]> granicus.if.org Git - python/log
python
9 years agoasyncio: Error if awaiting in parallel on the same coroutine
Yury Selivanov [Wed, 18 Nov 2015 17:40:26 +0000 (12:40 -0500)]
asyncio: Error if awaiting in parallel on the same coroutine

See https://github.com/python/asyncio/pull/293 for details.

9 years agoasyncio: Error if awaiting in parallel on the same coroutine
Yury Selivanov [Wed, 18 Nov 2015 17:39:45 +0000 (12:39 -0500)]
asyncio: Error if awaiting in parallel on the same coroutine

This change won't do anything in CPython 3.4

See https://github.com/python/asyncio/pull/293 for details.

9 years agoIssue #23200: Merge zlib doc from 3.4 into 3.5
Martin Panter [Wed, 18 Nov 2015 02:45:51 +0000 (02:45 +0000)]
Issue #23200: Merge zlib doc from 3.4 into 3.5

9 years agoIssue #23200: Document that max_length=0 is not supported
Martin Panter [Wed, 18 Nov 2015 00:59:17 +0000 (00:59 +0000)]
Issue #23200: Document that max_length=0 is not supported

9 years agoIssue #20468: Merge getpagesize doc from 3.4 into 3.5
Martin Panter [Tue, 17 Nov 2015 22:23:21 +0000 (22:23 +0000)]
Issue #20468: Merge getpagesize doc from 3.4 into 3.5

9 years agoIssue #20468: Remove incorrect information about maxrss and page size
Martin Panter [Tue, 17 Nov 2015 22:13:47 +0000 (22:13 +0000)]
Issue #20468: Remove incorrect information about maxrss and page size

Extract of patch by Ronald Oussoren.

9 years agoMerge 3.4
Yury Selivanov [Tue, 17 Nov 2015 17:20:02 +0000 (12:20 -0500)]
Merge 3.4

9 years agoasyncio: Cleanup Future API
Yury Selivanov [Tue, 17 Nov 2015 17:19:41 +0000 (12:19 -0500)]
asyncio: Cleanup Future API

See https://github.com/python/asyncio/pull/292 for details.

9 years agoMerge 3.4
Yury Selivanov [Tue, 17 Nov 2015 17:02:32 +0000 (12:02 -0500)]
Merge 3.4

9 years agoasyncio: Sync with github
Yury Selivanov [Tue, 17 Nov 2015 17:02:21 +0000 (12:02 -0500)]
asyncio: Sync with github

9 years agoIssue #25615: Merge glob doc from 3.4 into 3.5
Martin Panter [Tue, 17 Nov 2015 00:19:10 +0000 (00:19 +0000)]
Issue #25615: Merge glob doc from 3.4 into 3.5

9 years agoIssue #25615: Document unsorted behaviour of glob; patch by Dave Jones
Martin Panter [Mon, 16 Nov 2015 23:46:22 +0000 (23:46 +0000)]
Issue #25615: Document unsorted behaviour of glob; patch by Dave Jones

9 years agoMerge 3.4
Yury Selivanov [Mon, 16 Nov 2015 20:12:44 +0000 (15:12 -0500)]
Merge 3.4

9 years agoasyncio: Optimize Task._wakeup
Yury Selivanov [Mon, 16 Nov 2015 20:12:10 +0000 (15:12 -0500)]
asyncio: Optimize Task._wakeup

See https://github.com/python/asyncio/pull/289 for details.

9 years agoasyncio.docs: Document Transport.is_closing (merge 3.4)
Yury Selivanov [Mon, 16 Nov 2015 17:47:15 +0000 (12:47 -0500)]
asyncio.docs: Document Transport.is_closing (merge 3.4)

9 years agoasyncio.docs: Document Transport.is_closing
Yury Selivanov [Mon, 16 Nov 2015 17:46:41 +0000 (12:46 -0500)]
asyncio.docs: Document Transport.is_closing

9 years agoMerge 3.4
Yury Selivanov [Mon, 16 Nov 2015 17:43:35 +0000 (12:43 -0500)]
Merge 3.4

9 years agoasyncio: Add Transport.is_closing()
Yury Selivanov [Mon, 16 Nov 2015 17:43:21 +0000 (12:43 -0500)]
asyncio: Add Transport.is_closing()

See https://github.com/python/asyncio/pull/291 for details.

