* Add also min_char attribute to _PyUnicodeWriter structure (currently unused)
* _PyUnicodeWriter_Init() has no more argument (except the writer itself):
min_length and overallocate must be set explicitly
* In error handlers, only enable overallocation if the replacement string
is longer than 1 character
* CJK decoders don't use overallocation anymore
* Set min_length, instead of preallocating memory using
_PyUnicodeWriter_Prepare(), in many decoders
* _PyUnicode_DecodeUnicodeInternal() checks for integer overflow
* Check also that right is a Unicode object
* call directly resize_compact() instead of unicode_resize() for a more
explicit error handling, and to avoid testing some properties twice
(ex: unicode_modifiable())
strings are not convincing. For UCS2 (16-bit wchar_t type), use a dummy loop
instead of wmemcmp(). The dummy loop is as fast, or a little bit faster.
wchar_t is only 16-bit long on Windows. wmemcmp() is still used for 32-bit
wchar_t.
Inline the BLOOM_MEMBER() to only call PyUnicode_READ() only once (per loop
iteration). Store also the length of the seperator in a variable to avoid calls
to PyUnicode_GET_LENGTH().
"PyUnicode_READ_CHAR() is less efficient than PyUnicode_READ() because it calls
PyUnicode_KIND() and might call it twice." according to its documentation.
Write a function to enable more optimizations:
* If the substring is the whole string and overallocation is disabled, just
keep a reference to the string, don't copy characters
* Avoid a call to the expensive _PyUnicode_FindMaxChar() function when
possible
"honor direction and do a forward or backwards search": the runtime speed may
be different, but I consider that it doesn't really matter in practice. The
direction was never honored before: Python 2.7 uses memcmp() for the str type
for example.
ASCII/surrogateescape codec is now used, instead of the locale encoding, to
decode the command line arguments. This change fixes inconsistencies with
os.fsencode() and os.fsdecode() because these operating systems announces an
ASCII locale encoding, whereas the ISO-8859-1 encoding is used in practice.
ASCII/surrogateescape codec is now used, instead of the locale encoding, to
decode the command line arguments. This change fixes inconsistencies with
os.fsencode() and os.fsdecode() because these operating systems announces an
ASCII locale encoding, whereas the ISO-8859-1 encoding is used in practice.
computation as the overflow behavior of signed integers is undefined.
NOTE: This change is smaller compared to 3.2 as much of this cleanup had
already been done. I added the comment that my change in 3.2 added so that the
code would match up. Otherwise this just adds or synchronizes appropriate UL
designations on some constants to be pedantic.
In practice we require compiling everything with -fwrapv which forces overflow
to be defined as twos compliment but this keeps the code cleaner for checkers
or in the case where someone has compiled it without -fwrapv or their
compiler's equivalent. We could work to get rid of the -fwrapv requirement
in 3.4 but that requires more planning.
Found by Clang trunk's Undefined Behavior Sanitizer (UBSan).
Cleanup only - no functionality or hash values change.
computation as the overflow behavior of signed integers is undefined.
NOTE: This change is smaller compared to 3.2 as much of this cleanup had
already been done. I added the comment that my change in 3.2 added so that the
code would match up. Otherwise this just adds or synchronizes appropriate UL
designations on some constants to be pedantic.
In practice we require compiling everything with -fwrapv which forces overflow
to be defined as twos compliment but this keeps the code cleaner for checkers
or in the case where someone has compiled it without -fwrapv or their
compiler's equivalent.
Found by Clang trunk's Undefined Behavior Sanitizer (UBSan).
Cleanup only - no functionality or hash values change.
computation as the overflow behavior of signed integers is undefined.
In practice we require compiling everything with -fwrapv which forces overflow
to be defined as twos compliment but this keeps the code cleaner for checkers
or in the case where someone has compiled it without -fwrapv or their
compiler's equivalent.
Found by Clang trunk's Undefined Behavior Sanitizer (UBSan).
Cleanup only - no functionality or hash values change.
* Remove micro-optization:
(errors == "surrogateescape" || strcmp(errors, "surrogateescape") == 0).
Only use strcmp()
* Initialize 'arg' members in unicode_format_arg() to help the compiler to
diagnose real bugs and also make the code simpler to read
ASCII/surrogateescape codec is now used, instead of the locale encoding, to
decode the command line arguments. This change fixes inconsistencies with
os.fsencode() and os.fsdecode() because these operating systems announces an
ASCII locale encoding, whereas the ISO-8859-1 encoding is used in practice.
encoded/decoded to/from UTF-8/surrogateescape, instead of the locale encoding
(which may be ASCII if no locale environment variable is set), to avoid
inconsistencies with os.fsencode() and os.fsdecode() functions which are
already using UTF-8/surrogateescape.
encoded/decoded to/from UTF-8/surrogateescape, instead of the locale encoding
(which may be ASCII if no locale environment variable is set), to avoid
inconsistencies with os.fsencode() and os.fsdecode() functions which are
already using UTF-8/surrogateescape.
* Remove unicode_widen(): replaced with _PyUnicodeWriter_Prepare()
* Remove unicode_putchar(): replaced with
PyUnicodeWriter_Prepare() + PyUnicode_WRITER()
* When handling an decoding error, only overallocate the buffer by +25%
instead of +100%
This commit rewrites the docstring for int() to incorporate the documentation
changes made in issue #16036. It also switches the docstrings for int(),
str(), range(), and slice() to use multi-line signatures.
* Simplify the code: replace 4 steps with one unique step using the
_PyUnicodeWriter API. PyUnicode_Format() has the same design. It avoids to
store intermediate results which require to allocate an array of pointers on
the heap.
* Use the _PyUnicodeWriter API for speed (and its convinient API):
overallocate the buffer to reduce the number of "realloc()"
* Implement "width" and "precision" in Python, don't rely on sprintf(). It
avoids to need of a temporary buffer allocated on the heap: only use a small
buffer allocated in the stack.
* Add _PyUnicodeWriter_WriteCstr() function
* Split PyUnicode_FromFormatV() into two functions: add
unicode_fromformat_arg().
* Inline parse_format_flags(): the format of an argument is now only parsed
once, it's no more needed to have a subfunction.
* Optimize PyUnicode_FromFormatV() for characters between two "%" arguments:
search the next "%" and copy the substring in one chunk, instead of copying
character per character.
- Use _PyLong_FormatWriter() instead of formatlong() when possible, to avoid
a temporary buffer
- Enable the fast path when width is smaller or equals to the length,
and when the precision is bigger or equals to the length
- Add unit tests!
- formatlong() uses PyUnicode_Resize() instead of _PyUnicode_FromASCII()
to resize the output string
* Formatting string, int, float and complex use the _PyUnicodeWriter API. It
avoids a temporary buffer in most cases.
* Add _PyUnicodeWriter_WriteStr() to restore the PyAccu optimization: just
keep a reference to the string if the output is only composed of one string
* Disable overallocation when formatting the last argument of str%args and
str.format(args)
* Overallocation allocates at least 100 characters: add min_length attribute
to the _PyUnicodeWriter structure
* Add new private functions: _PyUnicode_FastCopyCharacters(),
_PyUnicode_FastFill() and _PyUnicode_FromASCII()
The speed up is around 20% in average.
Victor Stinner
Serhiy Storchaka
Raymond Hettinger
Benjamin Peterson
Andrew Svetlov
Gregory P. Smith
Antoine Pitrou
Ezio Melotti
Mark Dickinson
Christian Heimes
Chris Jerdonek
Stefan Krah
Nick Coghlan
Kristján Valur Jónsson
Kristjan Valur Jonsson