I have been developing a PyQt5 software for a while now. I have managed to package my Python 3 PyQt5 software with py2exe fine in past with everything working perfectly.
However, now I have encountered an issue where the packaged exe-program will crash when user exits. More specifically I get APPCRASH with following details
Problem signature:
Problem Event Name: APPCRASH
Application Name: Sotilasmatrikkelit.exe
Application Version: 0.0.0.0
Application Timestamp: 54467a51
Fault Module Name: PyQt5.QtCore.pyd
Fault Module Version: 0.0.0.0
Fault Module Timestamp: 549be77e
Exception Code: c0000005
Exception Offset: 0010c185
OS Version: 6.1.7601.2.1.0.256.4
Locale ID: 1035
Additional Information 1: 0a9e
Additional Information 2: 0a9e372d3b4ad19135b953a78882e789
Additional Information 3: 0a9e
Additional Information 4: 0a9e372d3b4ad19135b953a78882e789
I never get this crash when running the software from Pycharm during the development. Any idea of how to debug this or what could be the cause in general?
I suspect this might have something to do with memory management (that PyQt doesn't delete all the resources properly on exit and therefore segfaults) but does anyone have any good suggestions to figure out the exact problem without better debug information? Should I try to do some kind of cleanup on exit? Atm I start the software like this:
def start():
import sys
app = QApplication(sys.argv)
fixingtool = Mainwindow(app)
fixingtool.show()
sys.exit(app.exec_())
Additional investigation seems to suggest that I get Access Violation which is caused by C++ null-pointer. Sounds scary since I don't know too much of PyQt debugging. Anyway, I found an area on my businesslogic code which if removed will remove the problem. However, this code has nothing to do with PyQt and is just regular Python code and shouldn't differ in any way. Strangest part is that if I remove certain functions from the code, problem disappears even though functions are not called during runtime meaning that just importing the file with those functions cause the problem. Below is a code sample:
import re
from books.karelians.extraction.extractors.baseExtractor import BaseExtractor
from books.karelians.extraction.extractionExceptions import *
from books.karelians.extraction.extractors.dateExtractor import DateExtractor
from shared import textUtils
from books.karelians.extractionkeys import KEYS
from interface.valuewrapper import ValueWrapper
from shared import regexUtils
from books.karelians.extraction.extractors.professionextractor import ProfessionExtractor
class SpouseExtractor(BaseExtractor):
def extract(self, text, entry):
super(SpouseExtractor, self).extract(text)
self.entry = entry
self.PATTERN = r"Puol\.?(?P<spousedata>[A-ZÄ-Öa-zä-ö\s\.,\d-]*)(?=(Lapset|poika|tytär|asuinp))"
self.NAMEPATTERN = r"(?P<name>^[\w\s\.-]*)"
self.OPTIONS = (re.UNICODE | re.IGNORECASE) #TODO: TRY IGNORE CASE?
self.REQUIRES_MATCH_POSITION = False
self.SUBSTRING_WIDTH = 100
self.hasSpouse = False
self.spouseName = ""
self.profession = {KEYS["profession"] : ValueWrapper("")}
self.initVars(text)
self._findSpouse(text)
return self._constructReturnDict()
def initVars(self,text):
pass
def _findSpouse(self, text):
try:
self.foundSpouse = regexUtils.safeSearch(self.PATTERN, text, self.OPTIONS)
self.hasSpouse = True
self._findSpouseName(self.foundSpouse.group("spousedata"))
self._setFinalMatchPosition()
except regexUtils.RegexNoneMatchException:
pass
def _findSpouseName(self, text):
try:
name = regexUtils.safeSearch(self.NAMEPATTERN, text, self.OPTIONS)
self.spouseName = name.group("name").strip()
self._findProfession(text[name.end():])
except regexUtils.RegexNoneMatchException:
self.errorLogger.logError(SpouseNameException.eType, self.currentChild)
def _findProfession(self, text):
professionExt = ProfessionExtractor(self.entry, self.errorLogger, self.xmlDocument)
professionExt.setDependencyMatchPositionToZero()
self.profession = professionExt.extract(text, self.entry)
def _setFinalMatchPosition(self):
#Dirty fix for inaccuracy in positions which would screw the Location extraction
self.matchFinalPosition = self.foundSpouse.end() + self.matchStartPosition - 4
def _constructReturnDict(self):
print(self.profession)
return {KEYS["spouse"]: ValueWrapper({ KEYS["hasSpouse"]: ValueWrapper(self.hasSpouse),KEYS["spouseName"]: ValueWrapper(self.spouseName), KEYS["spouseProfession"]: ValueWrapper(self.profession[KEYS["profession"]].value) })}
Here if I remove or comment away all functions "initVars()" program exits properly. What gives?
This isn't really a solution to a problem itself but I'm leaving this here in case someone encounters a similar issue.