9 years agoIssue #25498: Fixed contributors name.
Serhiy Storchaka [Mon, 16 Nov 2015 16:45:39 +0000 (18:45 +0200)]
Issue #25498: Fixed contributors name.

9 years agoIssue #25498: Fixed contributors name.
Serhiy Storchaka [Mon, 16 Nov 2015 16:45:23 +0000 (18:45 +0200)]
Issue #25498: Fixed contributors name.

9 years agoMerge with 3.4
Terry Jan Reedy [Mon, 16 Nov 2015 12:36:33 +0000 (07:36 -0500)]
Merge with 3.4

9 years agoIssue #24750: whitespace
Terry Jan Reedy [Mon, 16 Nov 2015 12:36:22 +0000 (07:36 -0500)]
Issue #24750: whitespace

9 years agoMerge with 3.4
Terry Jan Reedy [Mon, 16 Nov 2015 12:32:46 +0000 (07:32 -0500)]
Merge with 3.4

9 years agoIssue #24750: Improve appearance of IDLE editor window status bar.
Terry Jan Reedy [Mon, 16 Nov 2015 12:32:26 +0000 (07:32 -0500)]
Issue #24750: Improve appearance of IDLE editor window status bar.
Patch by Mark Roseman.

9 years agoIssue #20220: Merge time zone workaround from 3.4 into 3.5
Martin Panter [Mon, 16 Nov 2015 09:42:41 +0000 (09:42 +0000)]
Issue #20220: Merge time zone workaround from 3.4 into 3.5

9 years agoIssue #9051: Added tests for pickling and copying the timezone objects.
Serhiy Storchaka [Mon, 16 Nov 2015 09:19:31 +0000 (11:19 +0200)]
Issue #9051: Added tests for pickling and copying the timezone objects.

9 years agoIssue #9051: Added tests for pickling and copying the timezone objects.
Serhiy Storchaka [Mon, 16 Nov 2015 09:12:58 +0000 (11:12 +0200)]
Issue #9051: Added tests for pickling and copying the timezone objects.

9 years agoIssue #20220: Add DST rules to work around glibc quirk
Martin Panter [Mon, 16 Nov 2015 07:21:38 +0000 (07:21 +0000)]
Issue #20220: Add DST rules to work around glibc quirk

This was triggering intermittent failures in unrelated tests.

9 years agomake the PyUnicode_FSConverter cleanup set the decrefed argument to NULL (closes...
Benjamin Peterson [Mon, 16 Nov 2015 05:57:39 +0000 (21:57 -0800)]
make the PyUnicode_FSConverter cleanup set the decrefed argument to NULL (closes #25630)

9 years agoFixes #23564: Fix a partially broken sanity check in the _posixsubprocess
Gregory P. Smith [Mon, 16 Nov 2015 05:15:26 +0000 (21:15 -0800)]
Fixes #23564: Fix a partially broken sanity check in the _posixsubprocess
internals regarding how fds_to_pass were passed to the child.  The bug
had no actual impact as subprocess.py already avoided it.

9 years agoFix issue #6973: When we know a subprocess.Popen process has died, do
Gregory P. Smith [Mon, 16 Nov 2015 02:26:11 +0000 (18:26 -0800)]
Fix issue #6973: When we know a subprocess.Popen process has died, do
not allow the send_signal(), terminate(), or kill() methods to do
anything as they could potentially signal a different process.

9 years agoFix issue #6973: When we know a subprocess.Popen process has died, do
Gregory P. Smith [Mon, 16 Nov 2015 02:19:10 +0000 (18:19 -0800)]
Fix issue #6973: When we know a subprocess.Popen process has died, do
not allow the send_signal(), terminate(), or kill() methods to do
anything as they could potentially signal a different process.

9 years agomerge 3.4 (#25578)
Benjamin Peterson [Sat, 14 Nov 2015 23:14:42 +0000 (15:14 -0800)]
merge 3.4 (#25578)

9 years agorm trailing ws
Benjamin Peterson [Sat, 14 Nov 2015 23:12:38 +0000 (15:12 -0800)]
rm trailing ws

9 years agofix possible memory lea k in _get_aia_uri (closes #25578)
Benjamin Peterson [Sat, 14 Nov 2015 23:12:18 +0000 (15:12 -0800)]
fix possible memory lea k in _get_aia_uri (closes #25578)

9 years agoIssue #25388: Fixed tokenizer crash when processing undecodable source code
Serhiy Storchaka [Sat, 14 Nov 2015 13:11:17 +0000 (15:11 +0200)]
Issue #25388: Fixed tokenizer crash when processing undecodable source code
with a null byte.

