Flushing sys.stdout and sys.stderr in Py_FatalError() can call again
Py_FatalError(). Add a reentrant flag to detect this case and just abort at the
second call.
It should help to see exceptions when stderr if buffered: PyErr_Display() calls
sys.stderr.write(), it doesn't write into stderr file descriptor directly.
* Display the current Python stack if an exception was raised but the exception
has no traceback
* Disable faulthandler if an exception was raised (before it was only disabled
if no exception was raised)
* To display the current Python stack, call PyGILState_GetThisThreadState()
which works even if the GIL was released
Issue #23654: Turn off ICC's tail call optimization for the stack_overflow
generator. ICC turns the recursive tail call into a loop.
Patch written by Matt Frank.
Some http servers will reject PUT, POST, and PATCH requests if they
do not have a Content-Length header.
Patch by James Rutherford, with additional cleaning up of the
'request' documentation by me.
Being able to read non-python text files is not a purpose of linecache, but it
does work and people use it. This changeset adjusts the language to make it
clear that Python files are not treated uniquely, but does not go so far as to
say reading non-python files is explicitly supported.
some circunstances while NamedTemporaryFile object was living. This causes
failing test_csv. Changed the implementation of NamedTemporaryFile.__iter__
to make tests passed.
add private method to enum to support replacing global constants with Enum members:
- search for candidate constants via supplied filter
- create new enum class and members
- insert enum class and replace constants with members via supplied module name
- replace __reduce_ex__ with function that returns member name, so previous Python versions can unpickle
modify IntEnum classes to use new method
Use a Python source file (linecache.__file__) instead of /etc/passwd.
Modify also linecache docstrings to clarify the linecache is written to cache
Python source files, not any text files.
The current documentation only mentions heap[0] as the smallest element in the
beginning, and not in any of the methods' docs. There's no method to access the
minimal element without popping it, and the documentation of nsmallest is
confusing because it may suggest that min() is the way to go for n==1.
Serhiy Storchaka
Victor Stinner
Ezio Melotti
Raymond Hettinger
R David Murray
Benjamin Peterson
Ned Deily
Ethan Furman
Antoine Pitrou
Vinay Sajip
Berker Peksag
Eli Bendersky
Steve Dower
Brett Cannon