I have a memory leak in my pyautogui script I made. I have no idea where it's coming from or what might be causing it.
The script is made to perfectly press space for a skill check in a game. So it just checks for when a pixel (out of an array) turns white and then changes color indicating the skill check bar has passed over it.
I don't feel like I'm doing anything unusual here that would cause a memory leak, it wouldn't be on the pyautogui side would it?
import pyautogui as auto
import keyboard
import random
import time
import random
circleLocations = [(1023, 522), (1025, 527), (1025, 537), (1024, 548), (1022, 560), (1018, 568)]
def delay(value1 = 0.05, value2 = 0.2):
return random.uniform(value1,value2)
def perfectSpacePress(x, y):
while auto.pixelMatchesColor(x, y, (255,255,255)):
print("waiting")
time.sleep(delay(0.1,0.15 ))
auto.press(' ')
print("PressedCircleSpace")
def checkCircle():
for location in circleLocations:
if auto.pixelMatchesColor(location[0], location[1], (255,255,255)):
perfectSpacePress(location[0], location[1])
try:
while True:
checkCircle()
time.sleep(0.1)
except KeyboardInterrupt:
print('\nDone.')
Related
Terminal Running
I'm making a Python code to get me the real timevalues on some currencies to test the API, how do I keep it runnning and printing just the updated value, but without print the lines all over again, basically printing the currency names one time and keep updating the values without closing the terminal.
Here's my code:
import requests
import json
import colored
from colored import fg
print("Valor atualizado:\n\n")
cotacoes = requests.get("https://economia.awesomeapi.com.br/last/USD-BRL,EUR-BRL,BTC-BRL,ETH-BRL")
cotacoes = cotacoes.json()
dolar = cotacoes['USDBRL']["bid"]
euro = cotacoes['EURBRL']["bid"]
bitcoin = cotacoes['BTCBRL']["bid"]
eth = cotacoes['ETHBRL']["bid"]
red = fg('red')
blue = fg('blue')
yellow = fg('yellow')
green = fg('green')
white = fg('white')
print(green + "Dólar:", dolar,"reais")
print(red + "Euro:", euro, "reais")
print(yellow + "Bitcoin:", bitcoin, "reais")
print(blue + "Etherium:", eth,"reais")
print(white)
Generally to override already written text you need to do some cursor positioning, but the simplest way is to clear the terminal in every iteration.
import os
import time
while True:
os.system('cls' if os.name == 'nt' else 'clear') # clears terminal
<get needed values>
<print everything>
time.sleep(1)
I am running a python script on a raspberry-pi.
Essentially, I would like a camera to take a picture every 5 seconds, but only if I have set a boolean to true, which gets toggled on a physical button.
initially I set it to true, and then in my while(true) loop, I want to check to see if the variable is set to true, and if so, start taking pictures every 5 seconds. The issue is if I use something like time time.sleep(5), it essentially freezes everything, including the check. Combine that with the fact that I am using debouncing for the button, it then becomes impossible for me to actually toggle the script since I would have to press it exactly after the 5s wait time, right for the value check... I've been searching around and I think the likely solution would have to include threading, but I can't wrap my head around it. One kind of workaround I thought of would be to look at the system time and if the seconds is a multiple of 5, then take picture (all within the main loop). This seems a bit sketchy.
Script below:
### Imports
from goprocam import GoProCamera, constants
import board
import digitalio
from adafruit_debouncer import Debouncer
import os
import shutil
import time
### GoPro settings
goproCamera = GoProCamera.GoPro()
### Button settings
pin = digitalio.DigitalInOut(board.D12)
pin.direction = digitalio.Direction.INPUT
pin.pull = digitalio.Pull.UP
switch = Debouncer(pin, interval=0.1)
save = False #this is the variable
while(True):
switch.update()
if switch.fell:
print("Pressed, toggling value")
save = not save
if save:
goproCamera.take_photo()
goproCamera.downloadLastMedia()
time.sleep(5)
Here's something to try:
while(True):
switch.update()
if switch.fell:
print("Pressed, toggling value")
save = not save
if save:
current_time = time.time()
if current_time - last_pic_time >= 5:
goproCamera.take_photo()
goproCamera.downloadLastMedia()
last_pic_time = current_time
Depending on exactly what sort of behavior you want, you may have to fiddle with when and how often time.time() is called.