Today I decided to try cx_freeze instead of py2exe figuring that maybe the problem is with py2exe since the whole issue is not happening when running the application with normal python interpreter.
Turns out I was right and problem seemed to magically disappear after I packaged the app using cx_freeze instead of py2exe. I didn't do any changes to code. Someone more knowledgeable than me about how py2exe and cx_freeze work might be able to explain the difference. My wild guess is that for some reason the exiting from the Python interpreter is not handled perfectly in py2exe case somehow messing the end garbage cleanup. I have no idea if this is py2exe's or my fault by not configuring py2exe properly.
In any case I'm happy it works now since yesterday was a really frustrating day.
Related
I'm trying to add a custom IMessageFilter in a Winforms app using Python.net, but I'm getting a segfault.
Here's a minimal sample application:
import clr
clr.AddReference("System.Windows.Forms")
import System.Windows.Forms as WinForms
class MessageFilter(WinForms.IMessageFilter):
__namespace__ = 'System.Windows.Forms'
def PreFilterMessage(self, message):
print('filter', message)
return False
class HelloApp(WinForms.Form):
def __init__(self):
self.textbox = WinForms.TextBox()
self.textbox.Text = "Hello World"
self.Controls.Add(self.textbox)
def main():
form = HelloApp()
app = WinForms.Application
f = MessageFilter()
app.AddMessageFilter(f)
app.Run(form)
if __name__ == '__main__':
main()
If you run this code as presented, the application window displays, but you get a segfault immediately (I presume this is when the first message is passed to the filter). The segfault is entirely opaque. There's no stack trace or other helpful details - it's just the OS-level segfault handler.
If you comment out line 25 (app.AddMessageFilter(f), installing the actual filter), the code works fine.
If you modify MessageFilter so that it doesn't subclass Winforms.IMessageFilter, you get an error saying there's no AddMessageFilter method matching the given arguments.
If you rename or remove the PreFilterMessage() method, you get an error that the Python object doesn't have a PreFilterMessage method.
Any suggestions on what I'm doing wrong, and/or how to fix it? Or how to get more debugging information that could point at the source of the segfault?
It appears that this is a bug in Python.net itself, relating to a problem with marshaling byref arguments. Details can be found here.
I am trying to read my mail and see the received time in outlook 2016 using MAPI.
I am able to see the subject of the mail not able to see the receivedTime of the mail. I know that "Receivedtime" is there to get the received time of the mail, but while the program is executed,
a popup is coming, telling that python has stopped working
I know it is not due to any machine problem rather some problem in my code.
Here is my code.
def arrange(mailbox):
global spam
timeperiod() # stores required date in spam[] list
msgs=mailbox.Items
msgs.Sort("[ReceivedTime]", True)
p=msgs.restrict(" [ReceivedTime] >= '"+spam[2]+"'") #and [ReceivedTime] >= '" +spam[1]+"'
print(len(p))
'''
for m in list1:
if m.Unread:
m.Unread=False
'''
return p
#Calling it
ctm1=arrange(ctm)
print(len(ctm1)) #Working fine
for message in ctm1:
print (message.subject) #Also works good
print (message.receivedTime) # Here is the problem, it's not showing
]1
i have tried Senton property as well, but it's not working . So any guesses why the senton or receivedTime properties are not working???
updated code :
def printlist(box1) :
print(len(box1))
for message in box1:
if message.Class==43 :
# print('true')
print (message)
#.senderEmailAddress) #working
#print(message.SentOn.strftime("%d-%m-%y")) #not working
#print(message.body)
#print(message.UnRead)
#print (message.receivedTime) #not working
#print('-----------')
I was also having trouble with the .ReceivedTime breaking the program when I compile the .py script to an .exe using auto-py-to-exe.
Here's where the error occurs underneath this try: statement
try:
received_year = str(email.ReceivedTime)[0:4]
received_month = str(email.ReceivedTime)[5:7]
received_day = str(email.ReceivedTime)[8:10]
This works PERFECTLY well inside of my IDE (PyCharm), but once I've compiled it to .exe, it breaks here.
I've updated pywin32 to the most up-to-date version (228) and also tried using 224 to see if it was a version issue (it wasn't).
BUT!! Going through this, I FOUND THE BUG! When you compile to .exe with auto-py-to-exe, it does not include the package "win32timezone" which the .ReceivedTime portion needs to run correctly. So you need to import this package for it to work.
All you have to do to fix this is include this at the top of your .py script before compiling to .exe:
import win32timezone
Let me know if this works for anyone else facing this issue!
Most likely you run into an item other than MailItem - you can also have ReportItem and MeetingItem objects in your Inbox; none of them expose the ReceivedTime property.
Check that message.Class property == 43 (olMail) before accessing any other MailItem specific properties.
