Comment 1 for bug 913905

Revision history for this message
Richard Everson (r-m-everson) wrote : Re: [Bug 913905] Re: py-up-exception remaining

Hi Andreas,

Sorry not to have got back to you sooner. The problem seems to be that with IPython the file is not listed as "<stdin>", but as, eg, "/Users/reverson/projects/edge/<ipython-input-2-1cebb13502c5> in <module>()", so the check in py-find-next-exception needs to check this. Here's a possible patch (against revno 772) which works for me -- I haven't checked the line number which the "<stdin>" check does.

ax:Emacs/el/python-mode/$ bzr log |head
------------------------------------------------------------
revno: 772
fixes bug(s): https://launchpad.net/bugs/913905
committer: Andreas Roehler <email address hidden>
branch nick: python-mode
timestamp: Tue 2012-01-10 13:48:47 +0100
message:
  `py-up-exception' again, lp:913905
------------------------------------------------------------
revno: 771
ax:Emacs/el/python-mode/$ bzr diff -p1
=== modified file 'python-mode.el'
--- old/python-mode.el 2012-01-10 12:48:47 +0000
+++ new/python-mode.el 2012-01-10 18:09:23 +0000
@@ -6110,7 +6110,8 @@
                   (setq file (substring file 0 -1)))
                 (when (string= errwhere "Bottom") (goto-char pos))
                 (if (and file line)
- (if (and (string= "<stdin>" file) (eq 1 line))
+ (if (or (and (string= "<stdin>" file) (eq 1 line))
+ (string-match "<ipython-input-" file))
                         (error "%s of traceback" errwhere)
                       (py-jump-to-exception file line py-line-number-offset))
                   (error "%s of traceback" errwhere)))

Best,
Richard.

On 10 Jan 2012, at 13:07, Andreas Roehler wrote:

> ** Changed in: python-mode
> Status: In Progress => Fix Committed
>
> --
> You received this bug notification because you are subscribed to the bug
> report.
> https://bugs.launchpad.net/bugs/913905
>
> Title:
> py-up-exception remaining
>
> Status in An Emacs mode for editing Python code:
> Fix Committed
>
> Bug description:
> Hi Andreas,
>
> A good improvement. There's still a minor bug with IPython and moving
> up the stacks with C-c -. In the demo1.py example, pressing C-c -
> repeatedly goes up the stack until the
>
> ----> 1 demo1.f()
>
> line is reached, at which point string-match says "wrong-type-argument
> stringp nil"; here's the backtrace:
>
> Debugger entered--Lisp error: (wrong-type-argument stringp nil)
> string-match(".+.pyc" nil)
> (if (string-match ".+.pyc" file) (progn (setq file ...)))
> (when (string-match ".+.pyc" file) (setq file (substring file 0 -1)))
> (progn (setq py-last-exeption-buffer (current-buffer)) (if (save-match-data ...) (progn ...) (setq line ... pos ... file ...)) (when (string-match ".+.pyc" file) (setq file ...)) (when (string= errwhere "Bottom") (goto-char pos)) (if (and file line) (if ... ... ...) (error "%s of traceback" errwhere)))
> (if (not (save-match-data ...)) (progn (setq py-last-exeption-buffer ...) (if ... ... ...) (when ... ...) (when ... ...) (if ... ... ...)) (goto-char orig) (error "%s of traceback" errwhere))
> (if (save-match-data (eq ... origline)) (progn (forward-line ...) (py-find-next-exception start buffer searchdir errwhere)) (if (not ...) (progn ... ... ... ... ...) (goto-char orig) (error "%s of traceback" errwhere)))
> (if (funcall searchdir py-traceback-line-re nil t) (if (save-match-data ...) (progn ... ...) (if ... ... ... ...)))
> (let ((orig ...) (origline ...) file line pos) (goto-char (py-point start)) (if (funcall searchdir py-traceback-line-re nil t) (if ... ... ...)))
> py-find-next-exception(bol "*IPython*" re-search-backward "Top")
> (if (string= start "TOP") (py-find-next-exception (quote bob) buffer (quote re-search-forward) "Top") (py-find-next-exception (quote bol) buffer (quote re-search-backward) "Top"))
> (if (eq direction (quote up)) (if (string= start "TOP") (py-find-next-exception ... buffer ... "Top") (py-find-next-exception ... buffer ... "Top")) (if (string= start "BOTTOM") (py-find-next-exception ... buffer ... "Bottom") (py-find-next-exception ... buffer ... "Bottom")))
> (let* ((name ...) (buffer ...)) (when buffer (set-buffer ...)) (switch-to-buffer (current-buffer)) (if (eq direction ...) (if ... ... ...) (if ... ... ...)))
> py-find-next-exception-prepare(up nil)
> py-up-exception(nil)
> call-interactively(py-up-exception nil nil)
>
>
> At this point the cursor in the *IPython* buffer is sitting at the start
> of the "----> 1 demo1.f()" line. It's not much of a problem, because on
> pressing C-c - again, the mini-buffer reports, "Top of traceback".
>
>>>>> Secondly, the SyntaxError exceptions in the IPython buffer don't get
>>>>> matched. Pressing C-c - does nothing with the following:
>>>>>
>>>>> In [1]: import demo2
>>>>> File "demo3.py", line 4
>>>>> if a != 25
>>>>> ^
>>>>> SyntaxError: invalid syntax
>>>>>
> Works well for me!
>
>
>>>>> One thing that would be very useful would be to limit the searching for exception reporting lines to the *last* set of exceptions. If you have a long Python or IPython session there will probably be several exceptions and pressing C-c - repeatedly, not only goes up the stack for the last exception, but then revisits the stack for the previous one, and so on until the beginning of the buffer is reached. I don't know if it's feasible to narrow the search just to the last exception.
> This too works well for me with Python, but with IPython presssing C-c -
> repeatedly still goes back to previous stack traces. In case it helps,
> I've attached the contents my *IPython* buffer that does this.
>
> Thanks again.
> Cheers,
> Richard.
>
>
> ** Attachment added: "*IPython*"
> https://bugs.launchpad.net/bugs/913428/+attachment/2664701/+files/%2AIPython%2A
>
> -- You received this bug notification because you are subscribed to
> the bug report. https://bugs.launchpad.net/bugs/913428 Title: py-down-
> exception IPython issues To manage notifications about this bug go to:
> https://bugs.launchpad.net/python-mode/+bug/913428/+subscriptions
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/python-mode/+bug/913905/+subscriptions