#9640: Change PARI error catching mechanism
----------------------------------------+-------------------------
Reporter: jdemeyer | Owner: was
Type: enhancement | Status: new
Priority: major | Milestone: sage-5.12
Component: interfaces | Resolution:
Keywords: pari error interrupt | Merged in:
Authors: Jeroen Demeyer | Reviewers:
Report Upstream: N/A | Work issues:
Branch: | Commit:
Dependencies: | Stopgaps:
----------------------------------------+-------------------------
Comment (by pbruin):
I am working on a patch that does the following things suggested by the
ticket description:
- Rewrite `pari_catch_sig_on()` and `pari_catch_sig_off()` to use the
callbacks `cb_pari_handle_exception` and `cb_pari_err_recover` (defined by
the PARI library) in combination with `sig_on()`/`sig_off()`, as opposed
to the additional `setjmp()` of the current implementation.
- Catch output sent to `pariErr` and store it in a string attribute
`pari_instance.error_string` for later use by `PariError`.
It seems to work quite well; the only doctest failures I've seen so far
are a few tests expecting a PARI warning message. With the patch, these
are not printed because they also go to `pariErr`.
To be done:
- Make `PariError` aware of the stored string to give a more informative
error message. This should probably simply be extracted from the lines
starting with `***` in `pari_instance.error_string`.
- Minor change in PARI: send any extra newline at the start of the error
message to `pariErr` instead of `pariOut` (this happens if the last
printed character was not a newline).
- Distinguish between errors and warnings; the latter should probably be
printed to `stderr` as before. This should be done either by filtering
the text sent to `pariErr` or by having `pari_catch_sig_off()` check if
`pari_instance.error_string` is non-empty even if no error occurred.
This may make #14894 redundant, although that ticket could be used for
general things to be done for the upgrade to PARI 2.6.
--
Ticket URL: <http://trac.sagemath.org/ticket/9640#comment:6>
Sage <http://www.sagemath.org>
Sage: Creating a Viable Open Source Alternative to Magma, Maple, Mathematica,
and MATLAB
--
You received this message because you are subscribed to the Google Groups
"sage-trac" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/sage-trac.
For more options, visit https://groups.google.com/groups/opt_out.
