파이썬(4)-pixabay

*이 글을 읽기전에 작성자 개인의견이 있으니, 다른 블로그와 교차로 읽는것을 권장합니다.*

1. 특정 키워드를 검색해서 화면이 나오면 이미지 선택하기

!pip install selenium
!pip install chromedriver_autoinstaller

 

import chromedriver_autoinstaller
import time
from selenium import webdriver
from urllib.request import Request, urlopen

driver = webdriver.Chrome()
def search(word):
    url = 'https://pixabay.com/ko/images/search/' + word
    driver.get(url)

search('winter')

image_xpath = '/html/body/div[1]/div[1]/div/div[2]/div[3]/div/div/div/div[23]/div/a/img'
image_url = driver.find_element('xpath', image_xpath).get_attribute('src')
print('image_url: ', image_url)

image_byte = Request(image_url, headers={'User-Agent':'Mozilla/5.0 (Windows NT 10.0; Win64; x64)'})
f = open('winter.jpg','wb') # writebyte 모드로 만듦
f.write(urlopen(image_byte).read())  # 이게 들어있는 파일에 이미지가 저장
f.close()

픽사베이 한 화면에 있는 사진 모두 다운하기 for문, time 사용

url = f'https://pixabay.com/ko/images/search/고양이'
driver = webdriver.Chrome()
driver.get(url)

image_xpath = '/html/body/div[1]/div[1]/div/div[2]/div[3]/div/div'
image_url = driver.find_element('xpath', image_xpath).get_attribute('src')
print(image_url.find_elements)

---

2. 여러개 이미지 수집하기

import chromedriver_autoinstaller
import time
from selenium import webdriver
from urllib.request import Request, urlopen
from selenium.webdriver.common.by import By

driver = webdriver.Chrome()
word = 'winter'
url = 'https://pixabay.com/ko/images/search/' + word
driver.get(url)
driver.implicitly_wait(2)  # (페이지 다부르기위한 용도)3초 기다리고 다음으로

# for _ in range(20):
#     driver.execute_script("window.scrollBy({ top: window.innerHeight, behavior: 'smooth' })")
#     driver.execute_script(f'window.scrollTo(0, document.body.scrollHeight/({i}+1))')
#     time.sleep(0.5)
for i in range(20):
    driver.execute_script(f'window.scrollTo(0, 1000 * {i})')
    time.sleep(0.1)

image_area_xpath = '/html/body/div[1]/div[1]/div/div[2]/div[3]/div/div'
image_area = driver.find_element(By.XPATH, image_area_xpath)
image_elements = image_area.find_elements(By.TAG_NAME, 'img')

image_urls = []
for image_element in image_elements:
    image_url = image_element.get_attribute('src')
    print(image_url)
    image_urls.append(image_url)

import os
from urllib import parse

# 0번~99번까지 강아지사진 저장 완료
for i in range(len(image_urls)):
    image_url = image_urls[i]

    url = parse.urlparse(image_url)
    name, ext = os.path.splitext(url.path) #path='/photo/2017/09/25/13/12/puppy-2785074_1280.jpg', params='', query='', fragment='')
    image_byte = Request(image_url, headers= {'User-Agent':'Mozilla/5.0 (Windows NT 10.0; Win64; x64)'})
    f = open(f'dog{i}.{ext}', 'wb')
    f.write(urlopen(image_byte).read())
    f.close()

---

3. 함수로 리팩토링

• crawl_and_save_image(keyword, pages)
• 강아지 이름으로 폴더 만들고, keyword이름으로 사진 넣기

import chromedriver_autoinstaller
import time
from selenium import webdriver
from urllib.request import Request, urlopen
from selenium.webdriver.common.by import By
import os
from urllib import parse

def searching(word):
    driver = webdriver.Chrome()
    url = 'https://pixabay.com/ko/images/search/' + word
    driver.get(url)
    driver.implicitly_wait(2)

def crawl_and_save_image(keyword, pages):
    image_urls = []
    for i in range(1, pages+1):
        url = f'https://pixabay.com/ko/images/search/{keyword}/?pagi={i}'
        driver.get(url)
        driver.implicitly_wait(2)

        time.sleep(2)
        for i in range(20):
            driver.execute_script(f'window.scrollTo(0, document.body.scrollHeight / 20 * {i})')
            time.sleep(0.1)
        
        image_area_xpath = '/html/body/div[1]/div[1]/div/div[2]/div[3]/div/div'
        image_area = driver.find_element(By.XPATH, image_area_xpath)
        image_elements = image_area.find_elements(By.TAG_NAME, 'img')
        
        for image_element in image_elements:
            image_url = image_element.get_attribute('src')
            print(image_url)
            image_urls.append(image_url)

    if not os.path.exists(keyword):
        os.mkdir(keyword)

    print(len(image_urls))
    # for i in range(len(image_urls)):
    #     image_url = image_urls[i]
    #     filename = image_url.split('/')[-1]
    #     image_byte = Request(image_url, headers={'User-Agent':'Mozilla/5.0 (Windows NT 10.0; Win64; x64)'})
    #     f= open(f'{keyword}/{filename}', 'wb')
    #     f.write(urlopen(image_byte).read())
    #     f.close()

driver = webdriver.Chrome()
# 사진 폴더안에 저장
crawl_and_save_image('tiger',2)

f.close()

확인