Home > Article > Backend Development > Completely master Python automatic operation GUI PyAutoGUI
This article brings you relevant knowledge about Python, which mainly sorts out the issues related to the automatic operation GUI PyAutoGUI. The desktop automation operation is by positioning the mouse on the desktop, and then based on The corresponding operation is performed at the positioned position. Let’s take a look at it. I hope it will be helpful to everyone.
【Related recommendations: Python3 video tutorial】
Our protagonist today ispyautogui
, pyautogui
is a pure Python GUI automation tool that allows the program to automatically control a series of mouse and keyboard operations to achieve the purpose of automated testing.
Module installation, open cmd, enter:
pip install pyautogui
Successful display: Successfully installed PyTweening-1.0.4 mouseinfo-0.1.3 pyautogui-0.9.53 pygetwindow-0.0 .9 pymsgbox-1.0.9 pyperclip-1.8.2 pyrect-0.2.0 pyscreeze-0.1.28
After installation, you can use it directly.
Guide package
import pyautogui
The most basic desktop operation is mouse operation. We can control the movement of the mouse:
# 移动鼠标 pyautogui.moveTo(200,400,duration=2) pyautogui.moveRel(200,500,duration=2)
The entire desktop has the upper left corner as the origin of the coordinate axis, and all operations are based on This origin is used to determine the operating position.
The first line moves the mouse to the specified pixel (200,400) position;
The second line of code moves the mouse 200px to the right and 500px downward according to the current point.
There is a common parameter duration in both lines of code. This parameter represents the movement time
, that is, the movement operation is completed within the specified time, and the unit is seconds
.
Run these two lines of code and observe the changes of the screen mouse. Isn’t it amazing?
Usually, our mouse has two buttons on the left and right, and an advanced mouse has a button in the middle.
My mouse only has two buttons, and there is no button in the middle. Alas~pyautogui
There are corresponding processing for these three button operations:
# 鼠标点击,默认左键 pyautogui.click(100,100) # 单击左键 pyautogui.click(100,100,button='left') # 单击右键 pyautogui.click(100,300,button='right') # 单击中间 pyautogui.click(100,300,button='middle')
Mouse click, if the button
parameter is not specified, the default is Click the left button, the first two parameters are the location of the click coordinates.
Run this code and see what happens to your desktop?
In addition to click operations, the mouse also has double-click operations:
# 双击左键 pyautogui.doubleClick(10,10) # 双击右键 pyautogui.rightClick(10,10) # 双击中键 pyautogui.middleClick(10,10) # 双击屏幕 pyautogui.tripleClick(10,10)
The operation function is also very simple, I believe everyone can understand it at a glance , if you can’t understand it at first glance, please take a few more glances!
Friends who are familiar with the front-end may immediately think that mouse operations have a process of pressing and releasing, and our screen operations also have corresponding controls:
# 鼠标按下 pyautogui.mouseDown() # 鼠标释放 pyautogui.mouseUp()
We can control the mouse to drag to the specified coordinate position, and set the operation time and keys:
pyautogui.dragTo(100,300,duration=1,button='right')
This operation effect is similar to the previous movement.
Based on the previous experience of moving, we also drag the mouse in the direction:
pyautogui.dragRel(100,300,duration=4,button='left')
In desktop operations, we sometimes need to scroll the mouse to reach up or to At this time, we can use the scroll
function to control:
pyautogui.scroll(30000)
parameter is an integer, indicating how many units to scroll up or down. This unit may be different depending on different operating systems. If you scroll up, pass in a positive integer, and if you scroll down, pass in a negative integer.
For example, how do you usually operate the keyboard by inputting an exclamation point (!)?
按住 shift 按键,然后再按住 1 按键,就可以了。用pyautogui
控制就是:
pyautogui.keyDown('shift') pyautogui.press('1') pyautogui.keyUp('shift')
运行上面的代码,如果你的鼠标是放在编辑框中,你会得到一个感叹号!
