(
Jump to the original submission)
Fri 18 Aug 2017 06:50:06 PM UTC, comment #8:
Good catch Kai. If you think that's a feature worth restoring at some point go ahead and file a new bug.
|
Thu 17 Aug 2017 08:33:27 PM UTC, comment #7:
I just want to document one minor regression regarding command-line functions. The solution is to run the script (like in the example below) prior to publishing it, as the necessary command-line functions seem to be present then.
At least we are still "better" than Matlab regarding command-line functions, such a thing does not work there.
|
Wed 16 Aug 2017 02:08:03 PM UTC, comment #6:
Thank you Mike, it works perfectly for me! A shame we cannot make a BIST for this bug, but it is only publish...
I think we can close this issue now.
|
Wed 16 Aug 2017 12:00:17 AM UTC, comment #5:
I have a fix that seems to work here, but "clear all" still can't be used in Octave's test suite because global variables are used, and "clear all" wipes out any globals.
I pushed this on the default branch without a test, please test if you are affected by this, and let me know if it works or if I broke anything.
https://hg.savannah.gnu.org/hgweb/octave/rev/ab1aa0e57b72
|
Tue 15 Aug 2017 06:44:00 PM UTC, comment #4:
I'm working on a change to fix this, to use an inline function definition similar to the way the test function works to create local scopes. If that works for me, I think I can get rid of the triple-underscore variable names as well, since they won't be needed for disambiguation.
|
Tue 15 Aug 2017 06:28:27 PM UTC, comment #3:
Regarding comment #2 that is true. Compared to Matlab my initial publish implementation only works with scripts and has a "primitive" evaluation. But it is pure m-code! Is there a way to create a new context, except my idea of using a functions context?
The problem of the original submission was discussed on the mailing list, just as a pointer: http://octave.1599824.n4.nabble.com/Publish-td4683415.html.
|
Tue 15 Aug 2017 05:45:42 PM UTC, comment #2:
The publish function no longer errors on this example, but it does still produce a spurious output in the resulting html:
Unfortunately this is because the subfunction that evaluates the script in question has local variables, and the script is run in the same context as that subfunction.
|
Wed 07 Jun 2017 12:18:42 AM UTC, comment #1:
I'm currently doing some significant refactoring of the symbol table, so maybe this will be easier to fix once that is done.
|
Tue 23 May 2017 12:52:31 PM UTC, original submission:
Trying to publish a script that calls makes publish fail with the error
This is due to bug #35881
Here a short example
|
(Note: upload size limit is set to 16384 kB, after insertion of the required
escape characters.)
Attach File(s):
Comment:
No files currently attached
Depends on the following items
Digest:
bug dependencies.
Items that depend on this one: None found
Follow 8 latest changes.