algod importer: Combines algod and algod_follower

[AlgoStephenAkiki]

Resolves #1449

Combines the algod and algod_follower plugins and uses a config value to switch between the logic of them

[codecov]

Codecov Report

Merging #1452 (7aca164) into develop (cb4581c) will decrease coverage by 0.13%.
The diff coverage is 68.29%.

@@             Coverage Diff             @@
##           develop    #1452      +/-   ##
===========================================
- Coverage    65.28%   65.16%   -0.13%     
===========================================
  Files           83       81       -2     
  Lines        11452    11369      -83     
===========================================
- Hits          7477     7409      -68     
+ Misses        3404     3395       -9     
+ Partials       571      565       -6     

Impacted Files	Coverage Δ
conduit/plugins/importers/algod/algod_importer.go	77.96% <68.29%> (-4.65%)	⬇️
fetcher/fetcher.go	63.98% <0.00%> (+1.27%)	⬆️

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[winder]

Two main changes:

1. Find a way to avoid a string comparison to toggle functionality in the GetBlock path.
2. Fix the tests. It looks like you applied the same pattern blindly to all tests. Most of them don't seem to add any coverage, the ones that do don't have any new assertions to check that the functionality is correct.

[shiqizng]

I'm wondering if this retry loop is necessary. there's already retries at the pipeline level. in this loop there's no wait before the next try, could this loop always finish before the next block is available?

[Eric-Warehime]

Yep, just added retires w/ backoff to the most recent version.

for r := 0; r < retries; r++ {
		time.Sleep(time.Duration(waitMultiplier*float64(r)) * initialWait)
...

We could just rely on the pipeline-level retires like you said, but retries with backoff make more sense in the case of the follower node since the catchup service gets started/stopped so frequently.