我们还可以直接输出内容:
pyautogui.typewrite('python', 1)
第一个参数是输出的内容,第二个参数是间隔时间,单位是秒。
运行上面代码,你的编辑器里面就会每隔1秒钟按顺序输出 python 的6个字母。
有时我们需要输入键盘的一些特殊的符号按键,比如 换行、方向键等,这些有相对应的键盘字符串表示:
pyautogui.typewrite(['p','y','t','h','o','n','enter']) pyautogui.typewrite(['p','y','t','h','o','n',' '])
运行上面代码,编辑器里面就会输出 python 之后换行。
其他特殊按键对应的字符串请参考官方说明。
如果我要复制一个内容,大部分情况下会使用快键键 ctrl + c,按照上面讲的,我们应该这么实现:
pyautogui.keyDown('ctrl') pyautogui.keyDown('c') pyautogui.keyUp('c') pyautogui.keyUp('ctrl')
实现的效果和上面的4行代码相同。
键盘字符串 说明
enter(或return 或 \n) 回车
esc ESC键
shiftleft, shiftright 左右SHIFT键
altleft, altright 左右ALT键
ctrlleft, ctrlright 左右CTRL键
tab (\t) TAB键
backspace, delete BACKSPACE 、DELETE键
pageup, pagedown PAGE UP 和 PAGE DOWN键
home, end HOME 和 END键
up, down, left,right 箭头键
f1, f2, f3…. f12 F1…….F12键
volumemute, volumedown,volumeup 声音变大变小静音(有些键盘没有)
pause PAUSE键,暂停键
capslock CAPS LOCK 键
numlock NUM LOCK 键
scrolllock SCROLLLOCK 键
insert INSERT键
printscreen PRINT SCREEN键
winleft, winright Win键(windows )
command command键(Mac OS X )
option option(Mac OS X)
我们先来假设一个场景:我现在要在屏幕上找到一个红色的点,你会怎么做?通常的做法是拿到这个红色点的颜色值,然后再对屏幕上的点逐个进行比较,直到找到为止。
pyautogui
为我们这个操作场景提供了支持,分别有三个函数可以完成这三件事情。
im = pyautogui.screenshot() im.save('screenshot.png') rgb = im.getpixel((100, 500)) print(rgb) match = pyautogui.pixelMatchesColor(500,500,(12,120,400)) print(match)
第一个是获取屏幕截图函数,它可以返回一个 Pillow 的 image 对象; 第二个是获取屏幕截图中指定坐标点的颜色,返回 rgb 颜色值;第三个是将指定坐标点的颜色和目标的颜色进行比对,返回布尔值。
我们再来升级一下需求:
我现在要在屏幕上找到 edge 浏览器的图标,你会怎么做?
通常的做法是先知道 edge 浏览器的图标长啥样,是绿色还是蓝色,是胖的还是瘦的,对吧?然后再在屏幕上去进行图标的匹配,直到找到一个图标跟我们目标图标一样,就得到了结果。
于是,我们的代码如下:
# 图像识别(一个) oneLocation = pyautogui.locateOnScreen('1.png') print(oneLocation) # 图像识别(多个) allLocation = pyautogui.locateAllOnScreen('1.png') print(list(allLocation))
你可以在桌面上将某个应用的图标截取下来,保存为图片,然后使用上面几行代码来识别,识别成功,你会返回类似下面的结果:
Box(left=20, top=89, width=33, height=34) [Box(left=20, top=89, width=33, height=34)]
这就是图片在桌面的位置,如果找不到图片,就会返回 None。
当你在模拟一个桌面操作的时候,如果有分支操作需要根据实际情况来判断,你是不是需要有一个地方可以让你选择走哪个分支?
