Thu, 28 Apr 2016 13:05:57 +0300fix issue with sometimes wrong return code of ProcessLogger
Andrey Skvortsov <andrej.skvortzov@gmail.com> [Thu, 28 Apr 2016 13:05:57 +0300] rev 1507
fix issue with sometimes wrong return code of ProcessLogger


As a result of wrong return code Beremiz gives folowing traceback:
Traceback (most recent call last):
File "./Beremiz.py", line 850, in OnMenu
getattr(self.CTR, method)()
File "/home/developer/WorkData/PLC/beremiz/beremiz/ProjectController.py", line 925, in _Build
IECGenRes = self._Generate_SoftPLC()
File "/home/developer/WorkData/PLC/beremiz/beremiz/ProjectController.py", line 568, in _Generate_SoftPLC
return self._Compile_ST_to_SoftPLC()
File "/home/developer/WorkData/PLC/beremiz/beremiz/ProjectController.py", line 661, in _Compile_ST_to_SoftPLC
C_files.remove("POUS.c")
ValueError: list.remove(x): x not in list

The problem is that both threads (for reading stdout and stderr) call self.Proc.poll(),
that updates internal returncode field. This call is done without any locking and the first thread gets correct result,
but other gets 0 as retval. If 0 gets thread, that afterwards calls callback finish, then wrong return code is returned
to the parent. Now only the thread with a callback polls for the return code, other thread just checked local value.

Additionally function spin() waits now until all threads finish reading their pipes, so the results are always correct.

Thu, 28 Apr 2016 12:58:58 +0300Fix error about missing attribute 'timeout' that happens sometimes during compilation
Andrey Skvortsov <andrej.skvortzov@gmail.com> [Thu, 28 Apr 2016 12:58:58 +0300] rev 1506
Fix error about missing attribute 'timeout' that happens sometimes during compilation

The fix [1476:49f1763a5613] of the problem with following trace was wrong.
Traceback (most recent call last):
File "./Beremiz.py", line 1229, in run_with_except_hook
run_old(*args, **kw)
File
"/home/developer/WorkData/PLC/beremiz/beremiz/util/ProcessLogger.py",
line 68, in run
self.endcallback(self.Proc.pid, err)
File
"/home/developer/WorkData/PLC/beremiz/beremiz/util/ProcessLogger.py",
line 169, in finish
if self.timeout: self.timeout.cancel()
AttributeError: ProcessLogger instance has no attribute 'timeout'
The problem was that compilation process was finished before the timeout attribute is set.
Now timeout is set before launcing of compilation process.

Wed, 27 Apr 2016 18:42:30 +0300fix copying elements with non-latin content like comment element, or
Andrey Skvortsov <andrej.skvortzov@gmail.com> [Wed, 27 Apr 2016 18:42:30 +0300] rev 1505
fix copying elements with non-latin content like comment element, or
variables with description in non-latin character set

before any non-latin text was converted after copying to strings like '&#1069;&#1090;&#1086;...'

Mon, 25 Apr 2016 17:23:04 +0300fix issue that if functional block is dragged from project panel into
Andrey Skvortsov <andrej.skvortzov@gmail.com> [Mon, 25 Apr 2016 17:23:04 +0300] rev 1504
fix issue that if functional block is dragged from project panel into
ST text editor and in dialog 'Please enter a block name' button
'cancel' is pressed, despite of cancel text about drag'n'dropped type
is pasted in ST program.

Fri, 22 Apr 2016 21:26:31 +0300fix regression introduced by changeset: 1431:28e9d479aa65 [Use '-p' command line option when running matiec (iec2c)]
Andrey Skvortsov <andrej.skvortzov@gmail.com> [Fri, 22 Apr 2016 21:26:31 +0300] rev 1503
fix regression introduced by changeset: 1431:28e9d479aa65 [Use '-p' command line option when running matiec (iec2c)]

Variable matiec_flags in c_ext.py and PythonFileCTNMixin.py is not
used for flags for matiec (IEC to C compiler) actually, but it's used
to compile C-code generated for extensions using normal C-compiler,
like gcc.

Fri, 22 Apr 2016 19:49:10 +0300fix display of ARRAY fields inside of structure data types.
Andrey Skvortsov <andrej.skvortzov@gmail.com> [Fri, 22 Apr 2016 19:49:10 +0300] rev 1502
fix display of ARRAY fields inside of structure data types.

Previously ARRAY fields in structures were displayed, for example,
like 'ARRAY [1..2] OF ARRAY' instead of 'ARRAY [1..2] OF INT'. Because
of this error the array type can be really edited.

