截图及识别

主要想通过识别微信聊天截图的文字，在触发关键字时，程序自动到网页去查询内容，然后通过键盘的模拟，将查询的结果发送到微主聊天中。

# -*- coding: UTF-8 -*-
import json
import string
import sys
import time
import pyautogui
import pyperclip
import requests
from paddleocr import PaddleOCR, draw_ocr
from PIL import Image
import re
from selenium import webdriver
from selenium.webdriver.chrome.service import Service
from webdriver_manager.chrome import ChromeDriverManager
from time import sleep
from bs4 import BeautifulSoup

chrome_service = Service(ChromeDriverManager().install())
ocr = PaddleOCR(use_angle_cls=True, lang="ch")

SEND_NAMES = ['aaa','bb','fptxyy', 'cc', 'ee', 'ff-Young','ss']

DELAY_TIME = 2
HELP = """使用方法 fo   问题或关键字 
"""
# 发送微信 用copy C copy V 等实现
def send_wx_message(target, msg):
    try:
        pyperclip.copy("")
        pyperclip.copy(target)
        pyautogui.hotkey("ctrl", "f")
        time.sleep(DELAY_TIME)
        pyautogui.hotkey("ctrl", "v")
        pyautogui.press("enter")
        time.sleep(DELAY_TIME)
        pyperclip.copy("")
        pyperclip.copy(msg)
        pyautogui.hotkey("ctrl", "v")
        time.sleep(DELAY_TIME)
        pyautogui.press("enter")
        print("发送微信消息")
        im = pyautogui.screenshot()
        im.save('screenshot.png')
    except Exception as ex:
        print("发送微信消息出现异常: " + str(ex))
        sys.exit(0)


# 识别图片加入换行
def ocr_img(language, img_path, result_img):
    ocr = PaddleOCR(use_angle_cls=True, use_gpu=False, lang=language)
    img_path = img_path
    result = ocr.ocr(img_path, cls=True)
    worlds = "\n".join(line[-1][0] for line in result)
    image = Image.open(img_path).convert('RGB')
    boxes = [line[0] for line in result]
    txts = [line[1][0] for line in result]
    scores = [line[1][1] for line in result]
    im_show = draw_ocr(image, boxes, txts, scores)
    im_show = Image.fromarray(im_show)
    im_show.save(result_img)
    return worlds

#输出帮助
def get_help(words):
    poem = HELP
    return poem


#淨空老法師影音圖文檢索中心
def get_fa(question):
    if question == "fo":
        return ""
    print('get_fa', question)
    if '问题或关键字' in question:
        return ""
    options = webdriver.ChromeOptions()
    options.add_argument('--no-gpu')
    options.add_argument('--disable-dev-shm-usage')
    options.add_argument('--headless')
    chrome_service = Service(ChromeDriverManager().install())
    # Use the Service object when creating the WebDriver
    driver = webdriver.Chrome(service=chrome_service, options=options)
    driver.get('https://ft.amtb.cn')
    # Find the input element with ID "query" on the page
    input_element = driver.find_element(by=webdriver.common.by.By.ID, value="query")
    input_element.send_keys(question)
    # Find the search button with ID "searchButton" on the page
    search_button = driver.find_element(by=webdriver.common.by.By.ID, value="searchButton")
    search_button.click()
    sleep(10)
    # Wait for the results to load
    # Get the updated page source after the search
    page_source = driver.page_source
    soup = BeautifulSoup(page_source, 'lxml')
    try:
        article = soup.find('article')
        text = article.get_text()
        # Use regex to extract Chinese characters
        chinese_characters = re.sub('<[^<]+?>', '', text)
        driver.quit()
        return chinese_characters
    except:
        return "无"

# 截图 用用户名进行区分
def cut_screen(name):
    pyperclip.copy("")
    pyperclip.copy(name)
    pyautogui.hotkey("ctrl", "f")
    time.sleep(2)
    pyautogui.hotkey("ctrl", "v")
    pyautogui.press("enter")
    time.sleep(2)
    pyperclip.copy("")
    time.sleep(2)
    screenshot = pyautogui.screenshot(region=(259, 64, 703, 637))
    screenshot.save(f"./pic/{name}screenshot_region.png")
# Process OCR results for different cases
def process_ocr_results(name, text):
    process_ocr_results_by_keyword(name, text, 'help', '操作说明', get_help)
    print('操作说明 finished')
    sleep(2)
    process_ocr_results_by_keyword(name, text, 'fo', '淨空老法師中心', get_fa)
    print('淨空老法師中心 finished')
    sleep(2)

def remove_prefix(s, prefix):
    if s.startswith(prefix):
        return s[len(prefix):]
    return s


# Process OCR results for a specific keyword
def process_ocr_results_by_keyword(name, text, keyword, txtform, service_func):
    shiname = f"{name}_{keyword}"

    if text!="" and keyword=="help" :
        pattern = re.compile(rf'{keyword}')
        matches = [i for i in pattern.findall(text) if i]
    else:
        pattern = re.compile(rf'{keyword}(.*)')
        matches = [i for i in pattern.findall(text) if i]
    if matches:
        result = matches[-1].strip()
        if result != '':
            last_result = last_question.get(shiname, "")
            if last_result != result:
                last_question[shiname] = result
                if '问题或关键字' not in result:
                    return_world: string = service_func(result)
                    if return_world != "":
                        send_wx_message(name, return_world + f"\n{txtform}")
                        time.sleep(DELAY_TIME)


if __name__ == '__main__':
    last_question = {}
    while True:
        for name in SEND_NAMES:
            print(name)
            cut_screen(name)
            language = 'ch'
            img_path = f"./pic/{name}screenshot_region.png"
            result_img = f"./pic/{name}screenshot_region_out.png"
            text = ocr_img(language, img_path, result_img)
            process_ocr_results(name, text)
        time.sleep(5)