Closed rikhuijzer closed 3 years ago
Nice!
How does jldoctest get rendered with Franklin? btw The SymbolicUtils documentation is generated with Franklin... We may have to hack https://github.com/JuliaSymbolics/SymbolicUtils.jl/blob/master/page/utils.jl#L5-L26 if it doesn't already render the code block properly.
How does jldoctest get rendered with Franklin?
I couldn't run the SymbolicUtils Page build locally due to some broken dependency (I'm running NixOS where I get those issues from time to time). Therefore, I've tested it in Franklin.jl and there the output is indeed incorrect for jldoctest
code blocks.
The workaround is added in d3423d2
, but note that I didn't test the workaround.
Merging #333 (de423d2) into master (aee8482) will not change coverage. The diff coverage is
n/a
.
@@ Coverage Diff @@
## master #333 +/- ##
=======================================
Coverage 83.89% 83.89%
=======================================
Files 12 12
Lines 1217 1217
=======================================
Hits 1021 1021
Misses 196 196
Impacted Files | Coverage Δ | |
---|---|---|
src/utils.jl | 60.37% <ø> (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 aee8482...de423d2. Read the comment docs.
The workaround is added in d3423d2
Looks good enough.
Thanks! This is great!
This is just a suggestion. In the Julia library, Documenter.jl is used to verify all the documentation via doctests. This means that Documenter's doctest logic should be quite robust. Since this package has a lot of examples in the docstrings, maybe it is a good idea to run doctests to avoid broken documentation.
EDIT: For example, other packages which also use doctests are DataFrames.jl and MCMCChains.jl.