Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
  • Nov 17 12:35
    halirutan commented #175
  • Nov 17 11:29
    HydrogenDeuterium assigned #175
  • Nov 17 11:29
    HydrogenDeuterium opened #175
  • Oct 13 11:30
    szhorvat commented #173
  • Oct 12 02:23
    halirutan commented #173
  • Sep 30 22:56
    halirutan closed #174
  • Sep 30 22:56
    halirutan commented #174
  • Sep 30 18:24
    subhodipbiswas assigned #174
  • Sep 30 18:24
    subhodipbiswas opened #174
  • Sep 18 08:25
    szhorvat commented #173
  • Sep 13 10:55
    halirutan labeled #173
  • Sep 13 10:55
    halirutan commented #173
  • Sep 13 06:52
    szhorvat assigned #173
  • Sep 13 06:52
    szhorvat opened #173
  • Aug 11 03:41
    halirutan closed #170
  • Aug 11 03:41
    halirutan commented #170
  • Aug 07 11:10
    halirutan closed #172
  • Aug 07 03:40
    simonkeys commented #172
  • Aug 07 01:56
    halirutan commented #172
  • Aug 07 00:23
    simonkeys edited #172
Kuba Podkalicki
@kubaPod
[[nested/path]] does not work and direct links open in a new tab
Patrick Scheibe
@halirutan
@kubaPod Well, this is odd because it worked for me. I reworked the side-bar for the JabRef project some time ago.
I'm not sure we had nested content but the links to internal md pages definitely open in the same tab for me.
Kuba Podkalicki
@kubaPod
@halirutan there is a flat list of .md files
I want to keep'em organized in a nested directory structure
Files, not the sidebar content (yet).
Patrick Scheibe
@halirutan
Hmm, that that maybe a good question for SO?
I don't know without trying.
Kuba Podkalicki
@kubaPod
I was surprised how hard it is to find a manual about _Sidebar syntax
Patrick Scheibe
@halirutan
Yep, took me also a while back then.
Kuba Podkalicki
@kubaPod
I think the answer is, you can't keep them in nested dirs if you want links to work well, because something about gollum
Gerli
@gerli
I'm not sure if I'm doing something wrong or there's a bug here (I only just discovered the proper module structure set-up in order for cross-module linking to work)... I have a project with two modules, module B is dependent on module A. Module A uses the same context (MyContext) in two different files MyContextComputation.m and MyContextObject.m. They both have a public section where they declare the usage messages of functions defined in those files. In Module B, functions defined in MyContextComputation.m correctly get underlined as coming from a different file but the functions defined in MyContextObject.m get a wriggly line with "Could not resolve symbol definition". I think the way I have it structured is not entirely unusual (although I don't know if having non-private contexts in multiple files is standard or not). Is this a known limitation, anything else I should try to tickle some indexing or similar?
Gerli
@gerli
Sorry, the other way around -- it's detecting the functions from MyContextObject.m (so the file that is later in alphabetical ordering) but not from MyContextComputation.m
Gerli
@gerli
And in a different project that I have, I also have the same context used across multiple files and I'm using the same private symbol in multiple files (defined in one and used in many) -- is there any way to get the correct linking working here at all?
Gerli
@gerli
I found an answer to that last question myself -- it works if I add a usage message to the private function (easy enough to do and makes sense)
Gerli
@gerli
And I figured out the first question too -- the two different files thing is a red herring. In the MyContextComputation file I don't have explicit symbol::usage = ... settings, I've factored it out into a addUsageMessage[symbol, "substring used in a template"] because I have quite many very similarly structured functions and I wanted a nice long usage message for all of them. And I guess the linking relies of this symbol::usage = ... line, yes?
Patrick Scheibe
@halirutan
@gerli To give some context: It's not possible to do correct resolving of symbols because for that, one would need to evaluate the code. Only when the kernel runs the code, we can be certain which symbols resolves to which definition. This is not possible in the IDE and we need to make crude assumptions to get a practical solution. Even getting the context of a symbol right is tricky at best and impossible in other cases.
What the plugin does is looking for symbol::usage = as you have correctly determined. This leads to the following things
  1. You need to give a symbol a usage message if you want it to be found in a different file
  2. Since I'm not checking the context where a symbol lives, you can also give private symbols a usage message and it will be found in a different package. So you have to be careful to import the correct context.
  3. If you have symbols that don't do anything (think of Node or Leaf in a tree) and IDEA warns you that it doesn't find a "definition", you can simply give it a usage message to make it resolve (which is both good programming style and informative for people reading your code)
Gerli
@gerli
Makes sense, thanks for the clarification!
Gerli
@gerli

Whether I defined this with or without full context in knownSymbols.m, it’s not recognized when written with full context in MATLink.m

This is from a discussion in October. I'm stumbling on this same issue -- symbols that I reference with their full context (usually because of pulling in symbols from a private context of some file) don't get the linking, even if I add a usage message explicitly with the full context. Should I create a bug report for it?