Fri, 22 Apr 2016 17:02:18 +0300fix Traceback if search icon on library panel is clicked, when no
Andrey Skvortsov <andrej.skvortzov@gmail.com> [Fri, 22 Apr 2016 17:02:18 +0300] rev 1501
fix Traceback if search icon on library panel is clicked, when no
category/item is selected in the library.

The problem is here, that if nothing is selected method GetSelection()
returns root item, that doesn't have any python data associated with it.
This issue seems to be wx3.0 specified, because it doesn't exist
with wx2.8 on Windows.

Fri, 22 Apr 2016 16:59:47 +0300update Russian about Beremiz dialog window
Andrey Skvortsov <andrej.skvortzov@gmail.com> [Fri, 22 Apr 2016 16:59:47 +0300] rev 1500
update Russian about Beremiz dialog window

Fri, 22 Apr 2016 16:24:06 +0300fix issue, that was impossible to change time interval for cyclic
Andrey Skvortsov <andrej.skvortzov@gmail.com> [Fri, 22 Apr 2016 16:24:06 +0300] rev 1499
fix issue, that was impossible to change time interval for cyclic
tasks, when wx 3.0 is used

Thu, 21 Apr 2016 20:42:37 +0300fix deprecation warnings about accessing properties m_x, m_y of wxKeyEvent
Andrey Skvortsov <andrej.skvortzov@gmail.com> [Thu, 21 Apr 2016 20:42:37 +0300] rev 1498
fix deprecation warnings about accessing properties m_x, m_y of wxKeyEvent

/home/beremiz/beremiz/controls/PouInstanceVariablesPanel.py:354: wxPyDeprecationWarning: Accessing deprecated property.
event.m_x = size.width / 2

Thu, 21 Apr 2016 19:14:28 +0300fix set_color_cycle from matplotlib deprecation warning
Andrey Skvortsov <andrej.skvortzov@gmail.com> [Thu, 21 Apr 2016 19:14:28 +0300] rev 1497
fix set_color_cycle from matplotlib deprecation warning

/usr/lib/python2.7/dist-packages/matplotlib/cbook.py:137: MatplotlibDeprecationWarning: The set_color_cycle attribute was deprecated in version 1.5. Use set_prop_cycle instead.
warnings.warn(message, mplDeprecation, stacklevel=1)

Thu, 21 Apr 2016 17:51:29 +0300update Russian translation
Andrey Skvortsov <andrej.skvortzov@gmail.com> [Thu, 21 Apr 2016 17:51:29 +0300] rev 1496
update Russian translation

Thu, 21 Apr 2016 17:49:39 +0300add i18n for some strings in DiscoveryDialog, DurationEditDialog, ProjectDialog and CodeFileEditor
Andrey Skvortsov <andrej.skvortzov@gmail.com> [Thu, 21 Apr 2016 17:49:39 +0300] rev 1495
add i18n for some strings in DiscoveryDialog, DurationEditDialog, ProjectDialog and CodeFileEditor

Thu, 21 Apr 2016 17:47:35 +0300fix wrong Row number in DiscoveryDialog.py that caused following traceback:
Andrey Skvortsov <andrej.skvortzov@gmail.com> [Thu, 21 Apr 2016 17:47:35 +0300] rev 1494
fix wrong Row number in DiscoveryDialog.py that caused following traceback:

Traceback (most recent call last):
File "/home/beremiz/beremiz/ProjectController.py", line 1536, in _Connect
dialog = DiscoveryDialog(self.AppFrame)
File "/home/beremiz/beremiz/dialogs/DiscoveryDialog.py", line 124, in __init__
self._init_ctrls(parent)
File "/home/beremiz/beremiz/dialogs/DiscoveryDialog.py", line 121, in _init_ctrls
self._init_sizers()
File "/home/beremiz/beremiz/dialogs/DiscoveryDialog.py", line 73, in _init_sizers
self._init_coll_ButtonGridSizer_Growables(self.ButtonGridSizer)
File "/home/beremiz/beremiz/dialogs/DiscoveryDialog.py", line 64, in _init_coll_ButtonGridSizer_Growables
parent.AddGrowableRow(1)
File "/usr/lib/python2.7/dist-packages/wx-3.0-gtk2/wx/_core.py", line 15343, in AddGrowableRow
return _core_.FlexGridSizer_AddGrowableRow(*args, **kwargs)
PyAssertionError: C++ assertion "!m_rows || idx < (size_t)m_rows" failed at ../src/common/sizer.cpp(1967) in AddGrowableRow(): invalid row index

Thu, 21 Apr 2016 17:44:17 +0300increase height of some dialog windows so they have enough space for buttons
Andrey Skvortsov <andrej.skvortzov@gmail.com> [Thu, 21 Apr 2016 17:44:17 +0300] rev 1493
increase height of some dialog windows so they have enough space for buttons