-
Notifications
You must be signed in to change notification settings - Fork 4
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Test-abort error for Tablicious, maybe around classes in namespaces, or categorical.missing #283
Comments
Edit: Nevermind. I mixed up the +folder and the .m file of similar names. |
Yeah, there's all of One thing the I think it's something to do with Octave's
But with Tablicious loaded,
On eilonwy, my macOS 14 M1 Mac (switched to that today), I loaded in octave-doctest (from its current HEAD of main, eb9429f) and did
Different stacktrace than for Same new stacktrace if I run it using the published OF doctest package, with
...aaaand I can reproduce the stacktrace from my original post here if I fire up a new Octave session and just forget to actually add Tablicious to the Octave path before running cd-ing to it and running
So that part at least is user error on my part. I'll update this ticket's description with a fixed repro. Looks like here it's cd-ing around, it's operating on and not recognizing that
I think at that point, because of the
Now,
So I think what's happening here (in my user-error case, not overall) is Still would like to figure out the "regular" error case when Tablicious is properly loaded. My ticket for handling the "dataset" naming ambiguity: |
A bit of diagnosis, stopped at that "error: unknown package 'categorical'" error inside
Kinda odd.
I get the same behavior when I fire up a fresh session and do the same I tried adding a texinfo block to my categorical.circshift.
Same behavior; it returned the helptext for the core circshift.
To see if it was specific to the name, I added a duplicate circshift method, with just the name changed and a "dupe" note added to its helptext.
For that "circshift2" one, I got my helptext.
SpeculationLooks to me like core Octave's Might be hard to do a full fix. Maybe |
Sorry this is a bit TL-DR... apologies for skimming.
|
See discussion in #283. Completely untested so far.
See discussion in #283. Completely untested so far.
Hi, doctest folks,
I'm working on the Tablicious package, and I'd like to make it "doctest clean", so it passes doctest, even if that doesn't do significant exercising of its code.
When running
doctest inst
against Tablicious HEAD of main a bit after v0.4.2, I get this "class not found: tblish.internal.dataset" error that seems to abort the whole doctest run.Repro code:
I get:
Note that it ends up in a different directory than I started in, too. With a
dbstop if error
, stopped at that same place:I'm thinking maybe this is a problem with support for methods in classes, classes in namespaces, some combination of both, or maybe I'm just writing my texinfo sections wrong?
I ran this against commit f0e346ab97eba11462d510305550bc371e797f4d in Tablicious (apjanke/octave-tablicious@f0e346a).
I'm trying to make a minimal repro case in branch WIP/repro-categorical-missing-error in repo apjanke/octo-docto-repro, but haven't had luck yet.
References
Here's my own ticket in Tablicious about ~this: apjanke/octave-tablicious#129
@mmuetzel originally reported something like this in gnu-octave/packages#401. Though there it was failing on something different, and in some of my other tests, it is choking on
categorical.missing
(a static method in Tablicious'scategorical
class) instead of thetblish.internal.dataset
error I'm getting here.The text was updated successfully, but these errors were encountered: