인스타그램 effect crawler 버그 수정

WebBasedCrawler/base/baseclasses.py

@@ -23,7 +23,6 @@ from selenium.webdriver.support import expected_conditions as EC
 from selenium.webdriver.support.ui import WebDriverWait
 from selenium.webdriver.common.desired_capabilities import DesiredCapabilities
 
-is_debug = False
 
 def is_debugger_attached():
     for frame in inspect.stack():
@@ -31,6 +30,8 @@ def is_debugger_attached():
             return True
     return False
 
+is_debug = is_debugger_attached()
+
 def printl(*objects, sep=' ', end='\n', file=None, flush=True):
     if is_debug:
         cur_frame = inspect.currentframe()

WebBasedCrawler/effect/effectinstagram.py

@@ -5,6 +5,7 @@ import json
 import requests
 import requests.exceptions
 import time
+import bs4
 
 import insta.instaheaders as instaheaders
 import insta.instaparser as instaparser
@@ -73,18 +74,20 @@ class InstaContent:
         self.has_previous = False
         self.cookies = {}
         self.proxies = proxies
+        self.query_id = ''
+        self.content = ''
         self.load_url(url, cookies, referer, self.proxies)
 
     def load_url(self, url, cookies, referer, proxies):
         self.__set_cookies(cookies)
         self.__r = requests.get(url, headers=instaheaders.get_headers_for_body_html(self.cookies), proxies=proxies,
                                 timeout=requests_timeout, stream=True)
-        content = requests_get(self.__r)
+        self.content = requests_get(self.__r)
         self.__r.raise_for_status()
         self.__referer = referer
         self.__code = self.__get_code(url)
         # self.body, self.reply, self.start_cursor, self.has_previous = instaparser.parse_body_html(self.__r.content)
-        self.body, self.reply, self.start_cursor, self.has_previous = instaparser.parse_body_html(content)
+        self.body, self.reply, self.start_cursor, self.has_previous = instaparser.parse_body_html(self.content)
         self.__set_cookies(self.__r.cookies)
         self.__r.close()
         return self.body, self.reply
@@ -95,20 +98,52 @@ class InstaContent:
     def get_reply(self):
         return self.reply
 
+    def get_query_ids(self, html):
+        doc = bs4.BeautifulSoup(html, "html.parser")
+
+        query_ids = []
+        for script in doc.find_all("script"):
+            if script.has_attr("src") and "_Commons.js" in script['src']:
+                text = requests.get("%s%s" % ('https://www.instagram.com', script['src'])).text
+                for query_id in re.findall("(?<=queryId:\")[0-9]{17,17}", text):
+                    query_ids.append(query_id)
+        return query_ids
+
+    def find_query_id(self):
+        potential_query_ids = self.get_query_ids(self.content)
+        query_id = ''
+        for potential_id in potential_query_ids:
+            url = 'https://www.instagram.com/graphql/query/?query_id={}&shortcode={}&first={}&after={}'.format(
+                potential_id, self.__code, len(self.reply), self.start_cursor)
+            try:
+                data = requests.get(url).json()
+                if data['status'] == 'ok':
+                    query_id = potential_id
+                    break
+            except Exception:
+                # no valid JSON retured, most likely wrong query_id resulting in 'Oops, an error occurred.'
+                pass
+
+        return query_id
+
     def load_reply_more(self):
-        form_data = instaheaders.get_form_data_for_reply(self.__code, self.start_cursor, num_of_reply_ajax)
-        headers = instaheaders.get_headers_for_ajax(self.cookies, self.__referer, form_data)
-        self.log_load_reply_more_before(form_data, headers)
-        self.__r = requests.post(insta_query, headers=headers, data=form_data, proxies=self.proxies,
-                                 timeout=requests_timeout, stream=True)
+        if not self.query_id:
+            self.query_id = self.find_query_id()
+
+        url = 'https://www.instagram.com/graphql/query/?query_id={}&shortcode={}&first={}&after={}'.format(
+            self.query_id, self.__code, len(self.reply), self.start_cursor)
+        self.__r = requests.get(url, headers=instaheaders.get_headers_for_body_html(self.cookies), proxies=self.proxies,
+                                timeout=requests_timeout, stream=True)
         content = requests_get(self.__r)
         self.__r.raise_for_status()
+        reply, self.start_cursor, self.has_previous = instaparser.parse_reply_more(content)
         self.__set_cookies(self.__r.cookies)
-        # self.reply, self.start_cursor, self.has_previous = instaparser.parse_reply_ajax(self.__r.content)
-        self.reply, self.start_cursor, self.has_previous = instaparser.parse_reply_ajax(content)
         self.__r.close()
-        self.log_load_reply_more_after()
-        return self.reply
+
+        self.reply = self.reply+reply
+        printl('{} - reply : {} (next : {})'.format(self.__referer, len(self.reply), self.start_cursor))
+
+        return reply
 
     def get_cookies(self):
         return self.cookies
@@ -332,13 +367,19 @@ class EffectInsta(object):
 
     def put_today_buzz(self, buzzs, today_acc_buzz_count):
         today = datetime.date.today().strftime('%Y%m%d')
-        today_buzz_count = today_acc_buzz_count - buzzs[-2][BUZZ_KEY[ACC]]
-        # if today_buzz_count < 0:
-        #     today_buzz_count = 0
 
         result_buzzs = buzzs.copy()
-        result_buzzs[-1][BUZZ_KEY[DAY]] = today_buzz_count if today_buzz_count >=0 else 0
-        result_buzzs[-1][BUZZ_KEY[ACC]] = result_buzzs[-2][BUZZ_KEY[ACC]] + today_buzz_count
+        if len(result_buzzs) == 0:
+            result_buzzs.append({BUZZ_KEY[ACC]:today_acc_buzz_count, BUZZ_KEY[DAY]:today_acc_buzz_count, BUZZ_KEY[DATE]:today})
+        elif len(result_buzzs) == 1:
+            result_buzzs[-1][BUZZ_KEY[ACC]] = today_acc_buzz_count
+            result_buzzs[-1][BUZZ_KEY[DAY]] = today_acc_buzz_count
+        else:
+            result_buzzs[-1][BUZZ_KEY[ACC]] = today_acc_buzz_count
+            result_buzzs[-1][BUZZ_KEY[DAY]] = today_acc_buzz_count - result_buzzs[-2][BUZZ_KEY[ACC]]
+
+        if result_buzzs[-1][BUZZ_KEY[DAY]] < 0:
+            result_buzzs[-1][BUZZ_KEY[DAY]] = 0
 
         return result_buzzs
 

WebBasedCrawler/effect/resultsender.py

@@ -43,8 +43,11 @@ class ResultSender:
         if not self.conn.open:
             self.connect()
         with self.conn.cursor() as cursor:
-            cursor.execute(query)
-            buzz = cursor.fetchone()
+            try:
+                cursor.execute(query)
+                buzz = cursor.fetchone()
+            except Exception as e:
+                print(e)
 
         return buzz['replybuzz'] if buzz != None else buzz
 

WebBasedCrawler/insta/instacrawl.py

@@ -731,9 +731,9 @@ class InstaContent:
         self.__set_cookies(self.__r.cookies)
         self.__r.close()
 
-        self.reply += reply
+        self.reply = self.reply + reply
         printl('{} - reply : {} (next : {})'.format(self.__referer, len(self.reply), self.start_cursor))
-        return self.reply
+        return reply
 
     def get_cookies(self):
         return self.cookies