Cheers!
Maybe something like this?
import threading
def set_interval(func, sec):
def func_wrapper():
set_interval(func, sec)
func()
t = threading.Timer(sec, func_wrapper)
t.start()
return t
We call the function above inside the main loop.
Wrap your while loop content on a function:
def take_photo:
goproCamera.take_photo()
goproCamera.downloadLastMedia()
Now we create a flag initially set to False to avoid creating multiple threads.
Notice that I did this before the while loop. We just need a starting value here.
active = False
while(True):
switch.update()
if switch.fell:
print("Pressed, toggling value")
save = not save
if save: # we need to start taking photos.
if not active: # it is not active... so it is the first time it is being called or it has been toggled to save as True again.
photo_thread = set_interval(take_photo, 5) # grabbing a handle to the thread - photo_thread - so we can cancel it later when save is set to False.
active = True # marking as active to be skipped from the loop until save is False
else:
try: # photo_thread may not exist yet so I wrapped it inside a try statement here.
photo_thread.cancel() # if we have a thread we kill it
active = False #setting to False so the next time the button is pressed we can create a new one.
Let me know if it works. =)
What I ended up doing:
### Imports
from goprocam import GoProCamera, constants
import board
import digitalio
from adafruit_debouncer import Debouncer
import os
import time
import threading
### GoPro settings
gopro = GoProCamera.GoPro()
### Button settings
pin = digitalio.DigitalInOut(board.D12)
pin.direction = digitalio.Direction.INPUT
pin.pull = digitalio.Pull.UP
switch = Debouncer(pin, interval=0.1)
### Picture save location
dir_path = os.path.dirname(os.path.realpath(__file__))
new_path = dir_path+"/pictures/"
save = False
### Functions
def takePhoto(e):
while e.isSet():
gopro.take_photo()
gopro.downloadLastMedia()
fname = '100GOPRO-' + gopro.getMedia().split("/")[-1]
current_file = dir_path+'/'+fname
if os.path.isfile(current_file):
os.replace(current_file, new_path+fname) #move file, would be cleaner to download the file directly to the right folder, but the API doesn't work the way I thought it did
e.wait(5)
### Initial settings
e = threading.Event()
t1 = threading.Thread(target=takePhoto, args=([e]))
print("Starting script")
while(True):
switch.update()
if switch.fell:
#toggle value
save = not save
if save:
e.set() #should be taking pictures
else:
e.clear() #not taking pictures
if not t1.is_alive(): #start the thread if it hasn't been yet
if e.is_set():
t1.start()
I would like to take input from pySimpleGUI, feed it into a normal Python var, then feed it into a music processor as I love music.
I had already tried to use wxPython for this but was unable to even get a simple fileDialog without crashing.
from pydub import AudioSegment
from os import listdir
import numpy as np
import math
import PySimpleGUI as sg
class Dankify():
song_dir = "songs"
attenuate_db = 0
accentuate_db = 2
yeet = sg.Window('Dankify ALL THE THINGS!'). Layout([[sg.Text('Filename')], [sg.Input(), sg.FileBrowse()], [sg.OK(), sg.Cancel()] ]).Read()
event, values = yeet.Read()
yeet1 = event, values
def bass_line_freq(track):
sample_track = list(track)
# c-value
est_mean = np.mean(sample_track)
# a-value
est_std = 3 * np.std(sample_track) / (math.sqrt(2))
bass_factor = int(round((est_std - est_mean) * 0.005))
return bass_factor
songfile = yeet1
for filename in listdir(songfile):
sample = AudioSegment.from_mp3(songfile)
filtered = sample.low_pass_filter(bass_line_freq(sample.get_array_of_samples()))
combined = (sample - attenuate_db).overlay(filtered + accentuate_db)
combined.export("exports/" + filename.replace(".mp3", "") + "-export.mp3", format="mp3")
However, it just does nothing, not even processing it. A reminder that I am using some open-source code and that I'm a beginner which knows nothing about how all this works and am trying to build real stuff to gain experience. Thanks!
I guess you are missing the "event loop".
