Use Python to simulate the like assistant

In the past, the company has been arranging various learning tasks, **There are many chapters**, some are technical and some are policy-related, and if you do not complete the learning tasks on time, you will be notified. However, these learning contents are not necessarily helpful, so everyone has developed the habit of brushing up on the class, that is, clicking on the ** and letting it hang there. But there is a problem,**The process will detect whether the user has operated,If the detection is not operated for a long time, the pop-up box will prompt,Only click the continue button on the pop-up box to continue**,This leads us to often pay attention to whether the page is paused when brushing the class,The efficiency of brushing the class is a bit low。 To this end, a desktop icon detection gadget was made, which detected that after a specific icon appeared on the desktop, it simulated mouse clicks to complete the function of continuing.

That's where one comes inpythonLibrary——pyautogui。Here's how this library is used.,And use this library to make an assistant for brushing likes.,But it's recommended not to use it to really brush likes.。

pyautoguiis a Python library for programmatically controlling mouse and keyboard operations that can be used to automate tests, tasks, and other requirements. Here's onepyautoguiBasic tutorials.

pip install pyautogui

import pyautogui

Move the mouse to the (100, 100) position of the screen. 
pyautogui.moveto(100, 100, duration=1)

useclickFunctions can simulate mouse clicks or specify where they are clicked.

Left-click at the current location. 

pyautogui.click()
Right-click on the screen (200, 200). 
pyautogui.click(200, 200, button='right')

Scroll up. 

pyautogui.scroll(200)
Scroll down. 
pyautogui.scroll(-200)

Press and release the spacebar. 

pyautogui.press('space')
Enter a string. 
pyautogui.write('hello, pyautogui!', interval=0.25)

pyautoguiScreenshots of the screen can be captured, which is useful for automated testing and monitoring.

Capture a full-screen screenshot. 

screenshot = pyautogui.screenshot()
screenshot.s**e('screenshot.png')

location = pyautogui.locateonscreen('button.png', confidence=0.9)

Open the Zhihu follow pageFirst of all, manually open the Zhihu follow page and prepare to start the automation operation. The screen detects the approval icon: The tool will constantly scan the screen for a specific endorsement icon. Simulate user reading: Once the approval icon is found, the tool simulates clicking on the line above the icon to open the full text, and then simulates scrolling up and down with the mouse to simulate the reading process. Auto-like: After reading, the tool simulates clicking the agree icon to complete the like operation. Exit reading and continue testing: After recognizing the collapsed icon, exit the reading state and simulate mouse swiping to detect the next approval icon. Through this series of operations, the tool can automatically like articles on Zhihu, simulating the normal browsing behavior of users.

Approve icon

Collapse icon

Liked icon.

Project ** is hosted ongithub, which can be accessed via the following link*** and icon: github-

import random

import time
import pyautogui
from pynput import keyboard
import threading
Listen for the ESC key identifier (stop running when the ESC key is heard).
listener_esc = false
The path of the image sample. 
agree_path = 'image/agree.png'
read_more_path = 'image/read_more.png'
close_path = 'image/close.png'
The length of the break (in s, preferably longer).
sleep_time = 1
Confidence level (it is recommended to increase it, otherwise it is easy to touch it by mistake).
confidence = 0.8
Mouse scroll speed. 
scroll_speed = 10
def on_press(key):
The action when the key is pressed. 
try:if key == keyboard.key.esc:
When the ESC key is detected, the message is output and the listening is stopped. 
print('The esc key is pressed to stop listening. ')
ESC has been pressed. 
global listener_esc
listener_esc = true
Returns false to stop ***
return false
except attributeerror:
passdef get_a_number(a, b):
Generate a random number [a, b].
return:
return random.random().randint(a, b)
def on_release(key):
The operation when the key is released. 
passdef start_listener():
Listen for the ESC key. 
return:
with keyboard.listener(on_press=on_press, on_release=on_release) as listener:
listener.join()
def reading():
Simulated reading. 1. Roll the mouse randomly. 
2. Scroll the number of pixel rows randomly. 
return:
Scroll the mouse up and down randomly ().
scroll_times = get_a_number(1, 3)
while scroll_times > 0 and not listener_esc:
The number of random scrolling pixel rows. 
lines = get_a_number(50, 100)
Swipe down. 
for _ in range(int(lines / scroll_speed)):
pyautogui.scroll(-scroll_speed)
time.sleep(sleep_time)
Swipe up to be able to return to the position just now. 
for _ in range(int(lines / scroll_speed)):
pyautogui.scroll(scroll_speed)
time.sleep(sleep_time)
scroll_times -= 1
def close():
The simulation closes the reading, and the loop is to prevent the coordinates from being inaccurate and not turning off. 
return:
try:close_location = pyautogui.locateonscreen(close_path, confidence=confidence)
while close_location:
Calculate the image center point. 
center = pyautogui.center(close_location)
pyautogui.click(center)
print(f"close-clicked on ")
time.sleep(sleep_time)
close_location = pyautogui.locateonscreen(close_path, confidence=confidence)
except pyautogui.failsafeexception:
print("close not found")
return
def do_read(x, y):
Simulate clicking "Read More".
param x:
param y: Agree, click about 50 pixels above to view the full article. 
return:
Look for an image on the screen to read the full article. 
if not listener_esc:
Click here to view the full article. 
pyautogui.click(x, y)
print(f"read more-clicked on , ")
Load the content. 
time.sleep(sleep_time)
Read. reading()
def do_click():
Mock likes. :return:
Check if the esc key is pressed, and if so, exit the loop. 
while not listener_esc:
Give yourself some time to get ready to switch to the target app window. 
time.sleep(sleep_time)
try: Find the favor image on the screen, and the confidence parameter can be adjusted to suit the accuracy requirements of the image match. 
agree_location = pyautogui.locateonscreen(agree_path, confidence=confidence)
if agree_location:
Calculate the image center point. 
center = pyautogui.center(agree_location)
x, y = center
Simulated reading. 
do_read(x, y - 50)
Thumbs up. if not listener_esc:
# pyautogui.click(x, y) prevents deviations caused by movement of coordinates and reacquires approval coordinates. 
agree_location = pyautogui.locateonscreen(agree_path, confidence=confidence)
center = pyautogui.center(agree_location)
pyautogui.click(center)
print(f"agree - clicked on ")
Click the Close button. 
close()
except pyautogui.imagenotfoundexception:
print("agree not found. trying again...")
Pause briefly so that loops don't execute too often. 
time.sleep(sleep_time)
Scroll down, about 200 pixels down. 
for _ in range(int(200 / scroll_speed)):
pyautogui.scroll(-scroll_speed)
print("exited.")
Create and start a thread listening to the esc key. 
listener_thread1 = threading.thread(target=start_listener)
listener_thread1.start()
The ** here can be executed without being blocked. 
print("like mode begin...")
do_click()
print("like mode end...")

pip install pyautogui pynput opencv-python

**The rest intervals in the demo were shortened and doubled.