Remove MissingUserDefinedType from public API

[Lorak-mmk]

Up until now CqlType::UserDefinedType contained a Result, with error type MissingUserDefinedType.
This error is an edge case that should be quite rare. It could possibly happen if we fetch metadata in the middle of schema change.
Possible scenario:

• We fetch UDTs
• New UDT is added
• New column is added to some table, with the type being this new UDT
• We fetch column types for all tables
• Now we have a column type with unknown UDT.

As this error should be extremely rare, it is okay to handle it a little bit less gracefully.
This simplifies user API (user no longer needs to handle the error) and makes progress towards unifying CqlType and ColumnType.

The new proposed way to handle this error is to drop the fetched metadata for the whole keyspace - re-using older version of it if available - and printing a warning.

As a side-fix, I found that query_tables_schema was unnecessarily called twice, so I deduplicated those calls.
query_tables_schema is likely the heaviest part of metadata reading, so this fix should lower the cluster load resulting from metadata fetches.

Progresses towards: #691

Pre-review checklist

• I have split my patch into logically separate commits.
• All commit messages clearly explain what they change and why.
• I added relevant tests for new features and bug fixes.
• All commits compile, pass static checks and pass test.
• PR description sums up the changes and reasons why they should be introduced.
• I have provided docstrings for the public items that I want to introduce.
• I have adjusted the documentation in ./docs/source/.
• I added appropriate Fixes: annotations to PR description.

[github-actions]

cargo semver-checks detected some API incompatibilities in this PR.
Checked commit: 800aabc

See the following report for details:

cargo semver-checks output

./scripts/semver-checks.sh --baseline-rev a216932412c375fc81a5737257edfe2236ac3e08
+ cargo semver-checks -p scylla -p scylla-cql --baseline-rev a216932412c375fc81a5737257edfe2236ac3e08
     Cloning a216932412c375fc81a5737257edfe2236ac3e08
    Building scylla v0.15.0 (current)
       Built [  22.639s] (current)
     Parsing scylla v0.15.0 (current)
      Parsed [   0.052s] (current)
    Building scylla v0.15.0 (baseline)
       Built [  22.443s] (baseline)
     Parsing scylla v0.15.0 (baseline)
      Parsed [   0.049s] (baseline)
    Checking scylla v0.15.0 -> v0.15.0 (no change)
     Checked [   0.108s] 107 checks: 106 pass, 1 fail, 0 warn, 0 skip

--- failure struct_missing: pub struct removed or renamed ---

Description:
A publicly-visible struct cannot be imported by its prior path. A `pub use` may have been removed, or the struct itself may have been renamed or removed entirely.
        ref: https://doc.rust-lang.org/cargo/reference/semver.html#item-remove
       impl: https://github.com/obi1kenobi/cargo-semver-checks/tree/v0.38.0/src/lints/struct_missing.ron

Failed in:
  struct scylla::cluster::metadata::MissingUserDefinedType, previously in file /home/runner/work/scylla-rust-driver/scylla-rust-driver/target/semver-checks/git-a216932412c375fc81a5737257edfe2236ac3e08/e302ccb9dfc77c2b9b626ab3e35888a4548dda73/scylla/src/cluster/metadata.rs:289

     Summary semver requires new major version: 1 major and 0 minor checks failed
    Finished [  46.389s] scylla
    Building scylla-cql v0.4.0 (current)
       Built [  10.894s] (current)
     Parsing scylla-cql v0.4.0 (current)
      Parsed [   0.034s] (current)
    Building scylla-cql v0.4.0 (baseline)
       Built [  11.035s] (baseline)
     Parsing scylla-cql v0.4.0 (baseline)
      Parsed [   0.034s] (baseline)
    Checking scylla-cql v0.4.0 -> v0.4.0 (no change)
     Checked [   0.110s] 107 checks: 107 pass, 0 skip
     Summary no semver update required
    Finished [  22.668s] scylla-cql
make: *** [Makefile:61: semver-rev] Error 1

[muzarski]

Looks good. I really dislike how complicated all the methods have become. They accept hashmap where value is a Result, and return a really complicated type with nested Result. I don't see how we could simplify this, though. It looks horrible, but does the job - we can propagate the error for each keyspace up to ClusterData::new() thanks to this.

[Lorak-mmk]

Rebased on #1163

[Lorak-mmk]

Rebased on main - should be ready to merge after getting an approve.

[muzarski]

Looks good!

[Lorak-mmk]

@wprzytula ping