mirrors/cpython
1
0
Fork 0
mirror of https://github.com/python/cpython.git synced 2025-08-02 16:13:13 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 8
91524 commits 7 branches 621 tags 2.7 GiB
Commit graph

5289 commits

Author SHA1 Message Date
Victor Stinner
01ada3996b Issue #25267: The UTF-8 encoder is now up to 75 times as fast for error 
handlers: ``ignore``, ``replace``, ``surrogateescape``, ``surrogatepass``.
Patch co-written with Serhiy Storchaka.
 2015-10-01 21:54:51 +02:00
Victor Stinner
d69dd8bd5e (Merge 3.5) Issue #25182: Fix compilation on Windows 2015-09-30 15:03:50 +02:00
Victor Stinner
ae86da9b20 (Merge 3.4) Issue #25182: Fix compilation on Windows 2015-09-30 15:03:31 +02:00
Victor Stinner
89719e1daf Issue #25182: Fix compilation on Windows 
Restore also errno value before calling PyErr_SetFromErrno().
 2015-09-30 15:01:34 +02:00
Serhiy Storchaka
85c386dee4 Issue #25182: The stdprinter (used as sys.stderr before the io module is 
imported at startup) now uses the backslashreplace error handler.
 2015-09-30 15:51:01 +03:00
Serhiy Storchaka
008fc77e1e Issue #25182: The stdprinter (used as sys.stderr before the io module is 
imported at startup) now uses the backslashreplace error handler.
 2015-09-30 15:50:32 +03:00
Serhiy Storchaka
a59018c7ab Issue #25182: The stdprinter (used as sys.stderr before the io module is 
imported at startup) now uses the backslashreplace error handler.
 2015-09-30 15:46:53 +03:00
Victor Stinner
c3713e9706 Optimize ascii/latin1+surrogateescape encoders 
Issue #25227: Optimize ASCII and latin1 encoders with the ``surrogateescape``
error handler: the encoders are now up to 3 times as fast.

Initial patch written by Serhiy Storchaka.
 2015-09-29 12:32:13 +02:00
Victor Stinner
0030cd52da Issue #25227: Cleanup unicode_encode_ucs1() error handler 
* Change limit type from unsigned int to Py_UCS4, to use the same type than the
  "ch" variable (an Unicode character).
* Reuse ch variable for _Py_ERROR_XMLCHARREFREPLACE
* Add some newlines for readability
 2015-09-24 14:45:00 +02:00
Victor Stinner
54385b206d Issue #24870: revert unwanted change 
Sorry, I pushed the patch on the UTF-8 decoder by mistake :-(
 2015-09-22 10:46:52 +02:00
Victor Stinner
5ebae87628 Issue #25207, #14626: Fix my commit. 
It doesn't work to use #define XXX defined(YYY)" and then "#ifdef XXX"
to check YYY.
 2015-09-22 01:29:33 +02:00
Victor Stinner
6174474bea _PyUnicodeWriter_PrepareInternal(): make the assertion more strict 2015-09-22 01:01:17 +02:00
Victor Stinner
ca9381ea01 Issue #24870: Add _PyUnicodeWriter_PrepareKind() macro 
Add a macro which ensures that the writer has at least the requested kind.
 2015-09-22 00:58:32 +02:00
Victor Stinner
5014920cb7 Issue #24870: Reuse the new _Py_error_handler enum 
Factorize code with the new get_error_handler() function.

Add some empty lines for readability.
 2015-09-22 00:26:54 +02:00
Victor Stinner
f96418de05 Issue #24870: Optimize the ASCII decoder for error handlers: surrogateescape, 
ignore and replace. Initial patch written by Naoki Inada.

The decoder is now up to 60 times as fast for these error handlers.

Add also unit tests for the ASCII decoder.
 2015-09-21 23:06:27 +02:00
Victor Stinner
026977717e Merge 3.5 2015-09-19 13:39:16 +02:00
Victor Stinner
5783fd2c58 Issue #24999: In longobject.c, use two shifts instead of ">> 2*PyLong_SHIFT" to 
avoid undefined behaviour when LONG_MAX type is smaller than 60 bits.

