Merge pull request #354 from davidteather/nightly

Update Sound & Profile Scraping
davidteather · Nov 14, 2020 · a385538 · a385538
2 parents 1ea788d + 9e696d1
commit a385538
Show file tree

Hide file tree

Showing 2 changed files with 55 additions and 22 deletions.
diff --git a/TikTokApi/tiktok.py b/TikTokApi/tiktok.py
@@ -2,6 +2,7 @@
 import requests
 import time
 import logging
+import json
 from urllib.parse import urlencode, quote
 from .browser import browser, get_playwright
 from playwright import sync_playwright
@@ -154,7 +155,7 @@ def getData(self, **kwargs) -> dict:
                 "sec-fetch-mode": "cors",
                 "sec-fetch-site": "same-site",
                 "user-agent": userAgent,
-                "cookie": "tt_webid_v2=" + did,
+                "cookie": "tt_webid_v2=" + did + ';s_v_web_id=' + kwargs.get("custom_verifyFp", "verify_khgp4f49_V12d4mRX_MdCO_4Wzt_Ar0k_z4RCQC9pUDpX"),
             },
             proxies=self.__format_proxy(proxy),
         )
@@ -685,6 +686,16 @@ def bySound(self, id, count=30, offset=0, **kwargs) -> dict:
     def getMusicObject(self, id, **kwargs) -> dict:
         """Returns a music object for a specific sound id.
 
+        :param id: The sound id to search by.
+        :param language: The 2 letter code of the language to return.
+                         Note: Doesn't seem to have an affect.
+        :param proxy: The IP address of a proxy to make requests from.
+        """
+        return self.getMusicObjectFull(id, **kwargs)['music']
+
+    def getMusicObjectFull(self, id, **kwargs):
+        """Returns a music object for a specific sound id.
+
         :param id: The sound id to search by.
         :param language: The 2 letter code of the language to return.
                          Note: Doesn't seem to have an affect.
@@ -697,14 +708,18 @@ def getMusicObject(self, id, **kwargs) -> dict:
             maxCount,
             did,
         ) = self.__process_kwargs__(kwargs)
-        kwargs['custom_did'] = did
-
-        query = {"musicId": id, "language": language}
-        api_url = "{}node/share/music/{}?{}&{}".format(
-            BASE_URL, self.get_music_title(id, **kwargs) + "-" + str(id), self.__add_new_params__(), urlencode(query)
-        )
-
-        return self.getData(url=api_url, **kwargs)
+        r = requests.get("https://www.tiktok.com/music/-{}".format(id), headers={
+                "Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3;q=0.9",
+                "authority": "www.tiktok.com",
+                "Accept-Encoding": "gzip, deflate",
+                "Connection": "keep-alive",
+                "Host": "www.tiktok.com",
+                "User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/86.0.4240.111 Safari/537.36",
+                "Cookie": "s_v_web_id=" + kwargs.get("custom_verifyFp", "verify_khgp4f49_V12d4mRX_MdCO_4Wzt_Ar0k_z4RCQC9pUDpX"),
+            }, proxies=self.__format_proxy(kwargs.get("proxy", None)))
+        t = r.text
+        j_raw = t.split('<script id="__NEXT_DATA__" type="application/json" crossorigin="anonymous">')[1].split("</script>")[0]
+        return json.loads(j_raw)['props']['pageProps']['musicInfo']
 
     def byHashtag(self, hashtag, count=30, offset=0, **kwargs) -> dict:
         """Returns a dictionary listing TikToks with a specific hashtag.
@@ -999,14 +1014,22 @@ def getUser(self, username, **kwargs) -> dict:
             maxCount,
             did,
         ) = self.__process_kwargs__(kwargs)
-        kwargs['custom_did'] = did
-        secUid = self.get_secUid(username)
-        query = {"uniqueId": username, "language": language, "isUniqueId": True, "validUniqueId": username, "sec_uid": "", "secUid": secUid}
-        api_url = "{}node/share/user/@{}?{}&{}".format(
-            BASE_URL, quote(username), self.__add_new_params__(), urlencode(query)
-        )
+        r = requests.get("https://tiktok.com/@{}?lang=en".format(username), headers={
+            "Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3;q=0.9",
+            "authority": "www.tiktok.com",
+            "path": "/{}".format(username),
+            "Accept-Encoding": "gzip, deflate",
+            "Connection": "keep-alive",
+            "Host": "www.tiktok.com",
+            "User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/86.0.4240.111 Safari/537.36",
+            "Cookie": "s_v_web_id=" + kwargs.get("custom_verifyFp", "verify_khgp4f49_V12d4mRX_MdCO_4Wzt_Ar0k_z4RCQC9pUDpX"),
+        }, proxies=self.__format_proxy(kwargs.get("proxy", None)))
 
-        return self.getData(url=api_url, **kwargs)
+        t = r.text
+
+        j_raw = t.split('<script id="__NEXT_DATA__" type="application/json" crossorigin="anonymous">')[1].split("</script>")[0]
+
+        return json.loads(j_raw)['props']['pageProps']
 
     def getSuggestedUsersbyID(
         self, userId="6745191554350760966", count=30, **kwargs
@@ -1334,14 +1357,24 @@ def get_Video_No_Watermark(self, video_url, return_bytes=0, **kwargs) -> bytes:
                 return r.content
 
     def get_music_title(self, id, **kwargs):
-        r = requests.get("https://www.tiktok.com/music/-{}".format(id), proxies=self.__format_proxy(kwargs.get("proxy", None)))
-        text = r.text.split('TikTok","desc":')[0]
-        on_tiktok = text.split(" | ")
-        return on_tiktok[len(on_tiktok)-2].split(" ")[1]
+        r = requests.get("https://www.tiktok.com/music/-{}".format(id), headers={
+            "Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3;q=0.9",
+            "authority": "www.tiktok.com",
+            "Accept-Encoding": "gzip, deflate",
+            "Connection": "keep-alive",
+            "Host": "www.tiktok.com",
+            "User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/86.0.4240.111 Safari/537.36",
+            "Cookie": "s_v_web_id=" + kwargs.get("custom_verifyFp", "verify_khgp4f49_V12d4mRX_MdCO_4Wzt_Ar0k_z4RCQC9pUDpX"),
+        }, proxies=self.__format_proxy(kwargs.get("proxy", None)))
+        t = r.text
+        j_raw = t.split('<script id="__NEXT_DATA__" type="application/json" crossorigin="anonymous">')[1].split("</script>")[0]
+        return json.loads(j_raw)['props']['pageProps']['musicInfo']['title']
 
     def get_secUid(self, username, **kwargs):
         r = requests.get("https://tiktok.com/@{}?lang=en".format(username), headers={
-            "Accept": "*/*",
+            "Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3;q=0.9",
+            "authority": "www.tiktok.com",
+            "path": "/{}".format(username),
             "Accept-Encoding": "gzip, deflate",
             "Connection": "keep-alive",
             "Host": "www.tiktok.com",

diff --git a/setup.py b/setup.py
@@ -8,7 +8,7 @@
 setuptools.setup(
     name="TikTokApi",
     packages=["TikTokApi"],
-    version="3.7.8",
+    version="3.7.9",
     license="MIT",
     description="The Unofficial TikTok API Wrapper in Python 3.",
     author="David Teather",