refactor: implement an updated parser [WIP]

[Midnighter]

• Upgrade package infrastructure.
• Implement a tokenizer (lexer) and a parser (compiler) that assembles the token values to SQLAlchemy models.
• Add tests for many situations.

This PR is work in progress. It completely updates the structure, configuration, and adds continuous integration. Once merged the BRENDA-Parser will be installable from PyPI.

What's missing:

1. More test cases.
2. Relationships on Enzyme class.
3. Some errors still occur during parsing.
4. Documentation
5. Automatic releases via Travis and readthedocs.

[codecov]

Codecov Report

❗ No coverage uploaded for pull request base (master@999fbca). Click here to learn what that means.
The diff coverage is 83.92%.

@@            Coverage Diff            @@
##             master       #6   +/-   ##
=========================================
  Coverage          ?   83.92%           
=========================================
  Files             ?       16           
  Lines             ?      591           
  Branches          ?        0           
=========================================
  Hits              ?      496           
  Misses            ?       95           
  Partials          ?        0

Impacted Files	Coverage Δ
src/brenda_parser/cli.py	0% <0%> (ø)
src/brenda_parser/parsing/__init__.py	100% <100%> (ø)
src/brenda_parser/exceptions.py	100% <100%> (ø)
src/brenda_parser/models/__init__.py	100% <100%> (ø)
src/brenda_parser/models/organism.py	100% <100%> (ø)
src/brenda_parser/data/__init__.py	100% <100%> (ø)
src/brenda_parser/models/enzyme.py	100% <100%> (ø)
src/brenda_parser/models/comment.py	100% <100%> (ø)
src/brenda_parser/models/field_entry.py	100% <100%> (ø)
src/brenda_parser/api.py	30.88% <30.88%> (ø)
... and 6 more

---

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 999fbca...95e9b3c. Read the comment docs.

[matthiaskoenig]

@Midnighter Is this working/usable?
Any documentation on how to use this?

I just tried the following in the hope it parses the brenda file and puts it into a sqlite database:

from brenda_parser import cli
cli.parse(flat_file="brenda.sqlite", filename="./brenda_parser/brenda_download.txt")

But resulted in

Traceback (most recent call last):
  File "/home/mkoenig/git/pancreas_model/pypancreas/webservices/brendaparser/brenda.py", line 5, in <module>
    cli.parse(flat_file="brenda.sqlite", filename="./brenda_parser/brenda_download.txt")
  File "/home/mkoenig/envs/pancreas/lib/python3.6/site-packages/click/core.py", line 764, in __call__
    return self.main(*args, **kwargs)
  File "/home/mkoenig/envs/pancreas/lib/python3.6/site-packages/click/core.py", line 716, in main
    with self.make_context(prog_name, args, **extra) as ctx:
  File "/home/mkoenig/envs/pancreas/lib/python3.6/site-packages/click/core.py", line 639, in make_context
    ctx = Context(self, info_name=info_name, parent=parent, **extra)
TypeError: __init__() got an unexpected keyword argument 'flat_file'

Would be great to finish this up and release on pypi.

[Midnighter]

Hey, I wasn't sure it's still worth since I thought BRENDA was now included in SABIO-RK and other sources. However, you're the third person asking for it so, I guess, I should finish it up.

It's definitely not useable yet. I had experimented with different ways of parsing the flat file

1. tried to improve the PLY setup
2. gave tatsu a shot

but I ran into a lot of problems with both (and they were also fairly slow). So I might use pyparsing instead.

[matthiaskoenig]

I just talked to the SABIO-RK developers and there is no plan on BRENDA integration. BRENDA is quit restrictive with the availability of their data (for instance no database dump, only somehow limiting text download).
I solved all my Brenda issues for now (I had some old parser code which I updated and is solving my use case).
Thanks Matthias