Closed bollwyvl closed 4 years ago
Thank you for the patch! I don't use TypeScript myself, so I have little idea what's going on in our TS support and am thus a poor judge. So I would put it to you: how do you think your fix compares with graup's fix, linked from https://github.com/mozilla/sphinx-js/issues/119#issuecomment-537357793? I am inclined to wait for a test to merge, as well, especially in the absence of better understanding of the code on my part. :-)
This is fixed in the great IR merge of #120, demonstrated by https://github.com/mozilla/sphinx-js/blob/705659ceb08de1ec1fd16bdefd53e7b2e0a8d848/tests/test_typedoc_analysis/test_typedoc_analysis.py#L375-L379.
Amazing stuff! Excited to take a look with the jupyterlab and lumino code bases!
Let me know how it goes! I'm interested to get some real-world proof that it works before releasing it; there were a lot of changes!
Here's a fun one... yields:
(nothing much else in the log)
it appears that
make_type_name
is always expected to return aList[Text]
: this change ensures atypeParameter
with a constraint does not becomeList[Union[Text, List[Text]]
... or worse! I see some of them are having the0
th item taken, would that be more appropriate?Happy to work up a test... i see there are some ts files that are getting round-tripped, which includes a pretty gnarly generic, as well as a build example... what would be most beneficial to add?
However, I can't claim to have gotten all the way down the rabbit hole due to our use of lerna, and trying to make typedoc-plugin-lerna-packages work as well... my environment is kind of a mess of hacks right now.