fix: change vtable columns naming

Summarize the changes

• fix: change vtable columns naming

Closes:

• #641 (closed)
• #642 (closed)

Ensure that

• New code is covered by unit and integration tests.
• Related issues would be automatically closed with gitlab's closing pattern (Closes #1, #2).
• Public modules are documented (check the rendered version with cargo doc --open).
• (if PEST grammar is changed) EBNF grammar reflects these changes (check the result with railroad diagram generator.

Next steps

• Update sbroad submodule in picodata/picodata.
• (if EBNF grammar is changed) create a follow-up issue in picodata/docs.

added 1 commit

• 04ce9ab1 - fix: change vtable columns naming

Compare with previous version

added 4 commits

• 01f19445 - fix: change vtable columns naming
• 1e6888d1 - feat: optimize left join
• ccacc977 - feat: optimize aggregates
• 0827760d - fix: fix tests

Compare with previous version

marked this merge request as ready

added 4 commits

• 19dd7b00 - fix: change vtable columns naming
• 1411ca46 - feat: optimize left join
• 9c32eb0d - feat: optimize aggregates
• bad5b46c - fix: fix tests

Compare with previous version

added 2 commits

• 7177dc36 - feat: optimize aggregates
• 7741437f - fix: fix tests

Compare with previous version

added 6 commits

• 7741437f...40e9e02e - 2 commits from branch main
• 3c6dc03f - fix: change vtable columns naming
• bf99b22b - feat: optimize left join
• 9c7df16c - feat: optimize aggregates
• 5720cf66 - fix: fix tests

Compare with previous version

requested review from @darthunix

added 4 commits

• c52274c7 - fix: change vtable columns naming
• d0167f81 - feat: optimize left join
• fcff4fef - feat: optimize aggregates
• 12b5d96e - fix: fix tests

Compare with previous version

requested review from @senya_volynets and removed review request for @darthunix

added 19 commits

• 12b5d96e...a16c6b7e - 15 commits from branch main
• ffb0e397 - fix: change vtable columns naming
• 1531cf85 - feat: optimize left join
• 1453136d - feat: optimize aggregates
• a57f1556 - fix: fix tests

Compare with previous version

assigned to @EmirVildanov

added 11 commits

• a57f1556...36502282 - 7 commits from branch main
• e513829a - fix: change vtable columns naming
• 800c2863 - feat: optimize left join
• 927f1ed8 - feat: optimize aggregates
• 98d41ef9 - fix: fix tests

Compare with previous version

added 1 commit

• dda67432 - fix: fix tests

Compare with previous version

added 3 commits

• cca4865b - feat: optimize aggregates
• 3fd216aa - fix: fix tests
• 27a809ad - fix: remove useless print

Compare with previous version

If am not mistaken, having table with columns COL_i will lead to errors in query like this:

picodata> select *, COL_1 from t2 join (select * from (select a from t2) t1 join (select a from t2) t2 on true) on true

I don't think it is a big deal, but should we document that names COL_i are reserved for internal use or maybe use other names (for example using uuid)?

I also noticed we have the following behaviour on this patch:

select t.a from t join (select * from (select a from t) 
t1 join (select a from t) t2 on true) on true
---
- metadata:
  - {'name': 'T.A', 'type': 'integer'}
  rows: []
...

select t.a from t
---
- metadata:
  - {'name': 'A', 'type': 'integer'}
  rows: []
...

Which looks a bit strange, @EmirVildanov is it expected or maybe it should be fixed as a follow up?

Изначально в тикете предлагалось сохранять имя скана внутри виртуальных таблиц, что кажется более естественным, чем использовать имена COL_i, почему ты решил отказаться от той идеи?

Просто интересно, что фундаметально мешает это сделать, потому что кажется в таком случае можно было бы избежать "починки" алиасов, что находятся над моушенами

added 19 commits

• 27a809ad...a741a9d7 - 18 commits from branch main
• 82bb0113 - fix: remove useless SubQuery nodes, fix logic of vtable columns naming...

Compare with previous version