2e9bd0f241
prc: fix compilation issue on MSVC
2019-05-12 20:58:13 +02:00
a7c743fd5e
Allow seek of IDecryptStream to begin (for looping encrypted audio)
2019-05-12 15:51:35 +02:00
60922fabc1
Give istream/ostream a friendlier file-like interface for Python
2019-05-12 15:46:14 +02:00
0ac36185a9
Merge branch 'release/1.10.x'
2019-05-02 21:34:25 +02:00
204cbe4464
interrogate: support unicode characters in Python 3 for 'char' arg
...
Fixes #626
2019-05-01 17:05:12 +02:00
5d6b4f4a77
dtool: add StreamReader/StreamWriter move ctor and assignment ops
2019-04-29 18:17:54 +02:00
daa57733cb
Merge branch 'release/1.10.x'
2019-04-14 23:11:03 +02:00
3ca3dfd13a
interrogate: fix in-place or (|=) operators (see #588 )
2019-04-14 17:10:52 +02:00
a8ceac919a
Raise version number to 1.10.3 on release/1.10.x branch
2019-03-12 18:08:19 +01:00
8d1bdf0118
Merge branch 'release/1.10.x'
2019-03-10 20:49:52 +01:00
a634e729c8
interrogate: fix uninitialized var, fixes sporadic nb_true_divide
...
This was a regression introduced by fa6c066b2ffixed #529 )
2019-03-08 11:53:45 +01:00
148134284d
Merge branch 'release/1.10.x'
2019-03-07 19:04:27 +01:00
12eeff0dec
py_panda: fix memory leak accessing map/seq properties
2019-03-07 17:33:21 +01:00
ba256a8b07
interrogate: fix regression in 2c91fdda1b
2019-03-04 22:51:20 +01:00
0f5da1c155
dtoolbase: fix warning when compiling on Windows
2019-03-04 16:58:18 +01:00
cbb3c182a5
Merge branch 'release/1.10.x'
2019-03-04 16:57:52 +01:00
8e4add0326
interrogate: fix refcount trouble assigning PyObject* properties
...
In particular, this fixes accessing PythonTask.__dict__
2019-03-04 16:46:28 +01:00
fa6c066b2f
interrogate: enable true division in Python 2 as well
...
This allows using "from __future__ import division" and get the expected result for divisions in Python 2.
Fixes #529
2019-03-03 10:29:39 +01:00
2c91fdda1b
interrogate: support in-place ops on properties (eg. light.color *= 2)
2019-02-19 13:46:38 +01:00
c37a33baef
interrogate: don't export bindings for global operators
...
linmath exports a global __mul__, which isn't really useful.
2019-02-19 13:45:37 +01:00
5e9c478b1a
Raise version number to 1.10.2 on release/1.10.x branch
...
[skip ci]
2019-02-18 12:50:28 +01:00
80bb5c09fe
Merge branch 'release/1.10.x'
2019-02-12 01:13:56 +01:00
5310945f4c
py_panda: fix use of macro that was removed in Python 3.8
2019-02-11 23:36:30 +01:00
3df3b27a4e
prc: prefer global-scope static MutexImpl over local one
...
MutexImpl guarantees nowadays to be initialized at constant init time, so this is safer on Windows, where we can't rely on magic statics.
2019-02-11 16:23:16 +01:00
b466e77fe3
dtoolbase: make TypeRegistry lock statically initialized
...
This is possible now that all MutexImpl have a constexpr constructor.
2019-02-11 16:23:16 +01:00
fb5191e812
pipeline: unify ConditionVarFull and ConditionVar
...
See #550 . Now that we use the Vista API for condition variables, the only place where the distinction is still relevant is Windows XP, and it's just not worth it for that one corner case.
2019-02-11 16:23:10 +01:00
ab569916b1
Disable WinXP mutex emulation when _WIN32_WINNT >= 0x0600
2019-02-11 14:52:21 +01:00
89392f0e2d
dtoolbase: new mutex and condition variable impl on Windows
...
