Issue-62: Add possibility to load a list of comma-separated files.

[JulienPeloton]

What is new in this PR?

This PR allows to load a list of comma-separated FITS files as described in #62:

// Scala
val fns = "/path/to/file1.fits,/path/to/file2.fits,..."
val df = spark.read.format("fits").option("hdu", 1).load(fns)

# Python
fns = "/path/to/file1.fits,/path/to/file2.fits,..."
df = spark.read.format("fits").option("hdu", 1).load(fns)

How this has been tested?

New unit test added.

Other changes?

Bump spark version to 2.4.0 for the build.

[codecov-io]

Codecov Report

Merging #63 into master will increase coverage by 0.06%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master      #63      +/-   ##
==========================================
+ Coverage   89.37%   89.44%   +0.06%     
==========================================
  Files           9        9              
  Lines         480      483       +3     
  Branches       87       90       +3     
==========================================
+ Hits          429      432       +3     
  Misses         51       51

Impacted Files	Coverage Δ
...strolabsoftware/sparkfits/FitsSourceRelation.scala	97.46% <100%> (+0.09%)	⬆️

---

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 3442a2c...aecc184. Read the comment docs.

[JulienPeloton]

Note however that loading a list of paths turns out to not be as easy as I would think.
The input parameters are implicitly read as Map[String, String] (from RelationProvider), hence requiring an array requires some deeper changes. In addition, Spark Data Source API is evolving quickly to V2 (spark-fits is V1), which makes things highly confusing for me :-(