Add TTL/Expiration, Increment/Decrement, and Pub/Sub to Cache service

[NitinKumar004]

Summary

The Cache driver interface only has basic Set/Get/Delete/Keys/FlushAll. It's missing TTL expiration, atomic counters, and pub/sub — the most fundamental Redis features.

What's Missing

TTL / Expiration

• SetWithTTL (set key with expiration)
• GetTTL / SetTTL / Persist (manage key expiration)
• Expired key auto-cleanup

Atomic Counters

• Incr / Decr / IncrBy / DecrBy
• IncrByFloat

Pub/Sub

• Subscribe / Unsubscribe / Publish
• PSubscribe (pattern-based)

Data Structures (stretch)

• LPush / RPush / LPop / RPop / LRange (Lists)
• SAdd / SRem / SMembers / SIsMember (Sets)
• HSet / HGet / HDel / HGetAll (Hashes)

Provider Mapping

Feature	AWS ElastiCache	Azure Cache for Redis	GCP Memorystore
TTL	EXPIRE/TTL	EXPIRE/TTL	EXPIRE/TTL
Counters	INCR/DECR	INCR/DECR	INCR/DECR
Pub/Sub	Pub/Sub	Pub/Sub	Pub/Sub

Why This Matters

• Cache without TTL is fundamentally incomplete — expiration is the Add 6 realistic cloud behaviors across AWS, Azure, and GCP providers #1 reason to use a cache
• Atomic counters are used for rate limiting, view counts, and distributed locks
• Pub/Sub is core to real-time messaging patterns

Acceptance Criteria

• Add TTL, counter, and pub/sub methods to cache/driver/driver.go
• Implement in all 3 providers (ElastiCache, Azure Redis, Memorystore)
• TTL expiration should work with FakeClock for deterministic testing
• Add tests covering all 3 providers