homepage

Hi there,
I suggest to improve the description of Lock.acquire()
[ http://docs.python.org/library/threading.html#threading.Lock.acquire ]
in the following way:
>>>>> current version >>>>>
Lock.acquire([blocking])
 Acquire a lock, blocking or non-blocking.
 When invoked without arguments, block until the lock is unlocked,
 then set it to locked, and return true.
 When invoked with the *blocking* argument set to true, do the same
 thing as when called without arguments, and return true.
 When invoked with the *blocking* argument set to false, do not
 block. If a call without an argument would block, return false
 immediately; otherwise, do the same thing as when called without
 arguments, and return true.
<<<<< current version <<<<<
>>>>> suggested version >>>>>
Lock.acquire([blocking])
 Acquire a lock, blocking or non-blocking.
 When invoked with the *blocking* argument set to true 
 (the default), block until the lock is unlocked, 
 then set it to locked, and return true.
 When invoked with the *blocking* argument set to false, do not
 block. If a call without an argument would block, return false
 immediately; otherwise, set the lock to locked, and return true.
<<<<< suggested version <<<<<
The idea is to simplify the structure of the explanation: get rid of an unnecessary "goto" -- "do the same thing as" as well as the extra branching ("when invoked without arguments" ... "when invoked with *blocking* argument set to true") .
The suggested patch for the text version of the documentation ( http://docs.python.org/download.html -> http://docs.python.org/archives/python-2.7.3-docs-text.tar.bz2 ) is attached. 
PS. I did not dare to capitalize the boolean variables ("true" -> "True") to adhere to the general style of the document (obviously inherited from Java). For the same reason I didn't either change the argument signature from "[blocking]" to "[blocking=True]".

Thanks for the suggestion and the patch.
The 'true' isn't inherited from java. The actual value can be either an integer or True/False. (If it were a function written in Python it would be any value that evaluates as true, but because it is written in C it is actually restricted to an integer...this is a bit of a wart, actually).
In the Python3 docs it is indeed documented as 'blocking=True'. The Python2 docs use an older signature style that we didn't bother to fix up.

Well, as threading is a Python wrapper, this could easily be fixed. (I am not certain whether it *should* be fixed or not -- perhaps things are fine just as they are, at least with that particular detail. ) 
But this is good to know, thank you.

New changeset 6ab4128acccc by R David Murray in branch '3.2':
#14823: Simplify threading.Lock.acquire argument discussion.
http://hg.python.org/cpython/rev/6ab4128acccc
New changeset b5e95bb79ba3 by R David Murray in branch 'default':
#14823: Simplify threading.Lock.acquire argument discussion.
http://hg.python.org/cpython/rev/b5e95bb79ba3
New changeset 251463919f3c by R David Murray in branch '2.7':
#14823: Simplify threading.Lock.acquire argument discussion.
http://hg.python.org/cpython/rev/251463919f3c 

Instead I decided to go ahead and document the argument as True/False. That something else is accepted is a CPython implementation detail that shouldn't be depended on.
By the way, I couldn't actually use your patch file, since it wasn't against the threading.rst file from the repository.

I suspect the patch was made against the source files served by Sphinx with a txt extension.

My bad. That's indeed what I did. Won't repeat the mistake, sorry.