CC: Memory and perf improvements

[carlosgjs]

Two main changes:

• Move the processing of a timespan (e.g. day) into a function and use a new ThreadPoolExecutor every time for better memory cleanup
• Process the paths so the check for already done correlations can be done in O(1) time

Test processing 3 days show memory steady:

[codecov-commenter]

Codecov Report

Patch coverage: 63.50% and project coverage change: +8.10% 🎉

Comparison is base (97326b4) 50.59% compared to head (27f5715) 58.70%.
Report is 1 commits behind head on master.

❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #216      +/-   ##
==========================================
+ Coverage   50.59%   58.70%   +8.10%     
==========================================
  Files          32       32              
  Lines        3751     3819      +68     
  Branches      522      528       +6     
==========================================
+ Hits         1898     2242     +344     
+ Misses       1793     1465     -328     
- Partials       60      112      +52     

Files Changed	Coverage Δ
src/noisepy/seis/main.py	43.75% <6.66%> (-1.58%)	⬇️
src/noisepy/seis/scheduler.py	36.58% <11.11%> (-0.73%)	⬇️
src/noisepy/seis/utils.py	92.75% <60.00%> (+13.06%)	⬆️
src/noisepy/seis/correlate.py	75.20% <77.77%> (+58.46%)	⬆️
src/noisepy/seis/constants.py	100.00% <100.00%> (ø)
src/noisepy/seis/hierarchicalstores.py	88.65% <100.00%> (ø)
src/noisepy/seis/numpystore.py	100.00% <100.00%> (ø)
src/noisepy/seis/zarrstore.py	100.00% <100.00%> (ø)

... and 4 files with indirect coverage changes

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[niyiyu]

thank you !