@b3m2a1 @Szabolcs @Kuba Since you're talking about documentation again, I thought I'd (shamelessly) link you to my approach to the problem: https://github.com/MMA-ForScience/ForScience#pacletutils Next to other things, the paclet contains a fully self-contained documentation creation and build framework - no need for the workbench at all, everything is done from within mathematica. The goal is to eventually support most documentation pages through this process. Currently, Symbol and Guide pages are fully implemented. A simple example might be:
DocumentationHeader[foo] = {"EXAMPLE SYMBOL", Gray, "Never introduced."};
Usage[foo] = "foo[a,b] does something.";
Details[foo] = {"[*foo*] does great things if applied correctly."};
Examples[foo, "Basic examples"] = {{"Use [*foo*] for something cool:", ExampleInput[foo[a, b]]}};
SeeAlso[foo] = {List, Set}; nb = DocumentationBuilder[foo];
It is currently used to document the paclet itself, so it's already in a usable state, although I have no idea what kind of functionality could be needed by others (feature requests are welcome ;) )
Currently, it builds only 11.0 style documentation pages (with documentation indexes for both pre 11.2 and post 11.2), but due to how everything inside the notebooks is built from a well-defined format, adding support for older documentation versions shouldn't be a big issue
Let me know if you have any questions/feedback/suggestions
@LukasLang nice to see I’m not the only one who’s built that sort of thing. Unfortunately the HTML docs it builds aren’t great for a ReadTheDocs type approach.
@b3m2a1 Well, if one really wanted to, it should be possible to just build your own HTML generator from the ground up - but that's a lot of work… If there's enough interest, I might look into making something that produces pages in the style of the online documentation - Let me know if anyone's interested
Also, it's really a shame that paclets built by others aren't easily discoverable (I learned of your paclets only very recently for example) - hopefully your paclet server project helps with that
It generally looks fine, with the exceptions of a bunch of places where I need to patch up some Permissions, but it's slow and unnecessarily finicky. It uses the Workbench NB->HTML XLST processing framework (which I literally just extracted from Workbench so that I didn't have to use it).
This framework can be customized at will, but you'd have to provide maybe ~100+ conversions and deal with its idiosyncrasies and bugs and things.
Whereas, say, this was built from raw Markdown which was originally generated from a notebook
And is less complicated but much easier to work with and gets the job done.
For an in-MMA bug tracker what are the most relevant properties? I currently have: {"Title", "Description", "Timestamp", "Examples", "Resolved", "Keywords", "Package"} which I hope gives me sufficient coverage to be able to do effective searches and formatting and things. Is there anything I missed that should be added in?
Hi all. I think I am missing something simple. Given a list, say Range[6], how do I partition it based on some list of sizes? Like given {2,3,1} sizes I would get:
@Halirutan Hello & Congrats. Do you know how J. M. is - in terms of health and current financial need? And also, if you know of a source for this info that you'd prefer I use instead of you, I certainly will. Thanks, Rabbit
Carl Woll pointed out that there's a function (TakeList or maybe ListTake) whose params are list bname and {length1, length2, ... lengthn}. This is new to 11.3
@VitailyKaurov Sorry I failed to address the note re TakeList to you.
