From 441b620b870f9c27e05537251f630fa2a2427eb1 Mon Sep 17 00:00:00 2001 From: Dylan Date: Sun, 27 Apr 2025 17:30:07 +0100 Subject: [PATCH] Add support for ?withFeed in API user requests --- requirements.txt | 3 ++- test_api.py | 11 +++++++++-- test_vx_extract.py | 5 ++--- twExtract/__init__.py | 39 ++++++++++++++++++++++++++++----------- twExtract/twUtils.py | 23 +++++++++++++++++++++-- twitfix.py | 12 ++++++++++-- vxApi.py | 2 +- 7 files changed, 73 insertions(+), 22 deletions(-) diff --git a/requirements.txt b/requirements.txt index c2a7318..f365729 100644 --- a/requirements.txt +++ b/requirements.txt @@ -7,4 +7,5 @@ Flask-Cors==4.0.0 Werkzeug==2.3.7 numerize==0.12 oauthlib==3.2.2 -PyRTF3==0.47.5 \ No newline at end of file +PyRTF3==0.47.5 +XClientTransaction==0.0.2 \ No newline at end of file diff --git a/test_api.py b/test_api.py index 58e752a..9843eb4 100644 --- a/test_api.py +++ b/test_api.py @@ -38,7 +38,14 @@ def test_api_include_rtf_nomedia(): assert not any(".rtf" in i for i in jData["mediaURLs"]) def test_api_user(): - resp = client.get(testUser.replace("https://twitter.com","https://api.vxtwitter.com")+"?include_rtf=true",headers={"User-Agent":"test"}) + resp = client.get(testUser.replace("https://twitter.com","https://api.vxtwitter.com"),headers={"User-Agent":"test"}) jData = resp.get_json() assert resp.status_code==200 - assert jData["screen_name"]=="jack" \ No newline at end of file + assert jData["screen_name"]=="jack" + +def test_api_user_feed(): + resp = client.get(testUser.replace("https://twitter.com","https://api.vxtwitter.com")+"?withFeed=true",headers={"User-Agent":"test"}) + jData = resp.get_json() + assert resp.status_code==200 + assert jData["screen_name"]=="jack" + assert len(jData["latestTweets"])>0 \ No newline at end of file diff --git a/test_vx_extract.py b/test_vx_extract.py index 5ffccc7..18613e0 100644 --- a/test_vx_extract.py +++ b/test_vx_extract.py @@ -101,7 +101,6 @@ def test_twextract_pollTweetExtract(): # basic check if poll data exists def test_twextract_NSFW_TweetExtract(): tweet = twExtract.extractStatus(testNSFWTweet,workaroundTokens=tokens) # For now just test that there's no error -''' def test_twextract_feed(): - tweet = twExtract.extractUserFeedFromId(testUserID,workaroundTokens=tokens) -''' \ No newline at end of file + tweets = twExtract.extractUserFeedFromId(testUserID,workaroundTokens=tokens) # For now just test that there's no error + assert len(tweets)>0 \ No newline at end of file diff --git a/twExtract/__init__.py b/twExtract/__init__.py index f96372b..8505474 100644 --- a/twExtract/__init__.py +++ b/twExtract/__init__.py @@ -14,6 +14,8 @@ v2bearer="Bearer AAAAAAAAAAAAAAAAAAAAANRILgAAAAAAnNwIzUejRCOuH5E6I8xnZz4puTs%3D1 androidBearer="Bearer AAAAAAAAAAAAAAAAAAAAAFXzAwAAAAAAMHCxpeSDG1gLNLghVe8d74hl6k4%3DRUMF4xAQLsbeBhTSRrCiQpJtxoGWeyHrDb5te2jpGskWDFW82F" tweetdeckBearer="Bearer AAAAAAAAAAAAAAAAAAAAAFQODgEAAAAAVHTp76lzh3rFzcHbmHVvQxYYpTw%3DckAlMINMjmCwxUcaXbAN4XqJVdgMJaHqNOFgPMK0zN1qLqLQCF" +requestUserAgent = "Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:138.0) Gecko/20100101 Firefox/138.0" + bearerTokens=[tweetdeckBearer,bearer,v2bearer,androidBearer] guestToken=None @@ -36,8 +38,8 @@ tweetDetailGraphqlFeatures='{"rweb_tipjar_consumption_enabled":true,"responsive_ tweetDetailGraphql_api="e7RKseIxLu7HgkWNKZ6qnw" # this is for UserTweets endpoint -tweetFeedGraphqlFeatures='{"profile_label_improvements_pcf_label_in_post_enabled":true,"rweb_tipjar_consumption_enabled":true,"responsive_web_graphql_exclude_directive_enabled":true,"verified_phone_label_enabled":false,"creator_subscriptions_tweet_preview_api_enabled":true,"responsive_web_graphql_timeline_navigation_enabled":true,"responsive_web_graphql_skip_user_profile_image_extensions_enabled":false,"premium_content_api_read_enabled":false,"communities_web_enable_tweet_community_results_fetch":true,"c9s_tweet_anatomy_moderator_badge_enabled":true,"responsive_web_grok_analyze_button_fetch_trends_enabled":false,"responsive_web_grok_analyze_post_followups_enabled":true,"responsive_web_jetfuel_frame":false,"responsive_web_grok_share_attachment_enabled":true,"articles_preview_enabled":true,"responsive_web_edit_tweet_api_enabled":true,"graphql_is_translatable_rweb_tweet_is_translatable_enabled":true,"view_counts_everywhere_api_enabled":true,"longform_notetweets_consumption_enabled":true,"responsive_web_twitter_article_tweet_consumption_enabled":true,"tweet_awards_web_tipping_enabled":false,"responsive_web_grok_analysis_button_from_backend":true,"creator_subscriptions_quote_tweet_preview_enabled":false,"freedom_of_speech_not_reach_fetch_enabled":true,"standardized_nudges_misinfo":true,"tweet_with_visibility_results_prefer_gql_limited_actions_policy_enabled":true,"rweb_video_timestamps_enabled":true,"longform_notetweets_rich_text_read_enabled":true,"longform_notetweets_inline_media_enabled":true,"responsive_web_grok_image_annotation_enabled":true,"responsive_web_enhance_cards_enabled":false}' -tweetFeedGraphql_api="Y9WM4Id6UcGFE8Z-hbnixw" +tweetFeedGraphqlFeatures='{"rweb_video_screen_enabled":false,"profile_label_improvements_pcf_label_in_post_enabled":true,"rweb_tipjar_consumption_enabled":true,"verified_phone_label_enabled":false,"creator_subscriptions_tweet_preview_api_enabled":true,"responsive_web_graphql_timeline_navigation_enabled":true,"responsive_web_graphql_skip_user_profile_image_extensions_enabled":false,"premium_content_api_read_enabled":false,"communities_web_enable_tweet_community_results_fetch":true,"c9s_tweet_anatomy_moderator_badge_enabled":true,"responsive_web_grok_analyze_button_fetch_trends_enabled":false,"responsive_web_grok_analyze_post_followups_enabled":true,"responsive_web_jetfuel_frame":false,"responsive_web_grok_share_attachment_enabled":true,"articles_preview_enabled":true,"responsive_web_edit_tweet_api_enabled":true,"graphql_is_translatable_rweb_tweet_is_translatable_enabled":true,"view_counts_everywhere_api_enabled":true,"longform_notetweets_consumption_enabled":true,"responsive_web_twitter_article_tweet_consumption_enabled":true,"tweet_awards_web_tipping_enabled":false,"responsive_web_grok_show_grok_translated_post":false,"responsive_web_grok_analysis_button_from_backend":true,"creator_subscriptions_quote_tweet_preview_enabled":false,"freedom_of_speech_not_reach_fetch_enabled":true,"standardized_nudges_misinfo":true,"tweet_with_visibility_results_prefer_gql_limited_actions_policy_enabled":true,"longform_notetweets_rich_text_read_enabled":true,"longform_notetweets_inline_media_enabled":true,"responsive_web_grok_image_annotation_enabled":true,"responsive_web_enhance_cards_enabled":false}' +tweetFeedGraphql_api="OAx9yEcW3JA9bPo63pcYlA" twitterUrl = "x.com" # doubt this will change but just in case class TwExtractError(Exception): @@ -99,7 +101,7 @@ def twitterApiGet(url,btoken=None,authToken=None,guestToken=None): def getAuthHeaders(btoken,authToken=None,guestToken=None): csrfToken=str(uuid.uuid4()).replace('-', '') - headers = {"x-twitter-active-user":"yes","x-twitter-client-language":"en","x-csrf-token":csrfToken,"User-Agent":"Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:109.0) Gecko/20100101 Firefox/116.0"} + headers = {"x-twitter-active-user":"yes","x-twitter-client-language":"en","x-csrf-token":csrfToken,"User-Agent":requestUserAgent} headers['Authorization'] = btoken