Try something like this, hope it helps.
import sys
if sys.version_info[0] >= 3:
import PySimpleGUI as sg
else:
import PySimpleGUI27 as sg
layout = [[sg.Text('Your typed chars appear here:'), sg.Text('', key='_OUTPUT_') ],
[sg.Input(do_not_clear=True, key='_IN_')],
[sg.Button('Show'), sg.Button('Exit')]]
window = sg.Window('Window Title').Layout(layout)
while True: # Event Loop
event, values = window.Read()
print(event, values)
if event is None or event == 'Exit':
break
if event == 'Show':
# change the "output" element to be the value of "input" element
window.FindElement('_OUTPUT_').Update(values['_IN_'])
window.Close()
You're doing 2 Read calls.
Try changing to this:
yeet = sg.Window('Dankify ALL THE THINGS!').Layout(
[[sg.Text('Filename')], [sg.Input(), sg.FileBrowse()], [sg.OK(), sg.Cancel()]])
event, values = yeet.Read()
Without the Read on the end of the first statement.
You are instantiating this class, right?
d = Dankify()
I have designed and created a program using Python 3 that reads unread messages in my Gmail inbox under two labels.
By using tkinter I have two lovely boxes that display the total messages in each label. One for sales of one particular product and the other for another.
They use the update loop to recheck each label every few seconds.
Then after the business day, I use a cleanup script in Gmail that flushes the inboxes two labels.
The program is for use on my team's sales floor. They can see daily, the number of sales, and get a real-time readout to the success of certain marketing campaigns. It has done wonders for morale.
Now I would like to implement some sounds when sales go up. A cliche bell ring, a "chhaaaching" perhaps, you get the drift.
So, I am currently tackling with my limited knowledge and have searched all throughout StackOverflow and other sites for an answer. My guess is that I need something like the following...
"if an integer value changes on the next loop from it's previous value, by an increment of 1 play soundA, or, play soundB if that value decreases by 1."
I can't for the life of me figure out what would be the term for 'increases by 1', I am also clueless on how to attach a sound to any changes made to the integer on the proceeding loop. Help!!
If I wasn't clear enough, I am more than happy to explain and go into this further.
Thank you so much guys.
Here is my code as it stands so far...
#! /usr/bin/python3
import imaplib
import email
import tkinter as tk
WIDTH = 1000
HEIGHT = 100
def update():
mail=imaplib.IMAP4_SSL('imap.gmail.com',993)
mail.login('email#gmail.com','MyPassword')
mail.select("Submissions")
typ, messageIDs = mail.search(None, "UNSEEN")
FirstDataSetSUBS = str(messageIDs[0], encoding='utf8')
if FirstDataSetSUBS == '':
info1['text'] = 'no submissions'
else:
SecondDataSetSUBS = FirstDataSetSUBS.split(" ")
nosubs = len(SecondDataSetSUBS)
nosubs = int(nosubs)
info1['text'] = '{} submission[s]'.format(nosubs)
subs.after(1000, update)
def update_2():
mail=imaplib.IMAP4_SSL('imap.gmail.com',993)
mail.login('email#gmail.com','MyPassword')
mail.select("Memberships")
typ, messageIDs = mail.search(None, "UNSEEN")
FirstDataSetMSGS = str(messageIDs[0], encoding='utf8')
if FirstDataSetMSGS == '':
info2['text'] = 'no memberships'
else:
SecondDataSetMSGS = FirstDataSetMSGS.split(" ")
memberships = len(SecondDataSetMSGS)
memberships = int(memberships)
info2['text'] = '{} membership[s]'.format(memberships)
membs.after(1000, update_2)
membs = tk.Tk()
subs = tk.Tk()
membs.title('memberships counter')
membs.configure(background="black")
subs.title('submissions counter')
subs.configure(background="black")
x = (subs.winfo_screenwidth()//5) - (WIDTH//5)
y = (subs.winfo_screenheight()//5) - (HEIGHT//5)
subs.geometry('{}x{}+{}+{}'.format(WIDTH, HEIGHT, x, y))
info1 = tk.Label(subs, text='nothing to display', bg="black", fg="green", font="Lucida_Console 40")
info1.pack()
x = (membs.winfo_screenwidth()//2) - (WIDTH//2)
y = (membs.winfo_screenheight()//2) - (HEIGHT//2)
membs.geometry('{}x{}+{}+{}'.format(WIDTH, HEIGHT, x, y))
info2 = tk.Label(membs, text='nothing to display', bg="black", fg="red", font="Lucida_Console 40")
info2.pack()
update()
update_2()
membs.mainloop
subs.mainloop()
for playing audio you can use pyaudio module
in ubuntu, install by pip3 install pyaudio include sudo if required
import pygame
pygame.init()
increased=0
inc_music=pygame.mixer.music
dec_music=pygame.mixer.music
inc_music.load("/home/pratik/Documents/pos.wav")
dec_music.load("/home/pratik/Documents/neg.wav")
def get_inc():
global increased
a=increased
return a
def pl():
global inc_music
global dec_music
while True:
increased=get_inc()
if increased == 1:
inc_music.play()
increased=increased-1
elif increased == -1:
dec_music.play()
increased=increased+1
here increased is a global variable. Make sure whenever your sales is increased it is set to +1 and whenever it is decreased it is set to -1 and call pl function in a separate thread by using threading library in the background. Since, it is a forever loop it will continuosly run in background and ring the bells.
