From ebfe58e6cbc866c5e2dc4876a6b5868837d504b3 Mon Sep 17 00:00:00 2001 From: James Taylor Date: Fri, 31 Aug 2018 16:23:19 -0700 Subject: Ability to reply to comments --- youtube/account_functions.py | 135 +++++++++++++++++++++++++++++++++++++++++++ youtube/comments.py | 7 ++- youtube/watch.py | 4 ++ youtube/youtube.py | 2 +- 4 files changed, 146 insertions(+), 2 deletions(-) create mode 100644 youtube/account_functions.py (limited to 'youtube') diff --git a/youtube/account_functions.py b/youtube/account_functions.py new file mode 100644 index 0000000..9dd3d10 --- /dev/null +++ b/youtube/account_functions.py @@ -0,0 +1,135 @@ +# Contains functions having to do with logging in or requiring that one is logged in +import urllib +import json +from youtube import common, proto, comments +import re +def _post_comment(text, video_id, session_token, cookie): + headers = { + 'User-Agent': 'Mozilla/5.0 (iPhone; CPU iPhone OS 10_3_1 like Mac OS X) AppleWebKit/603.1.30 (KHTML, like Gecko) Version/10.0 Mobile/14E304 Safari/602.1', + 'Accept': '*/*', + 'Accept-Language': 'en-US,en;q=0.5', + 'Accept-Encoding': 'gzip, deflate, br', + 'X-YouTube-Client-Name': '2', + 'X-YouTube-Client-Version': '2.20180823', + 'Content-Type': 'application/x-www-form-urlencoded', + 'Cookie': cookie, + } + + comment_params = proto.string(2, video_id) + proto.nested(5, proto.uint(1, 0)) + proto.uint(10, 1) + comment_params = proto.percent_b64encode(comment_params).decode('ascii') + + sej = json.dumps({"clickTrackingParams":"AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA=", "commandMetadata":{"webCommandMetadata":{"url":"/service_ajax","sendPost":True}},"createCommentEndpoint":{"createCommentParams": comment_params}}) + + data_dict = { + 'comment_text': text, + 'sej': sej, + 'session_token': session_token, + } + data = urllib.parse.urlencode(data_dict).encode() + + req = urllib.request.Request("https://m.youtube.com/service_ajax?name=createCommentEndpoint", headers=headers, data=data) + response = urllib.request.urlopen(req, timeout = 5) + '''with open('debug/post_comment_response', 'wb') as f: + f.write(response.read())''' + + +def _post_comment_reply(text, video_id, parent_comment_id, session_token, cookie): + headers = { + 'User-Agent': 'Mozilla/5.0 (iPhone; CPU iPhone OS 10_3_1 like Mac OS X) AppleWebKit/603.1.30 (KHTML, like Gecko) Version/10.0 Mobile/14E304 Safari/602.1', + 'Accept': '*/*', + 'Accept-Language': 'en-US,en;q=0.5', + 'Accept-Encoding': 'gzip, deflate, br', + 'X-YouTube-Client-Name': '2', + 'X-YouTube-Client-Version': '2.20180823', + 'Content-Type': 'application/x-www-form-urlencoded', + 'Cookie': cookie, + } + + comment_params = proto.string(2, video_id) + proto.string(4, parent_comment_id) + proto.nested(5, proto.uint(1, 0)) + proto.uint(6,0) + proto.uint(10, 1) + comment_params = proto.percent_b64encode(comment_params).decode('ascii') + + sej = json.dumps({"clickTrackingParams":"AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA=", "commandMetadata":{"webCommandMetadata":{"url":"/service_ajax","sendPost":True}},"createCommentReplyEndpoint":{"createReplyParams": comment_params}}) + + data_dict = { + 'comment_text': text, + 'sej': sej, + 'session_token': session_token, + } + data = urllib.parse.urlencode(data_dict).encode() + + req = urllib.request.Request("https://m.youtube.com/service_ajax?name=createCommentReplyEndpoint", headers=headers, data=data) + response = urllib.request.urlopen(req, timeout = 5) + '''with open('debug/post_comment_response', 'wb') as f: + f.write(response.read())''' + + + +xsrf_token_regex = re.compile(r'''XSRF_TOKEN"\s*:\s*"([\w-]*(?:=|%3D){0,2})"''') +def post_comment(query_string, fields): + with open('data/cookie.txt', 'r', encoding='utf-8') as f: + cookie_data = f.read() + + parameters = urllib.parse.parse_qs(query_string) + try: + video_id = fields['video_id'][0] + except KeyError: + video_id = parameters['video_id'][0] + + # Get session token for