if authToken is not None: @@ -114,7 +116,7 @@ def getGuestToken(): global guestToken global guestTokenUses if guestToken is None: - r = requests.get(f"https://{twitterUrl}",headers={"User-Agent":"Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:109.0) Gecko/20100101 Firefox/116.0","Cookie":"night_mode=2"},allow_redirects=False) + r = requests.get(f"https://{twitterUrl}",headers={"User-Agent":requestUserAgent,"Cookie":"night_mode=2"},allow_redirects=False) m = re.search(gt_pattern, r.text) if m is None: r = requests.post(f"https://api.{twitterUrl}/1.1/guest/activate.json", headers={"Authorization":bearer}) @@ -510,25 +512,40 @@ def extractUserFeedFromId(userId,workaroundTokens): # TODO: https://api.twitter.com/graphql/x31u1gdnjcqtiVZFc1zWnQ/UserWithProfileTweetsQueryV2?variables={"cursor":"?","includeTweetImpression":true,"includeHasBirdwatchNotes":false,"includeEditPerspective":false,"includeEditControl":true,"count":40,"rest_id":"12","includeTweetVisibilityNudge":true,"autoplay_enabled":true}&features={"longform_notetweets_inline_media_enabled":true,"super_follow_badge_privacy_enabled":true,"longform_notetweets_rich_text_read_enabled":true,"super_follow_user_api_enabled":true,"unified_cards_ad_metadata_container_dynamic_card_content_query_enabled":true,"super_follow_tweet_api_enabled":true,"articles_api_enabled":true,"android_graphql_skip_api_media_color_palette":true,"creator_subscriptions_tweet_preview_api_enabled":true,"freedom_of_speech_not_reach_fetch_enabled":true,"tweetypie_unmention_optimization_enabled":true,"longform_notetweets_consumption_enabled":true,"subscriptions_verification_info_enabled":true,"blue_business_profile_image_shape_enabled":true,"tweet_with_visibility_results_prefer_gql_limited_actions_policy_enabled":true,"immersive_video_status_linkable_timestamps":false,"super_follow_exclusive_tweet_notifications_enabled":true} continue try: - vars = json.loads('{"userId":"x","count":20,"includePromotedContent":true,"withQuickPromoteEligibilityTweetFields":true,"withVoice":true,"withV2Timeline":true}') + vars = json.loads('{"userId":"0","count":20,"includePromotedContent":true,"withCommunity":true,"withVoice":true}') vars['userId'] = str(userId) vars['includePromotedContent'] = False # idk if this works reqHeaders = getAuthHeaders(bearer,authToken=authToken) - reqHeaders["x-client-transaction-id"] = twUtils.generate_transaction_id("GET","/i/api/graphql/x31u1gdnjcqtiVZFc1zWnQ/UserWithProfileTweetsQueryV2") - feed = requests.get(f"https://{twitterUrl}/i/api/graphql/{tweetFeedGraphql_api}/UserTweets?variables={urllib.parse.quote(json.dumps(vars))}&features={urllib.parse.quote(tweetFeedGraphqlFeatures)}", reqHeaders) + endpoint=f"/i/api/graphql/{tweetFeedGraphql_api}/UserTweetsAndReplies" + reqHeaders["x-client-transaction-id"] = twUtils.generate_transaction_id("GET",endpoint) + feed = requests.get(f"https://{twitterUrl}{endpoint}", {'variables':json.dumps(vars),'features':tweetFeedGraphqlFeatures,'fieldToggles':'{"withArticlePlainText":false}'},headers=reqHeaders) + if feed.status_code == 403: + raise TwExtractError(403, "Extract error") output = feed.json() if "errors" in output: # pick the first error and create a twExtractError error = output["errors"][0] raise TwExtractError(error["code"], error["message"]) - return output + timelineInstructions = output['data']['user']['result']['timeline']['timeline']['instructions'] + #tweetIds=None + tweets=None + for instruction in timelineInstructions: + if 'type' in instruction and instruction['type'] == 'TimelineAddEntries': + entries = instruction['entries'] + #tweetIds = [] + tweets = [] + for entry in entries: + if entry['entryId'].startswith("tweet-"): + # get the tweet ID from the entryId + #tweetId = entry['entryId'].split("-")[1] + #tweetIds.append(tweetId) + tweet = entry['content']['itemContent']['tweet_results']['result'] + tweets.append(tweet) + return tweets except Exception as e: continue raise TwExtractError(400, "Extract error") -def extractUserFeed(username,workaroundTokens): - pass - def lambda_handler(event, context): if ("queryStringParameters" not in event): return { diff --git a/twExtract/twUtils.py b/twExtract/twUtils.py index 26dbf1f..20f72ae 100644 --- a/twExtract/twUtils.py +++ b/twExtract/twUtils.py @@ -2,6 +2,9 @@ import math import hashlib import base64 import uuid +from x_client_transaction import ClientTransaction +from x_client_transaction.utils import handle_x_migration +import requests digits = "0123456789abcdefghijklmnopqrstuvwxyz" def baseConversion(x, base): @@ -31,5 +34,21 @@ def calcSyndicationToken(idStr): c = '0' return c -def generate_transaction_id(method: str, path: str) -> str: - return "?" # not implemented \ No newline at end of file +def get_twitter_homepage(headers=None): + if headers is None: + headers = {"Authority": "x.com", + "Accept-Language": "en-US,en;q=0.9", + "Cache-Control": "no-cache", + "Referer": "https://x.com", + "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/127.0.0.0 Safari/537.36", + "X-Twitter-Active-User": "yes", + "X-Twitter-Client-Language": "en"} + if 'Authorization' in headers: + del headers['Authorization'] + response = requests.get("https://x.com/home", headers=headers) + return response + +def generate_transaction_id(method: str, path: str,headers=None) -> str: + ct = ClientTransaction(get_twitter_homepage(headers=headers)) + transaction_id = ct.generate_transaction_id(method=method, path=path) + return transaction_id \ No newline at end of file diff --git a/twitfix.py b/twitfix.py index 1ef763e..bdae23c 100644 --- a/twitfix.py +++ b/twitfix.py @@ -290,9 +290,17 @@ def getTweetData(twitter_url,include_txt="false",include_rtf="false"): addVnfToLinkCache(twitter_url,tweetData) return tweetData -def getUserData(twitter_url): +def getUserData(twitter_url,includeFeed=False): rawUserData = twExtract.extractUser(twitter_url,workaroundTokens=config['config']['workaroundTokens'].split(',')) userData = getApiUserResponse(rawUserData) + + if includeFeed: + feed = twExtract.extractUserFeedFromId(userData['id'],workaroundTokens=config['config']['workaroundTokens'].split(',')) + apiFeed = [] + for tweet in feed: + apiFeed.append(getApiResponse(tweet)) + userData['latestTweets'] = apiFeed + return userData @app.route('/') # Default endpoint used by everything @@ -322,7 +330,7 @@ def twitfix(sub_path): username=sub_path.split("/")[0] extra = sub_path.split("/")[1] if extra in [None,"with_replies","media","likes","highlights","superfollows","media",''] and username != "" and username != None: - userData = getUserData(f"https://twitter.com/{username}") + userData = getUserData(f"https://twitter.com/{username}","withFeed" in request.args) if isApiRequest: if userData is None: abort(404) diff --git a/vxApi.py b/vxApi.py index 2282bde..e44aca5 100644 --- a/vxApi.py +++ b/vxApi.py @@ -98,7 +98,7 @@ def getApiResponse(tweet,include_txt=False,include_rtf=False): if "hashtags" in tweetL["entities"]: for i in tweetL["entities"]["hashtags"]: hashtags.append(i["text"]) - elif "card" in tweet and tweet['card']['name'] == "player": + elif "card" in tweet and 'name' in tweet['card'] and tweet['card']['name'] == "player": width = None height = None vidUrl = None