Patrick Scheibe
@halirutan
@gerli No need for a bug-report, because I'm aware of this and unfortunately, it's exactly where reality hits the hardest. To be specific: If you enter Foo`bar`fun[3] then of course the plugin knows that you mean fun in the context Foo`bar` . However, at the position where you defined fun[x_]:=.. inside your package, I would need to evaluate the whole code to find that when the symbol fun is put into the symbol-table, it is put into the context Foo`bar` .
Also, the new style package format makes this entirely hidden from the user and the rules what is done to $Context and $ContextPath during loading of a file changed with the Mathematica version.
Patrick Scheibe
@halirutan
To give an unrealistic example, why it is required to really have a Mathematica Kernel that runs the code in order to know where a symbol lives look at this:
myContext = "C" <> ToString[1 + 1] <> "`";
Begin[myContext];
var = 3;
End[]
While this example is extreme and no one would do this in practice, we have similar things all over the place that might look innocent at first. Consider for instance the commonly used idiom
If[isMessageDefined, func::usage = "blub"]
which is often used at the top of packages. Evaluating the If and knowing that the usage message of func is defined makes all the difference between exporting a symbol from a package.
Patrick Scheibe
@halirutan
In conclusion, only by inspecting the code and not evaluating it, we will not be able to know exactly in which context a symbol lives and the plugin simply assumes that definitions are available with their name without looking at the contexts.
Gerli
@gerli
@halirutan sorry for the late reply but I mean that would you expect the current linking to work if you have the full context in the usage message definition, ie
Foo`bar`fun::usage = "fun does this"
Szabolcs Horvát
@szhorvat
Just updated to IDEA 2019.3, nice to see that the plugin had an immediately available update too!
Szabolcs Horvát
@szhorvat
@halirutan Not sure where all this Global` stuff comes from:
image.png
I’ll take a closer look later, but need to finish implementing this function first
Patrick Scheibe
@halirutan
@szhorvat That's odd. I'm currently banging my head with halirutan/Wolfram-Language-IntelliJ-Plugin-Archive#146 and I think I resolved. But now I need to check if it broke something since it's really deep inside the parser.
Szabolcs Horvát
@szhorvat
@halirutan Wasn’t the following bug fixed already? I really thought it was fixed. Did it regress?
image.png
Patrick Scheibe
@halirutan
and please ping me with any problems.
Szabolcs Horvát
@szhorvat
@halirutan I can’t find how to install the beta. I already had the beta repository added, but it’s not offering the 2020 version. Therefore I removed the plugin, I removed the respository, and I re-added it. But it’s still not offering anything but 2019.3

The instructions say:

you can install the beta-version for “Wolfram Language” in the Browse Repositories window

Where is the Browse Repositories window? Could you show me a screenshot? I couldn’t find this. I was searching in the Marketplace, which offers 2019.3. Maybe this is why I couldn’t find 2020, maybe it’s in a different place?
image.png
Szabolcs Horvát
@szhorvat
@halirutan Do I need to have IDEA 2020 EAP to be able to install the beta plugin?
I can install the EAP, if that's needed to test the plugin, no problem.
Patrick Scheibe
@halirutan
@szhorvat Yes. I tried to fix as many deprication/removal API usages as possible and unfortunately, some of the new API changes don't work in 2019.3.
Usually, I don't try to be such an early adopter of changes, but right now Jetbrains does a lot of high quality cleanup. All these things make sense and make the API more stable, and, eventually, we need to implement them anyway.
Szabolcs Horvát
@szhorvat
The font changes to something rather unusual looking. It’s not unpleasant, but I wanted to make sure it changed for you too, as I’ve been messing with the fonts in previous versions.
Never mind, I should have read the blog post