from threading import Thread
th=Thread(target=pl)
th.setDaemon(True)
th.start()
While writing the above I assumed at a moment there is either an increase or a decrease in sales, both dont occour simultaneously. If they do, use another global variable decreased which is also initialised with 0 and decreased by -1 each time decrease in sales. And return both of them from get_inc() .
Below code produces a label that gets updated randomly, and plays the sound files located in "C:\Windows\Media\" based on the change:
import tkinter as tk
import random
from playsound import playsound
def sound(is_positive=True):
if is_positive:
playsound(r"C:\Windows\Media\chimes.wav", False)
else:
playsound(r"C:\Windows\Media\chord.wav", False)
def random_update():
_old = label['text']
_new = random.randint(1, 100)
if _new > _old: # use (_new - _old) == 1 for checking
sound() # increment of 1 exclusively
elif _new < _old: # use (_new - _old) == -1 for checking
sound(False) # decrement of 1 exclusively
label['text'] = _new
label.after(2000, random_update)
if __name__ == '__main__':
root = tk.Tk()
label = tk.Label(root, text=1)
label.after(2000, random_update)
label.pack()
root.mainloop()
playsound is not a built-in package so it needs to be installed separately. You can install using:
pip install playsound
or:
python -m pip install playsound
in the command prompt on windows.
I have a gstreamer pipeline with a videotestsrc. My plan is to change the pipeline, but without loosing the playback position of the videotestsrc. My current code changes the pipeline and changes the quarktv effect with the radioactv effect, but the videotestsrc starts from the beginning after resuming the pipeline. Is there a way to prevent that?
from gi.repository import Gst
import time
import os
os.environ["GST_DEBUG"] = "3"
Gst.init()
pipeline = Gst.Pipeline.new()
def gen_cb(a, b, *c):
print "gen_cb", a, b, c
return Gst.PadProbeReturn.OK
# creating elements
vs = Gst.ElementFactory.make('videotestsrc')
vs.set_property('pattern', 18)
vs.set_property('is-live', 1)
pipeline.add(vs)
vc = Gst.ElementFactory.make('videoconvert')
pipeline.add(vc)
av = Gst.ElementFactory.make('autovideosink')
pipeline.add(av)
quark = Gst.ElementFactory.make('quarktv')
pipeline.add(quark)
radioactv = Gst.ElementFactory.make('radioactv')
q1 = Gst.ElementFactory.make('queue')
pipeline.add(q1)
# linking
vs.link(q1)
q1.link(quark)
quark.link(vc)
vc.link(av)
# starting
pipeline.set_state(Gst.State.PLAYING)
# sleep some time
time.sleep(2)
# modify
probe_id = q1.pads[1].add_probe(Gst.PadProbeType.BLOCK_DOWNSTREAM, gen_cb)
quark.unlink(q1)
pipeline.remove(quark)
quark.set_state(Gst.State.NULL)
pipeline.add(radioactv)
q1.link(radioactv)
radioactv.link(vc)
radioactv.set_state(Gst.State.PLAYING)
q1.pads[1].remove_probe(probe_id)
# wait until end
time.sleep(4)
It looks like a bug, please file it in https://bugzilla.gnome.org/enter_bug.cgi?product=GStreamer
When you do modifications to the pipeline a 'reconfigure' event is generated so that elements adapt to a possible new configuration (a different format/resolution might be needed). As part of this it seems like videotestsrc is also resetting its internal state so the drawing restarts from '0'.