Make block/epoch alignment in tests more predictable

[pcarranzav]

(Temporarily targeting the pcv/fix-tests branch as this builds on #554)

Now that blocks are mined predictably in our unit tests (i.e. one block per transaction), any tests that rely on a specific number of blocks passing can break if the number of transactions changes. This was evident when I incorporated the changes from #554 into #552, see: https://github.com/graphprotocol/contracts/runs/5728132209?check_suite_focus=true - adding a transaction in the fixtures bringup caused the test to spill over into the next epoch, changing the results.

So this PR makes a few changes to make tests hopefully more predictable and robust:

• Change the advanceToNextEpoch function to always advance to the first block in the next epoch, rather than always advancing a full epoch
• Change several tests that advanced an epochLength to use advanceToNextEpoch instead
• In takeRewards tests that expect an amount for indexer rewards, align to the first block in an epoch and recalculate the expected rewards using the number of blocks, that is now a bit easier to calculate by just counting the number of transactions in the test code.

[abarmat]

Great improvements ⚡️

[pcarranzav]

(Closed and reopened to trigger CI)

[codecov]

Codecov Report

Merging #555 (e38e530) into dev (3538e00) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##              dev     #555   +/-   ##
=======================================
  Coverage   90.58%   90.58%           
=======================================
  Files          37       37           
  Lines        1795     1795           
  Branches      293      293           
=======================================
  Hits         1626     1626           
  Misses        169      169           

Flag	Coverage Δ
unittests	90.58% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 87a2468...e38e530. Read the comment docs.