Allows use <params> in parsers defined steps

[elchupanebrej]

This covers #409 #404 #293

Blocked by #391, it has to be merged first

[codecov]

Codecov Report

Merging #433 (231b108) into master (c7faa9f) will increase coverage by 0.06%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #433      +/-   ##
==========================================
+ Coverage   96.14%   96.20%   +0.06%     
==========================================
  Files          50       50              
  Lines        1684     1715      +31     
  Branches      159      161       +2     
==========================================
+ Hits         1619     1650      +31     
  Misses         38       38              
  Partials       27       27              

Impacted Files	Coverage Δ
pytest_bdd/parser.py	99.55% <100.00%> (ø)
pytest_bdd/scenario.py	92.77% <100.00%> (+0.22%)	⬆️
tests/feature/test_outline.py	100.00% <100.00%> (ø)
tests/feature/test_parametrized.py	100.00% <100.00%> (ø)

---

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 c7faa9f...231b108. Read the comment docs.

[elchupanebrej]

@olegpidsadnyi , please review

[olegpidsadnyi]

Interesting. Can you share the use case where the examples are mixed with the step arguments?

[olegpidsadnyi]

This looks good to me. What do you think, @youtux?

[olegpidsadnyi]

@elchupanebrej Does it really need to be dependent on #391? I'm not sure yet about that one. Cucumber uses rule option no-unused-variables for raising errors on that. Is #391 related to your change or you can make it an independent one?

[elchupanebrej]

@olegpidsadnyi I'm interested in 391 also. I'll prepare a new patch with adequate API for it and will try to not use it in this patch

[elchupanebrej]

@olegpidsadnyi, there is no more dependence on #391

[olegpidsadnyi]

This looks okay to me.

[olegpidsadnyi]

@elchupanebrej could you make cov happy again?

[elchupanebrej]

@olegpidsadnyi done

[elchupanebrej]

Just one thing I don't like about this PR: values from Examples section are converted using converters and after that serialized again to str. I would like to have a possibility to substitute Step variables directly, but this would require huge refactoring

[olegpidsadnyi]

Sorry, I was having the second thoughts and reverted the merge. I don't really understand why do we need to substitute anything at all in the name of the step. Previously it was just a total match. Otherwise a parser match.
Substitution with the stringified fixture value is a slippery slope. Especially this introduces a lot of additional code that is probably not even needed.
What was the original problem again? Need to escape the <>?

[elchupanebrej]

#293 (comment)
and whole #293

For my project I've solved this by using regular expressions parser and special converter, but I really prefer DRY way.

For this PR I check two approaches - full substitution of variables and approach presented in this PR. The second one is more flexible, but more dangerous