9 years agoIssue #25388: Fixed tokenizer crash when processing undecodable source code
Serhiy Storchaka [Sat, 14 Nov 2015 13:10:35 +0000 (15:10 +0200)]
Issue #25388: Fixed tokenizer crash when processing undecodable source code
with a null byte.

9 years agomerge 3.4 (#25569)
Benjamin Peterson [Sat, 14 Nov 2015 08:11:09 +0000 (00:11 -0800)]
merge 3.4 (#25569)

9 years agofix build with older openssl (#25569)
Benjamin Peterson [Sat, 14 Nov 2015 08:09:22 +0000 (00:09 -0800)]
fix build with older openssl (#25569)

9 years agoFix grammar in whatsnew/3.5.rst and elsewhere
Martin Panter [Sat, 14 Nov 2015 01:14:54 +0000 (01:14 +0000)]
Fix grammar in whatsnew/3.5.rst and elsewhere

Mainly missing grammatical articles (the, a).

9 years agoMerge typo and grammar fixes from 3.4 into 3.5
Martin Panter [Sat, 14 Nov 2015 01:14:25 +0000 (01:14 +0000)]
Merge typo and grammar fixes from 3.4 into 3.5

9 years agoFix a few grammar problems in the documentation and comments
Martin Panter [Sat, 14 Nov 2015 01:07:43 +0000 (01:07 +0000)]
Fix a few grammar problems in the documentation and comments

9 years agoCorrect Content-Type syntax in documentation
Martin Panter [Sat, 14 Nov 2015 00:58:32 +0000 (00:58 +0000)]
Correct Content-Type syntax in documentation

9 years agoIssue #25590: Merge rlcompleter change from 3.4 into 3.5
Martin Panter [Fri, 13 Nov 2015 23:10:39 +0000 (23:10 +0000)]
Issue #25590: Merge rlcompleter change from 3.4 into 3.5

9 years agoIssue #25590: Make rlcompleter only call getattr() once per attribute
Martin Panter [Fri, 13 Nov 2015 22:47:00 +0000 (22:47 +0000)]
Issue #25590: Make rlcompleter only call getattr() once per attribute

Previously it was called another time via hasattr(), and both calls were
made once for dir(f) and again for dir(f.__class__).  This includes a
backport of changing from a list to a set from revision 4dbb315fe667.

9 years agoIssue #25498: Update error message for 3.5
Martin Panter [Fri, 13 Nov 2015 22:12:58 +0000 (22:12 +0000)]
Issue #25498: Update error message for 3.5

9 years agoIssue #25498: Merge ctypes crash fix from 3.4 into 3.5
Martin Panter [Fri, 13 Nov 2015 22:12:12 +0000 (22:12 +0000)]
Issue #25498: Merge ctypes crash fix from 3.4 into 3.5

9 years agoIssue #25498: Fix GC crash due to ctypes objects wrapping a memoryview
Martin Panter [Fri, 13 Nov 2015 21:43:39 +0000 (21:43 +0000)]
Issue #25498: Fix GC crash due to ctypes objects wrapping a memoryview

This was a regression caused by revision 1da9630e9b7f.  Based on patch by
Eryksun.

9 years agoMerge 3.4
Yury Selivanov [Fri, 13 Nov 2015 17:29:03 +0000 (12:29 -0500)]
Merge 3.4

9 years agoasyncio: Fix sporadic failing unittests in debug mode
Yury Selivanov [Fri, 13 Nov 2015 17:28:48 +0000 (12:28 -0500)]
asyncio: Fix sporadic failing unittests in debug mode

9 years agoIssue #25462: The hash of the key now is calculated only once in most
Serhiy Storchaka [Fri, 13 Nov 2015 12:48:36 +0000 (14:48 +0200)]
Issue #25462: The hash of the key now is calculated only once in most
operations in C implementation of OrderedDict.