pyautogui
贴心地考虑到了这种情况,你可以通过弹出一个选择框来中断当前的操作,选择操作分支。
way = pyautogui.confirm('领导,该走哪条路?', buttons=['农村路', '水路', '陆路']) print(way)
这里就是我们 HTML 页面的 confirm 选择框,选择了选项之后,我们可以获取到选择的选项,然后基于这个选项做判断,进入相应的操作分支。
除了选择确认框之外,还有其他一些提示信息框:
# 警告框 alert = pyautogui.alert(text='警告!敌军来袭!', title='警告框') print(alert) # 密码框 password = pyautogui.password('请输入密码') print(password) # 普通输入框 input = pyautogui.prompt('请输入指令:') print(input)
pyautogui
的基本知识就给大家介绍到这里,这个 python 模块的功能十分强大,函数都非常简单,对 python 初学者比较友好。学了这些基本知识之后,你可以运用这些基本知识的组合,去实现一些有趣的桌面自动化操作,快去尝试一把吧!
我们通过今天的基础知识的学习,我们可以通过
pyautogui
+PIL
结合来实现网页自动点赞的实战练习。
我们先介绍一个PIL库下一个裁剪图片的函数crop。
使用PIL裁切图片使用PIL需要引用Image,使用Image的open(file)方法可以返回打开的图片,使用crop((x1,y1,x2,y2))
方法可以对图片做裁切。
该区域是一个元组,(x1,y1)
是图片左上角的坐标,(x2,y2)
是图片右下角的坐标。
Python Imaging Library 使用左上角为 (0, 0)的坐标系统。
首先我们要知道点赞的图片的样子
# 截取整个屏幕 im=pyautogui.screenshot() # 截取点赞图片 om=im.crop((1754,595,1803,631)) #将图片保存供pyautogui.locateOnScreen()使用 om.save("dianzan.png")
根据截取的屏幕仅截取带赞的手势图片,可以用
pyautogui.mouseInfo()
获取图片的位置(1754,595,1803,631)
或者通过pyautogui.location()来确定首尾坐标
在当前目录下得到如下图片。
# 点赞函数 def zan(count): #为了防止一直运行下去,点赞次数由自己定义 time.sleep(0.5) # 寻找刚才保存点赞手势图片 xy = pyautogui.locateOnScreen('dianzan.png') # 寻找图片的中心 center = pyautogui.center(xy) # 点赞 pyautogui.click(center) print('点赞成功') count = 3 # 点赞次数 while range(count): if pyautogui.locateOnScreen('dianzan.png'): zan(count) # 调用点赞函数 count-=1 else: #本页没有图片后,滚动鼠标; pyautogui.scroll(-500) print('没有找到目标,屏幕下滚~') count -= 1 print('点赞结束!')
未点赞之前的图片是这样的:
自动点赞之后:
是不是很厉害呢?其实pyautogui
还可以做很多东西,比如说自动发信息,其原理还是不变的,找到对应的位置,输入内容,发送就可以了,在这里我就不演示啦!有兴趣的可以去试试哦~~~
完整代码
from PIL import Image import pyautogui import time # 截取整个屏幕 im=pyautogui.screenshot() # 截取点赞图片 om=im.crop((1754,595,1803,631)) #将图片保存供pyautogui.locateOnScreen()使用 om.save("dianzan.png") # 点赞函数 def zan(count): #为了防止一直运行下去,点赞次数由自己定义 time.sleep(0.5) # 寻找刚才保存点赞手势图片 xy = pyautogui.locateOnScreen('dianzan.png') # 寻找图片的中心 center = pyautogui.center(xy) # 点赞 pyautogui.click(center) print('点赞成功') count = 3 # 点赞次数 while range(count): if pyautogui.locateOnScreen('dianzan.png'): zan(count) # 调用点赞函数 count-=1 else: #本页没有图片后,滚动鼠标; pyautogui.scroll(-500) print('没有找到目标,屏幕下滚~') count -= 1 print('点赞结束!')
【相关推荐:Python3视频教程 】
The above is the detailed content of Completely master Python automatic operation GUI PyAutoGUI. For more information, please follow other related articles on the PHP Chinese website!