Bulk download images from specific site URLs with python

# -*- coding: utf-8 -*- 

import urllib
import urllib2
import os.path
import sys
from HTMLParser import HTMLParser

def download(url):
    img = urllib.urlopen(url)
    localfile = open(os.path.basename(url),'wb')
    localfile.write(img.read())
    img.close()
    localfile.close()

class imgParser(HTMLParser):

    def __init__(self):
        HTMLParser.__init__(self)

    def handle_starttag(self,tagname,attribute):
        if tagname.lower() == "img":
            for i in attribute:
                if i[0].lower() == "src":
                    img_url=i[1]
                    #Creating a file that collects the URLs of the acquired photos
                    f = open("collection_url.txt","a")
                    f.write("%s\t"%img_url)
                    f.close()
        
if __name__ == "__main__":

    print('Enter the URL of the site where you want to get the photo.')
    input_url = raw_input('>>>  ')
    serch_url = input_url
    htmldata = urllib2.urlopen(serch_url)
    
    print('Currently getting image files...')

    parser = imgParser()
    parser.feed(htmldata.read())

    parser.close()
    htmldata.close()

    #Read the generated file
    f = open("collection_url.txt","r")
    for row in f:
        row_url = row.split('\t')
        len_url = len(row_url)
    f.close()

    number_url = []

    for i in range(0,(len_url-1)):
        number_url.append(row_url[i])

    for j in range(0,(len_url-1)):
        url = number_url[j]
        download(url)

    print('The image download is complete.')

    #Delete file
    os.remove("collection_url.txt")

Recommended Posts