Support new versions of cache RP

[matthchr]

If applicable:

• this PR contains documentation
• this PR contains tests

[matthchr]

/ok-to-test sha=96cb206

[codecov-commenter]

Codecov Report

Merging #3206 (96cb206) into main (239eea7) will increase coverage by 0.09%.
Report is 1 commits behind head on main.
The diff coverage is 57.56%.

@@            Coverage Diff             @@
##             main    #3206      +/-   ##
==========================================
+ Coverage   54.39%   54.48%   +0.09%     
==========================================
  Files        1424     1442      +18     
  Lines      601351   614666   +13315     
==========================================
+ Hits       327097   334904    +7807     
- Misses     220754   224925    +4171     
- Partials    53500    54837    +1337     

Files Changed	Coverage Δ
v2/api/cache/customizations/redis_extensions.go	75.43% <ø> (ø)
v2/api/cache/v1api20230401/redis_types_gen.go	55.02% <ø> (ø)
...pi20230401storage/redis_firewall_rule_types_gen.go	55.17% <ø> (ø)
...pi20230401storage/redis_linked_server_types_gen.go	55.17% <ø> (ø)
...i20230401storage/redis_patch_schedule_types_gen.go	55.17% <ø> (ø)
.../api/cache/v1api20230401storage/redis_types_gen.go	59.75% <ø> (ø)
...01/redis_enterprise_database_spec_arm_types_gen.go	33.33% <ø> (ø)
...api20230701/redis_enterprise_database_types_gen.go	56.03% <ø> (ø)
...api20230701/redis_enterprise_spec_arm_types_gen.go	33.33% <ø> (ø)
.../cache/v1api20230701/redis_enterprise_types_gen.go	60.17% <ø> (ø)
... and 38 more

... and 134 files with indirect coverage changes

[super-harsh]

I'd suggest separating these samples out in their own folders like v1api20230401 and v1api20201201 for consistency with other groups

[matthchr]

Not every group does this actually, and because there's a bunch of API versions coexisting (because of redis + redisenterprise) it felt better to put it all into 1 folder like we do for compute so that people don't have to guess which API versions support which resources. Basically I think people are resource-centric and making them choose API versions first doesn't help in sample discovery. It's fine when every resource is in every API version (pick any one and you're right), but in this case not every resource is in every APIversion so it's harder.

[super-harsh]

LGTM! just a couple minors comments.

[matthchr]

/ok-to-test sha=ff9576d