findSolid errors

[marcus7070]

Workplane.findSolid had an incorrect docstring about it raising a ValueError when it didn't find any solids. It actually just returns None, and this is behaviour that a few other methods rely on (union for example). So I removed the line in the docstring.

There were several methods that treated the return from findSolid as just a Solid, and hence were trying to call methods on the None object and failing with an AttributeError. I added checks for None and ValueError exceptions to these methods.

Then I thought to myself, "This is mypy's job, why am I doing it?", and realised that findSolid needed the type hint Optional added to it's return value.

I also added tests for all the new exceptions. I realised cutEach isn't tested at all (except for probably in the examples), so I added a basic test for that as well.

[codecov]

Codecov Report

Merging #655 (d711a81) into master (e4ed881) will increase coverage by 0.16%.
The diff coverage is 97.14%.

@@            Coverage Diff             @@
##           master     #655      +/-   ##
==========================================
+ Coverage   94.03%   94.19%   +0.16%     
==========================================
  Files          31       31              
  Lines        6636     6683      +47     
  Branches      726      725       -1     
==========================================
+ Hits         6240     6295      +55     
+ Misses        258      253       -5     
+ Partials      138      135       -3     

Impacted Files	Coverage Δ
cadquery/cq.py	90.80% <95.00%> (+0.90%)	⬆️
tests/test_cadquery.py	99.09% <98.00%> (-0.04%)	⬇️
cadquery/occ_impl/shapes.py	91.89% <0.00%> (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 e4ed881...d711a81. Read the comment docs.

[marcus7070]

I forgot about the Workplane.findFace method, if I've added Optional to findSolid I should do it for findFace as well.

[adam-urbanczyk]

Thanks for picking this up. Open question: isn't it better if findSolid raises? This would result in less code duplication.

[marcus7070]

isn't it better if findSolid raises? This would result in less code duplication.

Firstly, if I had of realised how large this PR was going to be I would have made an issue first and discussed this there. But honestly I was intending to take a 10 minute break from work and fix one line in the findSolid docstring and it just snowballed from there. 😅

I thought about raising an exception in findSolid, but:

• that's a user facing method and it's had the behaviour of returning None for 8 years now, and
• it seemed like the smaller change to add a few raise ValueErrors than to change the behaviour of findSolid.

On the other hand:

• if no user has complained about the error in the docstring for 8 years, it's probably rarely used and a change won't be too disruptive,
• raising in findSolid would bring it in line with how I intend to do getPending in Improved error when pendingWires/pendingEdges is empty #621 (which I'm also working on now),
• it turned out to be more than just a few raise ValueErrors, so raising in findSolid would reduce a lot of duplication,
• at present findSolid is just an alias for Workplane._findType((Solid, Compound)), so it will be trivial to work around if you need the old behaviour.

Now I've written all that out, I prefer raising in findSolid. @adam-urbanczyk, @jmwright, are you OK with me changing the behaviour of findSolid to raise an exception instead of returning None?

[jmwright]

are you OK with me changing the behaviour of findSolid to raise an exception instead of returning None?

I am, yes.

[adam-urbanczyk]

+1

[marcus7070]

Workplane.largestDimension previously returned -1 when it couldn't find a solid/compound. Now it also raises a ValueError.

[jmwright]

@marcus7070 Looks good, thanks!

[adam-urbanczyk]

Thanks @marcus7070