Add E2E Test Suite

[robgpita]

This PR will close #33.

Reference data used for regression testing has been added. This adds 245M to the repo, which will slow down the git pull command. The collection (mip_17110008) used for regression testing was the smallest collection with all of the appropriate files and libraries. Unfortunately, the rating curves table in the ripple.gpkg is empty, so tests for the flows2fim validate command will have to occur with either a new set of reference data, ( instructions have been included), or as part of the next milestone (#35), when the ability to access S3 resources is incorporated as part of the test suite.

The bulk of this test suite is embodied within a bash script. This script has 3 options for arguments : controls, fim, & validate which will test their respective flows2fim utility. If no argument is provided, all flows2fim utilities will be tested.

The github actions workflow file has also been updated to run test_suite_linux.sh for each flows2fim utility, using the test matrix shown below for flows2fim, as well as the test suite's functionality itself.

Contents of test_matrix.md:

Test Matrix

	generate outputs	regression tests	Assert errors
controls
Generate 2,5,10,25,50,100 yr controls.csv files	☑	☑
missing -scsv			☑
missing -scvs, -o			☑
-db only			☑
-scsv empty			☑
-f empty			☑
-f columns swapped			☑
-f no values			☑
fim
Generate 2,5,10,25,50,100 yr fim.tif files	☑	☑
Generate fim files in different output formats	☑	☒
-rel false			☒
missing -c			☑
missing -l			☑
missing -o			☑
empty -c file			☑
validate

[ar-siddiqui]

@robgpita are we adding the whole collection FIM library to the reop or the bear minimum test cases that we need?
I don't think we should add 254MB of data to the repo.

[robgpita]

@ar-siddiqui yes, I added the whole collection for simplifying the regression testing (comparing against data that has already been generated without removing specific reaches and libraries). We can pare this down, but more specific instructions must be included as to which files need to be deleted. Do you have any ideas on how to keep the testdata lightweight, but test all of the flows2fim functionality?

Also, the reference flows files comprise the majority of the data footprint (240ish MB) and can be simplified to a subset of collections we intend to use for regression testing. Those collections have yet to be identified, so I did not delete unnecessary rows in the reference .csv flows files.

[ar-siddiqui]

Looks great. This will be of great help in development.

Please address few minor comments.