Remove lru_cache from get_active_msgids

[DanielNoord]

• Add yourself to CONTRIBUTORS if you are a new contributor.
• Write a good description on what the PR does.

Type of Changes

	Type
✓	🔨 Refactoring

Description

Ref #5670.

Instead of using lru_cache we can use a basic cache defined on the instance to achieve a similar performance boost.

Tested on pylint over pylint.
With new form of cache:

ncalls  tottime  percall  cumtime  percall filename:lineno(function)
54436    0.029    0.000    0.030    0.000 message_id_store.py:107(get_active_msgids)

Without new cache:

ncalls  tottime  percall  cumtime  percall filename:lineno(function)
54429    0.099    0.000    0.132    0.000 message_id_store.py:107(get_active_msgids)

Problem introduced in #5605

[coveralls]

Pull Request Test Coverage Report for Build 1692803372

• 8 of 8 (100.0%) changed or added relevant lines in 1 file are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage increased (+0.002%) to 93.831%

---

Totals
Change from base Build 1692435585:	0.002%
Covered Lines:	14647
Relevant Lines:	15610

---

💛 - Coveralls

[Pierre-Sassoulas]

Looks good, I did not understand the performance report in the description though. No cache was actually faster than using lru cache ?!

[DanielNoord]

Seems I miswrote the description. Updated.

The first is with this change, the second is without any form of cache. I haven't included lru_cache as that's not really a viable option.