Terry Jan Reedy [Mon, 16 Jun 2014 23:01:01 +0000 (19:01 -0400)]
Issue #21686: add unittest for idlelib.HyperParser. Original patch by Saimadhav
Heblikar. Correct a minor 3.x bug in HyperParser discovered by testing.
Victor Stinner [Mon, 16 Jun 2014 15:11:05 +0000 (17:11 +0200)]
Sync asyncio with Tulip: Fix test_tasks for Python 3.5
On Python 3.5, generator now gets their name from the function, no more from
the code. So we get the expected "notmuch" name instead of the generic "coro"
name.
Gregory P. Smith [Mon, 16 Jun 2014 00:51:04 +0000 (17:51 -0700)]
Isolate the subprocess test_close_fds_when_max_fd_is_lowered test so
that the rlimit calls happens in a child process rather than the
TestCase process to attempt to fix the gentoo buildbot's "Too many
open files" error.
Zachary Ware [Fri, 13 Jun 2014 18:44:39 +0000 (13:44 -0500)]
Issue #19493: Refactor ctypes test package.
Skipped tests are now marked as skipped, formerly commented-out or
renamed-so-it-doesn't-look-like-a-test tests are uncommented, properly named,
and unconditionally skipped, some tests that simply didn't run before
are now able to run, and a few are split into multiple methods instead of
skipping via 'return' in the middle of a method. Also, a couple of unused
files are removed completely.
Victor Stinner [Thu, 12 Jun 2014 16:39:26 +0000 (18:39 +0200)]
asyncio: Tulip issue 173: Enhance repr(Handle) and repr(Task)
repr(Handle) is shorter for function: "foo" instead of "<function foo at
0x...>". It now also includes the source of the callback, filename and line
number where it was defined, if available.
repr(Task) now also includes the current position in the code, filename and
line number, if available. If the coroutine (generator) is done, the line
number is omitted and "done" is added.
Victor Stinner [Tue, 10 Jun 2014 08:23:10 +0000 (10:23 +0200)]
Issue #21326: Add a new is_closed() method to asyncio.BaseEventLoop
Add BaseEventLoop._closed attribute and use it to check if the event loop was
closed or not, instead of checking different attributes in each subclass of
BaseEventLoop.
run_forever() and run_until_complete() methods now raise a RuntimeError('Event loop is
closed') exception if the event loop was closed.
BaseProactorEventLoop.close() now also cancels "accept futures".
Terry Jan Reedy [Tue, 10 Jun 2014 06:49:35 +0000 (02:49 -0400)]
Issue #21695: Catch AttributeError created when user closes grep output window
while still being written to. With no console, this closed Idle.
Also add missing import and a few other changes.
Zachary Ware [Thu, 5 Jun 2014 16:44:22 +0000 (11:44 -0500)]
Clean up Tcl/Tk building in the Windows buildbot scripts.
- Nix the 'noxp' option, it was for Win2k compatibility (which was dropped
with Python 3.3) and made default ttk ugly on post-Win2k systems (#21665)
- Use 'OPTS=symbols' instead of 'DEBUG=1'; symbols is the official method
- Use core, shell, dlls, install-binaries, and install-libraries targets
instead of all and install to avoid trying to 1) build packages that don't
work with MSVC 10+ and 2) install unnecessary bits and pieces.
Victor Stinner [Tue, 3 Jun 2014 22:11:52 +0000 (00:11 +0200)]
Issue #21119: asyncio now closes sockets on errors
Fix ResourceWarning: create_connection(), create_datagram_endpoint() and
create_unix_server() methods of event loop now close the newly created socket
on error.
Victor Stinner [Mon, 2 Jun 2014 12:16:44 +0000 (14:16 +0200)]
Issue #21634: Fix pystone micro-benchmark: use floor division instead of true
division to benchmark integers instead of floating point numbers. Set pystone
version to 1.2. Patch written by Lennart Regebro.
Attempt to fix the "too many open files" errors on several of the
buildbots that the new test_close_fds_when_max_fd_is_lowered test causes.
It now leaves 10 more low fd's available.
Don't restrict ourselves to a "max" fd when closing fds before exec()
when we have a way to get an actual list of all open fds from the OS.
Fixes issue #21618: The subprocess module would ignore fds that were
inherited by the calling process and already higher than POSIX resource
limits would otherwise allow. On systems with a functioning /proc/self/fd
or /dev/fd interface the max is now ignored and all fds are closed.