mobile + # youtube-dl uses disable_polymer=1 which uses a different request format which has an obfuscated javascript algorithm to generate a parameter called "bgr" + # Tokens retrieved from disable_polymer pages only work with that format. Tokens retrieved on mobile only work using mobile requests + # Additionally, tokens retrieved without sending the same cookie won't work. So this is necessary even if the bgr and stuff was reverse engineered. + headers = {'User-Agent': common.mobile_user_agent, + 'Cookie': cookie_data,} + mobile_page = common.fetch_url('https://m.youtube.com/watch?v=' + video_id, headers, report_text="Retrieved session token for comment").decode() + match = xsrf_token_regex.search(mobile_page) + if match: + token = match.group(1).replace("%3D", "=") + else: + raise Exception("Couldn't find xsrf_token") + + if 'parent_id' in parameters: + _post_comment_reply(fields['comment_text'][0], parameters['video_id'][0], parameters['parent_id'][0], token, cookie_data) + return comments.get_comments_page(query_string) + else: + _post_comment(fields['comment_text'][0], fields['video_id'][0], token, cookie_data) + return common.yt_basic_template.substitute( + page_title = "Success", + style = '', + header = common.get_header(), + page = 'Comment sucessfully posted', + ) + + +def get_post_comment_page(query_string): + parameters = urllib.parse.parse_qs(query_string) + video_id = parameters['v'][0] + style = ''' main{ + display: grid; + grid-template-columns: 3fr 2fr; +} +.left{ + display:grid; + grid-template-columns: 1fr 640px; +} +textarea{ + width: 462px; + height: 85px; +} +.comment-form{ + grid-column:2; +}''' + page = '''
+
+ + + +
+
+''' + return common.yt_basic_template.substitute( + page_title = "Post a comment", + style = style, + header = common.get_header(), + page = page, + ) \ No newline at end of file diff --git a/youtube/comments.py b/youtube/comments.py index 5490938..a1f08bf 100644 --- a/youtube/comments.py +++ b/youtube/comments.py @@ -206,7 +206,6 @@ def parse_comments_polymer(content, replies=False): print('Error parsing comments: ' + str(e)) comments = () ctoken = '' - raise else: print("Finished getting and parsing comments") return {'ctoken': ctoken, 'comments': comments, 'video_title': video_title} @@ -273,6 +272,10 @@ def get_comments_page(query_string): if replies: page_title = 'Replies' video_metadata = '' + comment_box = '''
+ + +
''' else: page_number = str(int(metadata['offset']/20) + 1) page_title = 'Comments page ' + page_number @@ -283,6 +286,7 @@ def get_comments_page(query_string): url = common.URL_ORIGIN + '/watch?v=' + metadata['video_id'], thumbnail = '/i.ytimg.com/vi/'+ metadata['video_id'] + '/mqdefault.jpg', ) + comment_box = '' comments_html, ctoken = get_comments_html(parsed_comments) @@ -293,6 +297,7 @@ def get_comments_page(query_string): return yt_comments_template.substitute( header = common.get_header(), + comment_box = comment_box, video_metadata = video_metadata, comments = comments_html, page_title = page_title, diff --git a/youtube/watch.py b/youtube/watch.py index ad3c448..ff7d6df 100644 --- a/youtube/watch.py +++ b/youtube/watch.py @@ -385,6 +385,10 @@ def get_watch_page(query_string): music_list_html += '''\n''' music_list_html += '''\n''' + with open('data/googlevideo-domains.txt', 'a+', encoding='utf-8') as f: + url = info['formats'][0]['url'] + subdomain = url[0:url.find(".googlevideo.com")] + f.write(subdomain + "\n") download_options = '' for format in info['formats']: diff --git a/youtube/youtube.py b/youtube/youtube.py index 3c31fa4..29fa6a0 100644 --- a/youtube/youtube.py +++ b/youtube/youtube.py @@ -80,7 +80,7 @@ def youtube(env, start_response): else: start_response('400 Bad Request', ()) return b'400 Bad Request' - elif path == "/post_comment": + elif path in ("/post_comment", "/comments"): start_response('200 OK', () ) return account_functions.post_comment(query_string, fields).encode() -- cgit v1.2.3