doctest_fix: try setting fixed default terminal size

[benlorenz]

These should be the correct defaults, I tried this with a few functions and it did not modify the output.
Would be great if someone else could try as well!

cc: @HechtiDerLachs

[codecov]

Codecov Report

Attention: Patch coverage is 0% with 20 lines in your changes missing coverage. Please review.

Project coverage is 84.69%. Comparing base (2fe164e) to head (5d2eb7a).
Report is 4 commits behind head on master.

Files with missing lines	Patch %	Lines
src/utils/docs.jl	0.00%	20 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #4171      +/-   ##
==========================================
+ Coverage   84.68%   84.69%   +0.01%     
==========================================
  Files         628      628              
  Lines       84471    84581     +110     
==========================================
+ Hits        71533    71636     +103     
- Misses      12938    12945       +7     

Files with missing lines	Coverage Δ
src/utils/docs.jl	18.34% <0.00%> (-0.52%)	⬇️

... and 5 files with indirect coverage changes

[fingolfin]

Good idea and seems sensible to me, but I have not (yet) tested it myself.

[joschmitt]

I just tried this and Oscar.build_doc(doctest = :fix) still produces changes like this for me:

--- a/experimental/FTheoryTools/src/AbstractFTheoryModels/attributes.jl
+++ b/experimental/FTheoryTools/src/AbstractFTheoryModels/attributes.jl
@@ -1677,7 +1677,8 @@ julia> components_of_dual_graph(qsm_model)
  "C7"
  "C8"
  "C9"
- ⋮
+ "C13"
+ "C14"
  "C16"
  "C17"
  "C21"

(It also adds stacktraces to several intentional errors.)

[benlorenz]

I just tried this and Oscar.build_doc(doctest = :fix) still produces changes like this for me:

--- a/experimental/FTheoryTools/src/AbstractFTheoryModels/attributes.jl
+++ b/experimental/FTheoryTools/src/AbstractFTheoryModels/attributes.jl
@@ -1677,7 +1677,8 @@ julia> components_of_dual_graph(qsm_model)
  "C7"
  "C8"
  "C9"
- ⋮
+ "C13"
+ "C14"
  "C16"
  "C17"
  "C21"

(It also adds stacktraces to several intentional errors.)

I only touched the Oscar.doctest_fix function and not build_doc, but I guess that would also make sense. Not sure if we can do anything about the stacktraces.

[joschmitt]

I just tried this and Oscar.build_doc(doctest = :fix) still produces changes like this for me:

--- a/experimental/FTheoryTools/src/AbstractFTheoryModels/attributes.jl
+++ b/experimental/FTheoryTools/src/AbstractFTheoryModels/attributes.jl
@@ -1677,7 +1677,8 @@ julia> components_of_dual_graph(qsm_model)
  "C7"
  "C8"
  "C9"
- ⋮
+ "C13"
+ "C14"
  "C16"
  "C17"
  "C21"

(It also adds stacktraces to several intentional errors.)

I only touched the Oscar.doctest_fix function and not build_doc, but I guess that would also make sense. Not sure if we can do anything about the stacktraces.

Oh, sorry, my mistake.

[joschmitt]

I can confirm that doctest_fix works better with this pull request.

[HechtiDerLachs]

Thank you @benlorenz ! I didn't try it, yet, but I will eventually.

[benlorenz]

I added the environment variables to build_doc as well. Running the whole docs with :fix now only caused the stacktraces to appear (which is a known documenter issue) but no other changes.

[joschmitt]

Thank you! Works on my machine.