Please try the below:
print([x.ReceivedTime for x in messages])
i have created a pretty simple MapReduce pipeline, but i am having a cryptic:
PipelineSetupError: Error starting production.cron.pipelines.ItemsInfoPipeline(*(), **{})#a741186284ed4fb8a4cd06e38921beff:
when i try to start it. This is the pipeline code:
class ItemsInfoPipeline(base_handler.PipelineBase):
"""
"""
def run(self):
output = yield mapreduce_pipeline.MapreducePipeline(
job_name="items_job",
mapper_spec="production.cron.mappers.items_info_mapper",
input_reader_spec="mapreduce.input_readers.DatastoreInputReader",
mapper_params={
"input_reader": {
"entity_kind": "production.models.Transaction"
}
}
)
yield ItemsInfoStorePipeline(output)
class ItemsInfoStorePipeline(base_handler.PipelineBase):
"""
"""
def run(self, statistics):
print statistics
return "OK"
Of course i have double checked that the mapper path is right, and take into account that ItemsInfoStorePipeline is not doing anything because i am focusing the have the pipeline started, which is not happening.
It is all triggered by a Flask view, the following:
class ItemsInfoMRJob(views.MethodView):
"""
It's based on transacions.
"""
def get(self):
"""
:return:
"""
pipeline = ItemsInfoPipeline()
pipeline.start()
redirect_url = "%s/status?root=%s" % (pipeline.base_path, pipeline.pipeline_id)
return flask.redirect(redirect_url)
I am using GoogleAppEngineMapReduce==1.9.22.0
Thanks for any help.
UPDATE
The above code works once deployed.
UPDATE 2
Apparently there's more people dealing with this:
https://github.com/GoogleCloudPlatform/appengine-mapreduce/issues/103
I'm updating this. I have a code base that's using pipelines and works fine in OSX. I had another developer using OSX that simply nothing I do seems to get it working, he gets this:
Encountered unexpected error from ProtoRPC method implementation: PipelineSetupError
I've tried swapping versions around and making our PC's match perfectly and it continue to happen. I finally broke down and built an Ubuntu image in docker. I'm also doing my best to match our versions of AppEngine and libraries perfectly.
It also refuses to start with the same message. I started working through the libraries uncommenting the part that's swallowing the error but that was a long rabbit hole I started down because lots of stuff above it also seems to be swallowing up whatever is going on.
I get this code on CEF Python 3 (link)
...
self.container = gtk.DrawingArea()
self.container.set_property('can-focus', True)
self.container.connect('size-allocate', self.OnSize)
self.container.show()
...
windowID = self.container.get_window().handle
windowInfo = cefpython.WindowInfo()
windowInfo.SetAsChild(windowID)
self.browser = cefpython.CreateBrowserSync(windowInfo,
browserSettings={},
navigateUrl=GetApplicationPath('example.html'))
...
This code [self.container.get_window().handle] don't work with PyGI and GTK3.
I trying port the code from GTK2 to GTK3, how I can do this?
Edited:
After some search, I found a tip to make get_window work: I call: self.container.realize() before self.container.get_window(). But I cant't get Window Handle yet.
I need put CEF3 window inside a DrawingArea or any element. How I can do this with PyGI?
Edited:
My environment is:
Windows 7
Python 2.7 and Python 3.2
Sadly there seems to be no progress on the python gobject introspection to fix this and make gdk_win32_window_get_handle available (reported a bug in the gnome bugtracker quite a while ago) - it is also quite needed for Python GStreamer and Windows ...
So I followed the suggestion of totaam and used ctypes to access gdk_win32_window_get_handle. Took me forever since I had no experience with this - and well it is somehow quite an ugly hack - but well when needed...
Here is the code:
Gdk.threads_enter()
#get the gdk window and the corresponding c gpointer
drawingareawnd = drawingarea.get_property("window")
#make sure to call ensure_native before e.g. on realize
if not drawingareawnd.has_native():
print("Your window is gonna freeze as soon as you move or resize it...")
ctypes.pythonapi.PyCapsule_GetPointer.restype = ctypes.c_void_p
ctypes.pythonapi.PyCapsule_GetPointer.argtypes = [ctypes.py_object]
drawingarea_gpointer = ctypes.pythonapi.PyCapsule_GetPointer(drawingareawnd.__gpointer__, None)
#get the win32 handle
gdkdll = ctypes.CDLL ("libgdk-3-0.dll")
hnd = gdkdll.gdk_win32_window_get_handle(drawingarea_gpointer)
#do what you want with it ... I pass it to a gstreamer videosink
Gdk.threads_leave()
You must first import GdkX11 for get_xid() to be available on the returned GdkX11Window.
from gi.repository import GdkX11
...