9 years agoMerge 3.4 (issue #25605)
Victor Stinner [Fri, 13 Nov 2015 08:14:05 +0000 (09:14 +0100)]
Merge 3.4 (issue #25605)

9 years agoIssue #25605: Document exceptions raised by fcntl.ioctl() and fcntl.flock()
Victor Stinner [Fri, 13 Nov 2015 08:13:48 +0000 (09:13 +0100)]
Issue #25605: Document exceptions raised by fcntl.ioctl() and fcntl.flock()

9 years agoMerge with 3.4
Terry Jan Reedy [Thu, 12 Nov 2015 20:24:50 +0000 (15:24 -0500)]
Merge with 3.4

9 years agoIndent
Terry Jan Reedy [Thu, 12 Nov 2015 20:24:33 +0000 (15:24 -0500)]
Indent

9 years agoMerge with 3.4
Terry Jan Reedy [Thu, 12 Nov 2015 20:06:20 +0000 (15:06 -0500)]
Merge with 3.4

9 years agoWhitespace
Terry Jan Reedy [Thu, 12 Nov 2015 20:06:07 +0000 (15:06 -0500)]
Whitespace

9 years agoMerge with 3.4
Terry Jan Reedy [Thu, 12 Nov 2015 20:03:17 +0000 (15:03 -0500)]
Merge with 3.4

9 years agoIssue #25313: Change the handling of new built-in text color themes to better
Terry Jan Reedy [Thu, 12 Nov 2015 20:02:57 +0000 (15:02 -0500)]
Issue #25313: Change the handling of new built-in text color themes to better
address the compatibility problem introduced by the addition of IDLE Dark.
Consistently use the revised idleConf.CurrentTheme everywhere in idlelib.

9 years agoIssue #25607: Restore old distutils logging threshold after running tests that
Serhiy Storchaka [Thu, 12 Nov 2015 17:47:14 +0000 (19:47 +0200)]
Issue #25607: Restore old distutils logging threshold after running tests that
parse command line arguments.

9 years agoIssue #25607: Restore old distutils logging threshold after running tests that
Serhiy Storchaka [Thu, 12 Nov 2015 17:46:23 +0000 (19:46 +0200)]
Issue #25607: Restore old distutils logging threshold after running tests that
parse command line arguments.

9 years agoRestore old distutils logging threshold after running test_log.
Serhiy Storchaka [Thu, 12 Nov 2015 11:15:56 +0000 (13:15 +0200)]
Restore old distutils logging threshold after running test_log.

9 years agoRestore old distutils logging threshold after running test_log.
Serhiy Storchaka [Thu, 12 Nov 2015 11:15:41 +0000 (13:15 +0200)]
Restore old distutils logging threshold after running test_log.

9 years agoIssue #22995: Default implementation of __reduce__ and __reduce_ex__ now
Serhiy Storchaka [Thu, 12 Nov 2015 09:31:51 +0000 (11:31 +0200)]
Issue #22995: Default implementation of __reduce__ and __reduce_ex__ now
rejects builtin types with not defined __new__.
Added tests for non-pickleable types.

9 years agoIssue #22995: Default implementation of __reduce__ and __reduce_ex__ now
Serhiy Storchaka [Thu, 12 Nov 2015 09:23:04 +0000 (11:23 +0200)]
Issue #22995: Default implementation of __reduce__ and __reduce_ex__ now
rejects builtin types with not defined __new__.
Added tests for non-pickleable types.

9 years agomerge 3.4 (#25530)
Benjamin Peterson [Thu, 12 Nov 2015 06:45:22 +0000 (22:45 -0800)]
merge 3.4 (#25530)

9 years agoalways set OP_NO_SSLv3 by default (closes #25530)
Benjamin Peterson [Thu, 12 Nov 2015 06:38:41 +0000 (22:38 -0800)]
always set OP_NO_SSLv3 by default (closes #25530)

9 years agomerge 3.5 (#25569)
Benjamin Peterson [Thu, 12 Nov 2015 06:14:08 +0000 (22:14 -0800)]
merge 3.5 (#25569)

9 years agofix memory leak in _get_crl_dp (closes #25569)
Benjamin Peterson [Thu, 12 Nov 2015 06:07:38 +0000 (22:07 -0800)]
fix memory leak in _get_crl_dp (closes #25569)

Patch started by Stéphane Wirtel.

9 years agoMerge with 3.4
Zachary Ware [Thu, 12 Nov 2015 05:33:50 +0000 (23:33 -0600)]
Merge with 3.4

9 years agoRewrite re.VERBOSE section.
Zachary Ware [Thu, 12 Nov 2015 05:32:14 +0000 (23:32 -0600)]
Rewrite re.VERBOSE section.

