增加七牛图片转换

requirements.txt

+git+ssh://git@git.wanmeizhensuo.com/backend/gm-upload.git@master
 Pillow >= 5.0.0
 matplotlib >= 2.1.2
 requests >= 2.18.4

settings/settings.py

+SECRET_KEY = 'h88p#*4t5l4&7^j4gfui%qs4x3wo(+u$1x2d1=p9cb)ogjr080'
+QINIU_ACCESS_KEY = "UPCOYIJkZOMcdd9FDzpBqYjzWUh55fBpVi3AhWpL"
+QINIU_SECRET_KEY = "z5YvpDDSam_JE345Z8J_f3TufzelOW2VOGNoBl9e"
+QINIU_HOST = "http://wanmeizhensuo.qiniudn.com/"
+QINIU_SCOPE = 'wanmeizhensuo'

zhihu_login.py

@@ -25,6 +25,14 @@ from scrapy.selector import Selector
 import brotli
 import random
 import time
+import re
+from gm_upload import upload, upload_file
+from urllib.request import urlretrieve
+import os
+import cv2
+import copy
+
+
 
 
 top_query_list = [
@@ -232,7 +240,7 @@ class ZhihuAccount(object):
             print(50*"*")
 
     # 知乎搜索词搜索
-    def zhihu_query_by_word(self,query_word,zhihu_spider_fd):
+    def zhihu_query_by_word(self,query_word,zhihu_spider_fd,cur_image_index):
 
         for begin_index in range(0,200,10):
             query_by_word_url = "https://www.zhihu.com/api/v4/search_v3?t=general&correction=1&lc_idx=62&" \
@@ -252,10 +260,27 @@ class ZhihuAccount(object):
                         try:
                             data_type = data_item["object"]["type"]
                             content = data_item["object"]["content"]
+                            # content = copy.deepcopy(tmp_content)
                             platform_id = data_item["object"]["id"]
                             user_id = random.choice(majia_user_list)
                             question_id = ""
 
+                            img_url_list = re.findall('img src="(.*?)"', content)
+                            for ori_img_url in img_url_list:
+                                cur_image_index += 1
+
+                                local_img_url_path = "./image/img_" + str(cur_image_index) + ".png"
+                                urlretrieve(ori_img_url, local_img_url_path)
+                                local_cv2_img = cv2.imread(local_img_url_path)
+                                height, weidth, channel = local_cv2_img.shape
+                                local_cropped_img = local_cv2_img[0:(height - 100), 0:weidth]
+                                local_cropped_img_url_path = "./image/cropped_image_" + str(cur_image_index) + ".png"
+                                cv2.imwrite(local_cropped_img_url_path,local_cropped_img)
+                                qiniu_url = upload_file(local_cropped_img_url_path)
+                                content = content.replace(ori_img_url,qiniu_url)
+
+
+
                             if data_type == "article":
                                 title = data_item["object"]["title"]
                             elif data_type == "answer":
@@ -289,12 +314,16 @@ class ZhihuAccount(object):
         return
 if __name__ == '__main__':
 
+    os.environ.setdefault("DJANGO_SETTINGS_MODULE", "settings.settings")
+
     account = ZhihuAccount('', '')
     account.login(captcha_lang='en', load_cookies=True)
     #account.test_member_article()
 
     zhihu_spider_data = "./zhihu_spider_data.txt"
     zhihu_spider_fd = open(zhihu_spider_data,"w")
-    account.zhihu_query_by_word(top_query_list[0],zhihu_spider_fd)
+
+    cur_image_index = 0
+    account.zhihu_query_by_word(top_query_list[0],zhihu_spider_fd,cur_image_index)
 
     zhihu_spider_fd.close()
\ No newline at end of file