This change should fix a warning with the ICC compiler.
 2015-09-19 13:39:03 +02:00
Victor Stinner
058258652a Merge 3.5 (pytime, odict) 2015-09-18 13:55:15 +02:00
Victor Stinner
4a0d1e7c36 odictobject.c: fix compiler warning 
PyObject_Length() returns a P_ssize_t, not an int. Use a Py_ssize_t to avoid
overflow.
 2015-09-18 13:44:11 +02:00
Serhiy Storchaka
56f6e76c68 Issue #15989: Fixed some scarcely probable integer overflows. 
It is very unlikely that they can occur in real code for now.
 2015-09-06 21:25:30 +03:00
Guido van Rossum
ba5f59089a Issue #24912: Prevent __class__ assignment to immutable built-in objects. (Merge 3.5 -> 3.6) 2015-09-05 15:20:57 -07:00
Guido van Rossum
37fdcbc4c3 Issue #24912: Prevent __class__ assignment to immutable built-in objects. (Merge 3.5.0 -> 3.5) 2015-09-05 15:20:08 -07:00
Guido van Rossum
7d293ee97d Issue #24912: Prevent __class__ assignment to immutable built-in objects. 2015-09-04 20:54:07 -07:00
Victor Stinner
fa9dfd4f82 Merge 3.5 (odict) 2015-09-03 17:50:30 +02:00
Victor Stinner
ca30b02abe Issue #24992: Fix error handling and a race condition (related to garbage 
collection) in collections.OrderedDict constructor.

Patch reviewed by Serhiy Storchaka.
 2015-09-03 17:50:04 +02:00
Victor Stinner
99bb14bf0c type_call() now detect bugs in type new and init 
* Call _Py_CheckFunctionResult() to check for bugs in type
  constructors (tp_new)
* Add assertions to ensure an exception was raised if tp_init failed
  or that no exception was raised if tp_init succeed

Refactor also the function to have less indentation.
 2015-09-03 12:16:49 +02:00
Eric V. Smith
ab2aa6dc91 Fixed an incorrect comment. 2015-08-26 14:10:32 -04:00
Stefan Krah
5f35725fb8 Merge #15944. 2015-08-08 13:38:59 +02:00
Stefan Krah
0c51595a78 Issue #15944: memoryview: Allow arbitrary formats when casting to bytes. 
Original patch by Martin Panter.
 2015-08-08 13:38:10 +02:00
Eric Snow
5060bc51ca Merge from 3.5 (issue #24667). 2015-08-07 17:47:35 -06:00
Eric Snow
8c7f9558eb Issue #24667: Resize odict in all cases that the underlying dict resizes. 2015-08-07 17:45:12 -06:00
Eric V. Smith
577d2069da Fix typo in comment. 2015-08-07 18:50:24 -04:00
Raymond Hettinger
4148195c45 Move the active entry multiplication to later in the hash calculation 2015-08-07 00:43:39 -07:00
Raymond Hettinger
b501a27ad8 Restore frozenset hash caching removed in cf707dd190a9 2015-08-06 22:15:22 -07:00
Zachary Ware
070bd62cfa Closes #21279: Merge with 3.5 2015-08-06 00:05:13 -05:00
Zachary Ware
d987a81d29 Issue #21279: Merge with 3.4 2015-08-06 00:04:23 -05:00
Zachary Ware
79b98df023 Issue #21279: Flesh out str.translate docs 
Initial patch by Kinga Farkas, Martin Panter, and John Posner.
 2015-08-05 23:54:15 -05:00
Raymond Hettinger
a286a51ae1 Fix comment typo 2015-08-01 11:07:11 -07:00
Raymond Hettinger
36c0500990 Tweak the comments 2015-08-01 10:57:42 -07:00
Raymond Hettinger
fbffdef47d Issue #24762: Speed-up frozenset_hash() and greatly beef-up the comments. 2015-08-01 09:53:00 -07:00
Raymond Hettinger
daffc916aa Issue #24681: Move the most likely test first in set_add_entry(). 2015-07-31 07:58:56 -07:00
Raymond Hettinger
70559b5c20 Issue #24681: Move the store of so->table to the code block where it is used. 2015-07-23 07:42:23 -04:00
Serhiy Storchaka
c8fe04484e Issue #23573: Restored optimization of bytes.rfind() and bytearray.rfind() 
for single-byte argument on Linux.
 2015-07-20 22:58:29 +03:00
Serhiy Storchaka
d92d4efe3d Issue #23573: Restored optimization of bytes.rfind() and bytearray.rfind() 
for single-byte argument on Linux.
 2015-07-20 22:58:02 +03:00
Raymond Hettinger
ff9e18a863 Issue #24583: Consolidate previous set object updates into a single function 
with a single entry point, named exit points at the bottom, more self-evident
refcount adjustments, and a comment describing why the pre-increment was
necessary at all.
 2015-07-20 07:34:05 -04:00
Raymond Hettinger
482c05cbb5 Issue #24583: Fix refcount leak. 2015-07-20 01:23:32 -04:00
Raymond Hettinger
061091a7c5 Issue #24583: Fix crash when set is mutated while being updated. 2015-07-15 23:54:02 -07:00
Raymond Hettinger
d702044bcd merge 2015-07-15 23:52:29 -07:00
Raymond Hettinger
a3626bc5bd Issue #24583: Fix crash when set is mutated while being updated. 2015-07-15 23:50:14 -07:00
Raymond Hettinger
5d2385ff6f Neaten-up a little bit. 2015-07-08 11:52:27 -07:00