It now has slightly better rationale and a less awkward wording in the
explanation.

Inspired by a report of a neither/or conflict by 'animalize' on docs@.

9 years agoIssue #25603: Add missing parenthesis.
Zachary Ware [Thu, 12 Nov 2015 04:59:44 +0000 (22:59 -0600)]
Issue #25603: Add missing parenthesis.

9 years agoFix awkward grammar in whatsnew.
Zachary Ware [Thu, 12 Nov 2015 04:53:47 +0000 (22:53 -0600)]
Fix awkward grammar in whatsnew.

Reported on docs@ by Keith Briggs.

9 years agoIssue #25595: Fixed test_deleted_cwd in test_importlib on AIX.
Serhiy Storchaka [Wed, 11 Nov 2015 06:26:59 +0000 (08:26 +0200)]
Issue #25595: Fixed test_deleted_cwd in test_importlib on AIX.

9 years agoIssue #6598: Avoid clock wrapping around in test_make_msgid_collisions.
Serhiy Storchaka [Tue, 10 Nov 2015 17:51:56 +0000 (19:51 +0200)]
Issue #6598: Avoid clock wrapping around in test_make_msgid_collisions.
Use time.monotonic instead of time.clock.

9 years agoIssue #6598: Avoid clock wrapping around in test_make_msgid_collisions.
Serhiy Storchaka [Tue, 10 Nov 2015 17:50:14 +0000 (19:50 +0200)]
Issue #6598: Avoid clock wrapping around in test_make_msgid_collisions.
Use time.monotonic or time.time instead of time.clock.

9 years agoFix Visual Studio warning.
Stefan Krah [Tue, 10 Nov 2015 17:35:19 +0000 (18:35 +0100)]
Fix Visual Studio warning.

9 years agoIaaue #25598: Fix memory_hex from #9951 for non-contiguous buffers.
Stefan Krah [Tue, 10 Nov 2015 17:17:22 +0000 (18:17 +0100)]
Iaaue #25598: Fix memory_hex from #9951 for non-contiguous buffers.

9 years agoIssue #25263: Fixed the tkinter.test.test_tkinter.test_font.FontTest test
Serhiy Storchaka [Tue, 10 Nov 2015 13:31:38 +0000 (15:31 +0200)]
Issue #25263: Fixed the tkinter.test.test_tkinter.test_font.FontTest test
failure when the test is ran the second time.

The root attribute was set in parent class in setUpClass and then
overridded in child class in tearDownClass.

9 years agoIssue #25263: Fixed the tkinter.test.test_tkinter.test_font.FontTest test
Serhiy Storchaka [Tue, 10 Nov 2015 13:30:35 +0000 (15:30 +0200)]
Issue #25263: Fixed the tkinter.test.test_tkinter.test_font.FontTest test
failure when the test is ran the second time.

The root attribute was set in parent class in setUpClass and then
overridded in child class in tearDownClass.

9 years agopytime.c: rename pygettimeofday_new() to pygettimeofday()
Victor Stinner [Tue, 10 Nov 2015 11:11:39 +0000 (12:11 +0100)]
pytime.c: rename pygettimeofday_new() to pygettimeofday()

I forgot to rename it in my previous refactoring of pytime.c.

9 years agoImprove appearance of example code
Raymond Hettinger [Tue, 10 Nov 2015 08:09:34 +0000 (00:09 -0800)]
Improve appearance of example code

9 years agoRemove confusing section from named tuple recipes.
Raymond Hettinger [Tue, 10 Nov 2015 08:00:00 +0000 (00:00 -0800)]
Remove confusing section from named tuple recipes.

9 years agoAdded missed periods at the ends of sentences in Misc/NEWS.
Serhiy Storchaka [Mon, 9 Nov 2015 22:31:10 +0000 (00:31 +0200)]
Added missed periods at the ends of sentences in Misc/NEWS.

9 years agoAdded missed periods at the ends of sentences in Misc/NEWS.
Serhiy Storchaka [Mon, 9 Nov 2015 22:20:43 +0000 (00:20 +0200)]
Added missed periods at the ends of sentences in Misc/NEWS.

9 years agoFixed markup in Misc/NEWS.
Serhiy Storchaka [Mon, 9 Nov 2015 21:43:56 +0000 (23:43 +0200)]
Fixed markup in Misc/NEWS.

