Add helper class to streamline access to expressions and metadata

[robertcv]

This partially implements spec.

What is implemented:

• extend ReSDK API similar to "Alternative 3" with some exceptions:
  • selection of the used collection is moved to class initialization as holding multiple collections in a single object seems a bit messy
  • instead of the name ResolweBio I used CollectionTables as this class functionality is, in a nutshell, exporting collections as tables of data
• fetching expressions, counts, metadata from ReSDK server
• caching data in memory and on disc as long as modified on samples doesn't change
• returned tables have samples in rows and gene symbols for columns

[JureZmrzlikar]

Looks good in general, i have some smaller comments. Will try to play with it after lunch to see it in action :-)

[JureZmrzlikar]

Looks good, I have a few additional small comments...

Please rebase to latest master to see if tests pass. You will probably see that you need to format your docstrings according to PEP8 and PEP257.

[codecov-io]

Codecov Report

Merging #225 (153d705) into master (5c0f624) will increase coverage by 2.85%.
The diff coverage is 99.39%.

@@            Coverage Diff             @@
##           master     #225      +/-   ##
==========================================
+ Coverage   77.86%   80.72%   +2.85%     
==========================================
  Files          25       27       +2     
  Lines        1455     1619     +164     
==========================================
+ Hits         1133     1307     +174     
+ Misses        322      312      -10     

Impacted Files	Coverage Δ
src/resdk/utils/table_cache.py	96.66% <96.66%> (ø)
src/resdk/collection_tables.py	100.00% <100.00%> (ø)
src/resdk/__about__.py	84.61% <0.00%> (+84.61%)	⬆️

---

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 5c0f624...153d705. Read the comment docs.

[JureZmrzlikar]

@robertcv This looks good for MVP, please make linters happy to ensure that tests pass. Thanks for your contribution!

[mzganec]

return {
    feature.feature_id: feature.name
    for feature in self.resolwe.feature.filter(
        feature_id__in=ensembl_ids,
        species=self._table_samples[0].metadata["general"]["species"],
        fields=["feature_id", "name"],
    ).iterate(CHUNK_SIZE)
}