres Professional
Matrix IM: @dhyan:nataraj.su