article_generate/test.py

import time
import os
import urllib
from selenium.webdriver.common.by import By
from pyvirtualdisplay import Display
import sys
import urlToarticle
from fastapi import FastAPI, Form, Request
from fastapi.responses import FileResponse, HTMLResponse
from fastapi.templating import Jinja2Templates
from pydantic import BaseModel
from typing import List
from pyvirtualdisplay import Display
from selenium.webdriver.chrome.service import Service
import undetected_chromedriver as uc

app = FastAPI()
templates = Jinja2Templates(directory="templates")
driver = None

driver_path = '/usr/local/bin/chromedriver'
user_data_dir = '/home/ling/.config/google-chrome'
profile_directory = 'Default'


def re_get_webdriver():
    global port
    global driver
    global portnum
    global is_docker
    result = []
    if driver is not None:
        print('closing....')
        driver.quit()
        print('quit....')
        driver = None
    try:
        display = Display(visible=0, size=(1920, 1080))
        display.start()

        service = Service(driver_path)
        options = uc.ChromeOptions()
        #options.add_argument("--window-size=200,100")  # 縮小視窗
        #options.add_argument("--window-position=-32000,-32000")  # 移到螢幕外
        options.add_argument("--no-sandbox")
        options.add_argument("--disable-dev-shm-usage")
        options.add_argument("--disable-blink-features=AutomationControlled")

        driver = uc.Chrome(options=options, version_main=132)

        driver.execute_script("Object.defineProperty(navigator, 'webdriver', {get: () => undefined})")
        time.sleep(3)
        print(driver.current_url)
    except:
        driver = None
        return driver

    return driver


def get_resource(kw):
    max_attempts = 2
    attempts = 0
    while attempts<max_attempts:
        driver = re_get_webdriver()
        print('re_get_webdriver')
        if driver is not None:
            break
        time.sleep(3)
        attempts+=1
    try:
        googleurl = 'https://www.google.com/search?q={}&num={}&hl={}&gl=tw'.format(urllib.parse.quote(kw), 100, 'zh-TW')
        # googleurl = 'https://www.google.com/search?q={}&num={}&hl={}&gl=tw&tbm=vid&tbs=vd:m'.format(urllib.parse.quote(kw), 100, 'zh-TW')
        # googleurl = f'https://www.google.co.jp/search?q={kw}&sca_esv=741dc4f98c90c9c4&source=hp&ei=djmOZ8inMYWk2roPk_yMiA4&iflsig=AL9hbdgAAAAAZ45HhiuBAUgi3Vf3Qd5FTyfcyUOySOxk&ved=0ahUKEwjIutTinoSLAxUFklYBHRM-A-EQ4dUDCA8&uact=5&oq=junho&gs_lp=Egdnd3Mtd2l6IgphbmdlbG8ga29vMgUQLhiABDIEEAAYHjIEEAAYHjIEEAAYHjIEEAAYHjIEEAAYHjIEEAAYHjIEEAAYHjIGEAAYChgeSL0YUABYqRZwAXgAkAEAmAGwAaABjQyqAQQwLjExuAEDyAEA-AEBmAIMoALYDMICCxAuGIAEGNEDGMcBwgIFEAAYgATCAgoQLhiABBhDGIoFwgILEC4YgAQYxwEYrwHCAgcQABiABBgKwgIHEC4YgAQYCsICDRAuGIAEGMcBGAoYrwGYAwCSBwQxLjExoAfBqQE&sclient=gws-wiz'
        driver.get(googleurl)

        time.sleep(6)
        print(driver.current_url)
        elmts = driver.find_elements(By.XPATH, "//div[@class='yuRUbf']//a")
        numresults = len(elmts)
        print('搜尋結果數量', numresults)
        if numresults == 0:
            print(driver.current_url)
            print(driver.title)
            sys.exit()

        resources_list = []

        for elmt in elmts[0:11]:
            href = elmt.get_attribute('href')
            resources_list.append(href)
        print(resources_list)
        return resources_list

    except Exception as e:
        print('exception')
        return None

    driver.quit()

urls = get_resource('書房 設計 北歐') # 取得搜尋結果第一頁網址來源
#urls = ['https://www.100.com.tw/article/4359', 'https://yes-99.com/news-info.asp?id=278']
#article_generator = urlToarticle.ArticleGenerator(user_data_dir, profile_directory)
#is_succesed = article_generator.generate_article(urls)
#if is_succesed:
#    print(is_succesed)