9 years agoFixed markup in Misc/NEWS.
Serhiy Storchaka [Mon, 9 Nov 2015 21:41:39 +0000 (23:41 +0200)]
Fixed markup in Misc/NEWS.

9 years agoIssue #25584: Added "escape" to the __all__ list in the glob module.
Serhiy Storchaka [Mon, 9 Nov 2015 21:18:19 +0000 (23:18 +0200)]
Issue #25584: Added "escape" to the __all__ list in the glob module.
From patch by Xavier de Gaye.

9 years agoIssue #25584: Fixed recursive glob() with patterns starting with '**'.
Serhiy Storchaka [Mon, 9 Nov 2015 21:12:07 +0000 (23:12 +0200)]
Issue #25584: Fixed recursive glob() with patterns starting with '**'.

9 years agoIssue #25582: Fixed 100 MB memory leak in test_ctypes.
Serhiy Storchaka [Mon, 9 Nov 2015 20:31:10 +0000 (22:31 +0200)]
Issue #25582: Fixed 100 MB memory leak in test_ctypes.

9 years agoIssue #24306: Sets component ID for launcher to match 3.5 and later to avoid downgrading.
Steve Dower [Sat, 7 Nov 2015 18:07:04 +0000 (10:07 -0800)]
Issue #24306: Sets component ID for launcher to match 3.5 and later to avoid downgrading.

9 years agoIssue #22643: Skip test_case_operation_overflow on computers with low memory.
Serhiy Storchaka [Sat, 7 Nov 2015 14:54:48 +0000 (16:54 +0200)]
Issue #22643: Skip test_case_operation_overflow on computers with low memory.

9 years agoIssue #892902: Added new tests for pickling recursive collections.
Serhiy Storchaka [Sat, 7 Nov 2015 09:16:10 +0000 (11:16 +0200)]
Issue #892902: Added new tests for pickling recursive collections.

9 years agoIssue #24802: Copy bytes-like objects to null-terminated buffers if necessary
Martin Panter [Sat, 7 Nov 2015 02:32:21 +0000 (02:32 +0000)]
Issue #24802: Copy bytes-like objects to null-terminated buffers if necessary

This avoids possible buffer overreads when int(), float(), compile(), exec()
and eval() are passed bytes-like objects. Similar code is removed from the
complex() constructor, where it was not reachable.

Patch by John Leitch, Serhiy Storchaka and Martin Panter.

9 years agoIssue #25555: Fix parser and AST: fill lineno and col_offset of "arg" node when
Victor Stinner [Fri, 6 Nov 2015 16:01:48 +0000 (17:01 +0100)]
Issue #25555: Fix parser and AST: fill lineno and col_offset of "arg" node when
compiling AST from Python objects.

9 years agoIssue #25582: Fixed 100 MB memory leak in test_ctypes.
Serhiy Storchaka [Mon, 9 Nov 2015 20:32:23 +0000 (22:32 +0200)]
Issue #25582: Fixed 100 MB memory leak in test_ctypes.

9 years agoImprove namedtuple doc string examples
Raymond Hettinger [Mon, 9 Nov 2015 16:24:53 +0000 (08:24 -0800)]
Improve namedtuple doc string examples

9 years agoFixes installer not allowing launcher to be selected.
Steve Dower [Mon, 9 Nov 2015 12:31:05 +0000 (07:31 -0500)]
Fixes installer not allowing launcher to be selected.

9 years agoIssue #24802: Update test case for ValueError in 3.5
Martin Panter [Sun, 8 Nov 2015 11:09:13 +0000 (11:09 +0000)]
Issue #24802: Update test case for ValueError in 3.5

9 years ago#25446: Fix regression in smtplib's AUTH LOGIN support.
R David Murray [Sun, 8 Nov 2015 06:03:52 +0000 (01:03 -0500)]
#25446: Fix regression in smtplib's AUTH LOGIN support.

The auth method tests simply weren't adequate because of the fact that
smtpd doesn't support authentication.  I borrowed some of Milan's
code for that from issue #21935 and added it to the smtplib tests.
Also discovered that the direct test for the 'auth' method wasn't actually
testing anything and fixed it.

The fix makes the new authobject mechanism work the way it is
documented...the problem was that wasn't checking for a 334 return code
if an initial-response was provided, which works fine for auth plain
and cram-md5, but not for auth login.