Update API and add performance enhancements to CollectionTables

[robertcv]

What started with just a few fixes ended up being closer to a rewrite.
The major changes are:

• change main methods to attributes: metadata() -> meta, expressions() -> exp, counts() -> rc
• add clear_cache() which removes all cache files from the default location
• add method id_to_symbol which holds gene id to symbole mapping
• exp and rc return gene ids instead of symboles
• sample name is used instead of sample slug for tables index
• add separate versioning for expression data
• seperatly fetch metadata and expressions
• add downloading progress bar
• mapping of gene ids to symboles is now cached on disc and is also used on other collection
• add CollectionTable to resdk namespace
• store expression type in returned tables attrs['exp_type'] attribute

Performance on 45 samples (exp and rc have the same performance):

how	what	old	new
/	init	0.4s	<0.01s
uncached	meta	127s*	1.7s
uncached	exp	255s*	64s**
uncached	id_to_symbol	/	86s
cached	meta	1.7s	0.4s
cached	exp	0.02s	1.8s
cached	id_to_symbol	/	2.8s

* calling just one of metadata() or expressions(), if used together resolwe server gets queried only once and performance improves for he second function for ~120s
** without gene id to symbol mapping

[robertcv]

@JureZmrzlikar @mstajdohar

[JureZmrzlikar]

I have few comments. Also, make sure that tests pass.

[robertcv]

@JureZmrzlikar Can you look at Jenkins and help me understand why it is failing?

[JureZmrzlikar]

@JureZmrzlikar Can you look at Jenkins and help me understand why it is failing?

I have no clue. @acopar Do you have any idea?

[acopar]

This problem happened because pip switched the default dependency resolver on the recent 20.3.* version. The new resolver has different dependency resolution, and causes some glitches with dependencies, in particular backtracking package versions until timeout. This PR should fix this problem:
https://github.com/genialis/genialis-base/pull/596

We can re-trigger build after genialis-base PR is merged.

[mstajdohar]

What is the status? Can we merge this PR?

[JureZmrzlikar]

I just have few comments. Otherwise looks good to me.

[codecov-io]

Codecov Report

Merging #229 (1196e38) into master (0d51338) will increase coverage by 0.20%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #229      +/-   ##
==========================================
+ Coverage   80.72%   80.92%   +0.20%     
==========================================
  Files          27       27              
  Lines        1619     1636      +17     
==========================================
+ Hits         1307     1324      +17     
  Misses        312      312              

Impacted Files	Coverage Δ
src/resdk/__init__.py	100.00% <100.00%> (ø)
src/resdk/collection_tables.py	100.00% <100.00%> (ø)
src/resdk/utils/table_cache.py	97.14% <100.00%> (+0.47%)	⬆️

---

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 0d51338...1196e38. Read the comment docs.