fix: exclude supabase managed schemas from inspect subcommand

[encima]

What kind of change does this PR introduce?

Feature/Fix: moving queries for inspect to the root inspect module and removing Supabase schemas from queries

What is the current behavior?

Queries are within the command and show results from Supabase schemas that could cause concern or make the user take actions on schemas they do not own

What is the new behavior?

• Queries moved into a single place (could be further developed to template similar queries)
• Supabase schemas removed from relevant queries
• Explanation of hit rates added to cache-hit command and a column explains whether action might be needed

Additional context

[coveralls]

Pull Request Test Coverage Report for Build 8665741536

Details

• 2 of 46 (4.35%) changed or added relevant lines in 20 files are covered.
• 5 unchanged lines in 1 file lost coverage.
• Overall coverage decreased (-0.07%) to 57.738%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
internal/inspect/bloat/bloat.go	0	1	0.0%
internal/inspect/calls/calls.go	0	1	0.0%
internal/inspect/index_sizes/index_sizes.go	0	1	0.0%
internal/inspect/outliers/outliers.go	0	1	0.0%
internal/inspect/role_connections/role_connections.go	0	1	0.0%
internal/inspect/table_index_sizes/table_index_sizes.go	0	1	0.0%
internal/inspect/total_index_size/total_index_size.go	0	1	0.0%
internal/inspect/vacuum_stats/vacuum_stats.go	0	1	0.0%
internal/inspect/blocking/blocking.go	0	2	0.0%
internal/inspect/index_usage/index_usage.go	0	2	0.0%

Files with Coverage Reduction	New Missed Lines	%
internal/gen/keys/keys.go	5	11.76%

Totals
Change from base Build 8657336860:	-0.07%
Covered Lines:	6376
Relevant Lines:	11043

---

💛 - Coveralls

[sweatybridge]

Just one more comment. Everything else looks good.

[sweatybridge]

I've verified all commands locally and fixed the result types where appropriate.

A new change I made is to include managed schemas like auth, storage, extensions, etc. when inspecting table sizes, ie. table-records-count, table-sizes, total-table-sizes. Because records in these schemas are mostly inserted by the user, it would be a good idea to surface their sizes. Schemas with pg_* prefix and information_schema are still excluded from these queries.

In the future, we can perhaps move queries.go to individual *.sql files for better syntax checks.