sql: allow strpos() builtin function to support bit and bytes array

[abhishek20123g]

Fixes #45849

This commit modified strpos builtin function to allow it to
support bit and byte array and add their respective testcases.

This PR creates separate overloads for both bit and bytes as:
strpos(varbit, varbit)
strpos(bytes, bytes)
since POSITION is an alias for strpos, at the parser level.
Therefore this PR affect POSITION too.

Release justification: low-risk change to existing functionality.

Release note (sql change): This PR modified strpos() function
to allow it support bit and byte array.

[cockroach-teamcity]

This change is 

[abhishek20123g]

small overview

1. Above implementation can be done without writing functions bytesOverload2 and bitsOverload2 but written to maintain the overall uniformity of code and these function may be used in future.
2. For bit array, this can be done without typecasting to string.

index := strings.Index(bitString.BitArray.String(),bitSubstring.BitArray.String())

for that, we have to implement a separate function like strings.Index inside
bitarray.BitArray which indeed end up the same implementation as
strings.Index.

[otan]

LGTM!

---

leaving this here as a comment to myself, but if you want to nerd it out you can have a squiz!

it looks like postgres handles POSITION as a position C function (which is not available as a builtin)

https://github.com/postgres/postgres/blob/de3bbfcc962f24c1a20a17b76c604e5973a05817/src/backend/parser/gram.y#L13946

whereas we use strpos

cockroach/pkg/sql/parser/sql.y

Line 8714 in 7dc8cc4

$$.val = &tree.FuncExpr{Func: tree.WrapFunction("strpos"), Exprs: $3.exprs()}

one could argue strpos should only be strings, not bits and bytes. but doing a position builtin will make parsing angry as it will fight with the line mentioned above. so whatever /shrug

[otan]

bors r+

[otan]

bors r-

bors is stick due to github issue

[otan]

bors r+

[abhishek20123g]

TF[YT]R

Hi @otan,
Sorry for asking this here, if you have any issue in mind other than good first issue which I could take a shot with my current knowledge, please let me know,
I will be very happy to tackle it 😊.

[otan]

looks like this never went in

bors r+

looks like we'll be out of builtins (for now) ...

#44616 and #41333 should be good.

[craig]

Build failed

• GitHub CI (Cockroach)

[otan]

flakes bors r+

…

On Mon, 6 Apr 2020, 9:29 pm craig[bot], ***@***.***> wrote: Build failed - GitHub CI (Cockroach) <https://teamcity.cockroachdb.com/viewLog.html?buildId=1857029&buildTypeId=Cockroach_UnitTests> — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#46875 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AA32FQ6QCYSQUZ6G3XEZ65DRLKT23ANCNFSM4LZP4OAA> .

[craig]

Build succeeded

• GitHub CI (Cockroach)