-windowID = self.container.get_window().handle
+windowID = self.container.get_window().get_xid()
The answer advising you to use .handle or .get_xid() works on GTK2, but not with GTK3 or on MS Windows, which are part of your question.
I've done a lot of digging and found that there is a function in GTK3 which does what you want: gdk_win32_window_get_handle, but sadly it is not available in the gi bindings.
You may be able to get to it using ctypes or Cython (which is what I am going to do).
I develop a critical application used by a multi-national company. Users in offices all around the globe need to be able to install this application.
The application is actually a plugin to Excel and we have an automatic installer based on Setuptools' easy_install that ensures that all a project's dependancies are automatically installed or updated any time a user switches on their Excel. It all works very elegantly as users are seldom aware of all the installation which occurs entirely in the background.
Unfortunately we are expanding and opening new offices which all have different proxy settings. These settings seem to change from day to day so we cannot keep up with the outsourced security guys who change stuff without telling us. It sucks but we just have to work around it.
I want to programatically detect the system-wide proxy settings on the Windows workstations our users run:
Everybody in the organisazation runs Windows XP and Internet Explorer. I've verified that everybody can download our stuff from IE without problems regardless of where they are int the world.
So all I need to do is detect what proxy settings IE is using and make Setuptools use those settings. Theoretically all of this information should be in the Registry.. but is there a better way to find it that is guaranteed not to change with people upgrade IE? For example is there a Windows API call I can use to discover the proxy settings?
In summary:
We use Python 2.4.4 on Windows XP
We need to detect the Internet Explorer proxy settings (e.g. host, port and Proxy type)
I'm going to use this information to dynamically re-configure easy_install so that it can download the egg files via the proxy.
UPDATE0:
I forgot one important detail: Each site has an auto-config "pac" file.
There's a key in Windows\CurrentVersion\InternetSettings\AutoConfigURL which points to a HTTP document on a local server which contains what looks like a javascript file.
The pac script is basically a series of nested if-statements which compare URLs against a regexp and then eventually return the hostname of the chosen proxy-server. The script is a single javascript function called FindProxyForURL(url, host)
The challenge is therefore to find out for any given server which proxy to use. The only 100% guaranteed way to do this is to look up the pac file and call the Javascript function from Python.
Any suggestions? Is there a more elegant way to do this?
Here's a sample that should create a bullet green (proxy enable) or red (proxy disable) in your systray
It shows how to read and write in windows registry
it uses gtk
#!/usr/bin/env python
import gobject
import gtk
from _winreg import *
class ProxyNotifier:
def __init__(self):
self.trayIcon = gtk.StatusIcon()
self.updateIcon()
#set callback on right click to on_right_click
self.trayIcon.connect('popup-menu', self.on_right_click)
gobject.timeout_add(1000, self.checkStatus)
def isProxyEnabled(self):
aReg = ConnectRegistry(None,HKEY_CURRENT_USER)
aKey = OpenKey(aReg, r"Software\Microsoft\Windows\CurrentVersion\Internet Settings")
subCount, valueCount, lastModified = QueryInfoKey(aKey)
for i in range(valueCount):
try:
n,v,t = EnumValue(aKey,i)
if n == 'ProxyEnable':
return v and True or False
except EnvironmentError:
break
CloseKey(aKey)
def invertProxyEnableState(self):
aReg = ConnectRegistry(None,HKEY_CURRENT_USER)
aKey = OpenKey(aReg, r"Software\Microsoft\Windows\CurrentVersion\Internet Settings", 0, KEY_WRITE)
if self.isProxyEnabled() :
val = 0
else:
val = 1
try:
SetValueEx(aKey,"ProxyEnable",0, REG_DWORD, val)
except EnvironmentError:
print "Encountered problems writing into the Registry..."
CloseKey(aKey)
def updateIcon(self):
if self.isProxyEnabled():
icon=gtk.STOCK_YES
else:
icon=gtk.STOCK_NO
self.trayIcon.set_from_stock(icon)
def checkStatus(self):
self.updateIcon()
return True
def on_right_click(self, data, event_button, event_time):
self.invertProxyEnableState()
self.updateIcon()
if __name__ == '__main__':
proxyNotifier = ProxyNotifier()
gtk.main()
As far as I know, In a Windows environment, if no proxy environment variables are set, proxy settings are obtained from the registry's Internet Settings section. .
Isn't it enough?
Or u can get something useful info from registry:
HKEY_CURRENT_USER\Software\Microsoft\Windows\CurrentVersion\Internet Settings\ProxyServer
Edit:
sorry for don't know how to format comment's source code, I repost it here.
>>> import win32com.client
>>> js = win32com.client.Dispatch('MSScriptControl.ScriptControl')
>>> js.Language = 'JavaScript'
>>> js.AddCode('function add(a, b) {return a+b;}')
>>> js.Run('add', 1, 2)
3