* bpo-351428: Updates documentation to reflect AsyncMock call_count after await.
* Adds skip and fixes warning.
* Removes extra >>>.
* Adds ... in front of await mock().
The link we have points to the version from Unicode 6.0.0, dated 2010.
There have been numerous updates to it since then:
https://www.unicode.org/reports/tr44/#Modifications
Change the link to one that points to the current version. Also, use HTTPS.
* Minor changes.
* Update Doc/faq/library.rst
Co-Authored-By: Kyle Stanley <aeros167@gmail.com>
* Apply suggestions from aeros167.
* Update Doc/faq/library.rst
Co-Authored-By: Kyle Stanley <aeros167@gmail.com>
* Apply suggestions from aeros167 + re-add a "a" that was accidentally deleted.
* Update documentation for plistlib
- Update "Mac OS X" to "Apple" since plists are used more widely than just macOS
- Re-add the UID class documentation (oops, removed in GH-15615)
* Rename PyThreadState_DeleteCurrent()
to _PyThreadState_DeleteCurrent()
* Move it to the internal C API
Co-Authored-By: Carol Willing <carolcode@willingconsulting.com>
The purpose of the `unicodedata.is_normalized` function is to answer
the question `str == unicodedata.normalized(form, str)` more
efficiently than writing just that, by using the "quick check"
optimization described in the Unicode standard in UAX #15.
However, it turns out the code doesn't implement the full algorithm
from the standard, and as a result we often miss the optimization and
end up having to compute the whole normalized string after all.
Implement the standard's algorithm. This greatly speeds up
`unicodedata.is_normalized` in many cases where our partial variant
of quick-check had been returning MAYBE and the standard algorithm
returns NO.
At a quick test on my desktop, the existing code takes about 4.4 ms/MB
(so 4.4 ns per byte) when the partial quick-check returns MAYBE and it
has to do the slow normalize-and-compare:
$ build.base/python -m timeit -s 'import unicodedata; s = "\uf900"*500000' \
-- 'unicodedata.is_normalized("NFD", s)'
50 loops, best of 5: 4.39 msec per loop
With this patch, it gets the answer instantly (58 ns) on the same 1 MB
string:
$ build.dev/python -m timeit -s 'import unicodedata; s = "\uf900"*500000' \
-- 'unicodedata.is_normalized("NFD", s)'
5000000 loops, best of 5: 58.2 nsec per loop
This restores a small optimization that the original version of this
code had for the `unicodedata.normalize` use case.
With this, that case is actually faster than in master!
$ build.base/python -m timeit -s 'import unicodedata; s = "\u0338"*500000' \
-- 'unicodedata.normalize("NFD", s)'
500 loops, best of 5: 561 usec per loop
$ build.dev/python -m timeit -s 'import unicodedata; s = "\u0338"*500000' \
-- 'unicodedata.normalize("NFD", s)'
500 loops, best of 5: 512 usec per loop
Adds a link to `dateutil.parser.isoparse` in the documentation.
It would be nice to set up intersphinx for things like this, but I think we can leave that for a separate PR.
CC: @pitrou
[bpo-37979](https://bugs.python.org/issue37979)
https://bugs.python.org/issue37979
Automerge-Triggered-By: @pitrou
