[issue2988] Invalid cookies crash web applications
Shish added the comment: I'm having problems with this too -- a third party app on the same domain as me has set an invalid cookie, and now my app crashes horribly :( (And even if cherrypy handled the exception and didn't crash completely, it would still not be able to use any cookies) -- nosy: +shish2k ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue2988 ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue3949] curses' sigwinch handler isn't visible from python
Shish shish+pyt...@shishnet.org added the comment: in its simplest form, this shows the bug -- signal.getsignal does not return the original signal handler (the tell the app to resize itself one that curses installs), and so, the original signal handler cannot be re-activated after being temporarily disabled ~~ import curses import signal curses.initscr() old_winch = signal.getsignal(signal.SIGWINCH) curses.endwin() print old_winch -- ___ Python tracker rep...@bugs.python.org http://bugs.python.org/issue3949 ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue3948] readline steals sigwinch
New submission from Shish [EMAIL PROTECTED]: I have an app which wants to use a mostly curses interface with some parts readline, however, doing so much as import readline causes readline to claim ownership of sigwinch, thus breaking the ability of the app to resize. Worse, it seems to claim it at the C level -- doing signal.getsignal (signal.SIGWINCH) returns SIG_DFL, so I can't see anything I can do at the python level to manually set the signals to be handled in the way I want :-/ I would think it best to set the handler at the start (for compatability, and because the vast majority of apps will expect it that way), but set it via python's signal module, so that app writers can take the signal handler and call it in their own way (In my case, I want the app to listen for the signal, and the app's handler will call readline's and curses's handlers in turn) -- components: Extension Modules messages: 73667 nosy: shish severity: normal status: open title: readline steals sigwinch type: behavior versions: Python 2.4 ___ Python tracker [EMAIL PROTECTED] http://bugs.python.org/issue3948 ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue3949] curses' sigwinch handler isn't visible from python
New submission from Shish [EMAIL PROTECTED]: after the first initscr() sigwinch is handled as expected, but then my app needs to ignore sigwinch, leave curses to view the output from something, come back into curses, and start listening for sigwinch again. The signal(SIGWINCH, SIG_IGN) call doesn't return the old, working signal handler; it returns 0 -- thus trying to re-set the signal handler doesn't work. (The reason for ignoring the signal is that if the window is resized while the external app is running, python's wait() is interrupted) Having the handler visible from python would also make curses play nicer with readline, as the app could mediate signal handling (see issue #3948) -- components: Extension Modules messages: 73668 nosy: shish severity: normal status: open title: curses' sigwinch handler isn't visible from python versions: Python 2.4 ___ Python tracker [EMAIL PROTECTED] http://bugs.python.org/issue3949 ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue2732] curses.textpad loses characters at the end of lines
Shish [EMAIL PROTECTED] added the comment: Also present in 2.5 -- components: +Library (Lib) -Extension Modules versions: +Python 2.5 __ Tracker [EMAIL PROTECTED] http://bugs.python.org/issue2732 __ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue2732] curses.textpad loses characters at the end of lines
New submission from Shish [EMAIL PROTECTED]: Even in the self-test suite, the bug appears, run: python /usr/lib/python2.4/curses/textpad.py then type so that the line wraps, eg 123456789123456789 to fill two lines, and then ctrl-g to return -- the result has had the end characters of each line removed, so the return value is 12345678\n12345678\n As a quick hack on my local install I changed line 56 from: last = min(self.maxx, last+1) to last = min(self.maxx, last)+1 and it seems to work, but I have no idea if this is the right way to fix it (ie, this might have side effects, or this off-by-one might happen in several places and need a global fix, not just one local one) -- components: Extension Modules messages: 66035 nosy: shish severity: normal status: open title: curses.textpad loses characters at the end of lines versions: Python 2.4 __ Tracker [EMAIL PROTECTED] http://bugs.python.org/issue2732 __ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue2732] curses.textpad loses characters at the end of lines
Changes by Shish [EMAIL PROTECTED]: -- type: - behavior __ Tracker [EMAIL PROTECTED] http://bugs.python.org/issue2732 __ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com