Give more helpful encoding exceptions in pack_sim

[mferrera]

Resolves #537

• Added pytest-xdist to parallelize the tests
• Pinned urllib3<2, as urllib3>=2 is pulled in through the dependency tree and is incompatible with RGS machines
• Gave a more helpful exception reason on utf-8 encoding errors. It roughly copies how ERT handles it but stops at the first bad character

A specific command to do the conversion could be given but it's probably better not to be responsible if it damages the file.

[codecov-commenter]

Codecov Report

Merging #545 (46f08ba) into master (87a3dde) will increase coverage by 0.04%.
The diff coverage is 89.47%.

❗ Current head 46f08ba differs from pull request most recent head c9818c7. Consider uploading reports for the commit c9818c7 to get more accurate results

@@            Coverage Diff             @@
##           master     #545      +/-   ##
==========================================
+ Coverage   86.17%   86.21%   +0.04%     
==========================================
  Files          49       49              
  Lines        7025     7046      +21     
==========================================
+ Hits         6054     6075      +21     
  Misses        971      971              

Impacted Files	Coverage Δ
src/subscript/pack_sim/pack_sim.py	90.10% <89.47%> (+0.82%)	⬆️

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[alifbe]

@mferrera very nice 👍

Do we need to check it as well for inspect_file and _md5checksum as well?

Also would it be nice/needed to print out the correct encoding to end-user? I see that komodo
has chardet library install which can guess the correct encoding format or alternatively file -ib

[mferrera]

I don't think we need to check in inspect_file or _md5checksum. The root .DATA file won't make it to inspect_file with an invalid character, but it is possible that INCLUDEd files have invalid characters. The script already checks for this but concludes that they are actually binary and just copies them without extra processing, so it won't fail (but maybe it should). I added another test case for this.

There are a number of possible improvements that could address possibly-unintentional logic like this in the script but maybe out of scope.

I'm not certain most users would gain a lot of actionable insight if we tell them the encoding. My guess is that the vast majority will probably find ISO-8859-1 a bit meaningless to them 😅 but if you think it's a good idea I'm happy to add it

[alifbe]

@mferrera looks good to me.

Regarding printing out correct encoding, I’m not sure either. I saw in slack channel that user can change/convert the file encoding by using iconv command and specify which encoding to convert from.