This uses SRWLock on Vista and above, and uses a hand-rolled implementation on Windows XP that uses Events (and a spinlock, if this is a multi-core system).
Since SRWLocks aren't recursive, ReMutexWin32Impl has been added to implement recursive mutices, using old-fashioned critical sections.
MutexImpl now has a constexpr constructor on all implementations.
2019-02-07 00:20:59 +01:00
7e0952e367
parser-inc: add synchapi.h and winnt.h headers
2019-02-07 00:20:59 +01:00
36dd8889e7
Merge branch 'release/1.10.x'
2019-02-06 22:52:19 +01:00
666568a513
prc: fix recursive use of non-recursive mutex
2019-02-06 22:10:39 +01:00
86c266fd88
Merge branch 'release/1.10.x'
2019-01-31 00:58:29 +01:00
3a6f7fcde1
dtoolbase: docstring clarification for record_python_type
...
[skip ci]
2019-01-31 00:11:21 +01:00
1d9334f578
parser-inc: add atomic_int, etc. typedefs
2019-01-30 23:59:34 +01:00
ffa0c74402
Merge branch 'release/1.10.x'
2019-01-26 20:18:02 +01:00
376cef51c4
interrogate: fix Python 3 crash with optional std::wstring args
...
This would cause PyMem_Free to be called on an uninitialized pointer if a default argument was not specified.
Fixes #542
2019-01-26 19:27:50 +01:00
403e56817a
dtoolbase: fix x86 compilation error on Android
2019-01-26 19:22:44 +01:00
fc90ca7fdf
Merge branch 'release/1.10.x'
2019-01-20 17:02:15 +01:00
ba4036e290
prc: clear error flag if loading deploy-ng blobinfo fails
2019-01-19 23:12:46 +01:00
6eca44464e
load_dso: prevent load_dso_error() from returning old error
...
This happens if load_dso fails due to being unable to find the file, but not setting the dlerror flag and therefore causing a misleading error message to be shown.
2019-01-19 23:10:46 +01:00
ba993aea0d
Merge branch 'release/1.10.x'
2019-01-10 23:32:29 +01:00
84a1315104
More fixes to work around broken libc++ eof() in Mac OS X 10.7
2019-01-10 21:53:02 +01:00
7c6e677a23
dtoolbase: add std::initializer_list constructors for pvector, epvector, pdeque, plist, pset, pmultiset
...
Closes #520
2019-01-10 01:42:44 +01:00
441e1e3d66
Bump version number on master to 1.11.0
2019-01-07 18:40:02 +01:00
7132d09e7e
Bump version to 1.10.1
...
[skip ci]
2019-01-06 09:34:55 +01:00
0c890f059d
interrogate: Export Python entry points with EXPORT_CLASS
...
Closes #505
2019-01-03 12:39:22 +01:00
ce4985a333
interrogatedb: Fix missing 'static' keyword
2019-01-02 10:18:59 -07:00
ac8be1a2e6
py_panda: fix int-to-enum conversion in Python 2
...
Fixes #498
2019-01-01 16:32:04 +01:00
0a43008313
py_panda: avoid duplicate symbol issues with LINK_ALL_STATIC
...
Fixes #478
2018-12-30 16:42:50 +01:00
818fdbd232
interrogate: Tidy up hash_string function
...
The main motivation behind this change is to get rid of a
signed integer overflow that sometimes happens in the prime
multiplication step, which is (per the C++ spec) undefined
behavior. However, it's probably for the best to use only
unsigned int when the function is clearly trying to avoid
negative values.
Not that I suspect it matters much, but I have also heavily
tested that the behavior of the function is unchanged (at
least on PC hardware - signed integer overflow doesn't
behave portably) although it may now be slightly faster due
to the fact that I have removed the floating-point math.
2018-12-30 04:21:34 -07:00
rdb
Brian Lach
Sam Edwards