Synced repo using 'sync_with_huggingface' Github Action

Dockerfile

@@ -27,6 +27,9 @@ RUN --mount=type=secret,id=access_token_secret,mode=0444,required=true \
 RUN --mount=type=secret,id=bearer_token,mode=0444,required=true \
    cat /run/secrets/bearer_token > /test
 
+RUN --mount=type=secret,id=hf_token,mode=0444,required=true \
+   cat /run/secrets/hf_token > /test
+
 # Set up a new user named "user" with user ID 1000
 RUN useradd -m -u 1000 user
 

LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2023 Nathan Raw
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

main.py

@@ -1,25 +1,40 @@
 import logging
 import os
+import re
 import time
-from pathlib import Path
+import traceback
 
 import openai
+import pandas as pd
+import requests
 import tweepy
+from huggingface_hub import upload_file
 
 logger = logging.getLogger()
 logging.basicConfig(level=logging.INFO)
 logger.setLevel(logging.INFO)
 
+bot_user_id = 1612106815096999936
+bot_user_name = "RoastOrToastGPT"
+
+persistent_storage_repo_id = "team6/roast-history"
+persistent_storage_file_name = "history.csv"
+persistent_storage_file_url = (
+    f"https://huggingface.co/datasets/{persistent_storage_repo_id}/resolve/main/{persistent_storage_file_name}"
+)
+
+# will be used to remove the @roastortoastgpt from the tweet text (case insensitive)
+pattern_at_mention = re.compile(re.escape("@roastortoastgpt"), re.IGNORECASE)
+
+print(persistent_storage_file_url)
+
 openai_key = os.environ.get("openai_api_key")
 consumer_key = os.environ.get("consumer_key")
 consumer_secret_key = os.environ.get("consumer_secret_key")
 access_token = os.environ.get("access_token")
 access_token_secret = os.environ.get("access_token_secret")
 bearer_token = os.environ.get("bearer_token")
-
-home_path = os.environ.get("HOME")
-last_id_filepath = home_path + "/app/last_id.txt"
-
+hf_token = os.environ.get("hf_token")
 
 client = tweepy.Client(
     bearer_token=bearer_token,
@@ -31,25 +46,6 @@ client = tweepy.Client(
 )
 
 
-def get_last_tweet(file_path: str):
-    path = Path(file_path)
-    if not path.exists():
-        mentions = client.get_users_mentions(
-            id=1612106815096999936,
-            expansions=["author_id", "in_reply_to_user_id", "referenced_tweets.id"],
-        )
-        since_id = mentions.data[0].id
-        put_last_tweet(file_path, since_id)
-        return since_id
-    return int(path.read_text().strip())
-
-
-def put_last_tweet(file_path: str, id: str):
-    Path(file_path).write_text(str(id))
-    logger.info("Updated the file with the latest tweet Id")
-    return
-
-
 def response(text):
     openai.api_key = openai_key
     response = openai.Completion.create(
@@ -78,71 +74,134 @@ def toast(text):
 
 
 def reply_to_mentions():
-    since_id = get_last_tweet(last_id_filepath)
+    df = pd.read_csv(persistent_storage_file_url)
+    last_tweet_id = df.iloc[-1]["id"]
+
+    # List of unique conversation ids that we've already responded to.
+    # This is to prevent us from responding to the same conversation twice.
+    all_convo_ids = df["conversation_id"].unique().tolist()
+
+    # get the mentions. These are both direct mentions and replies to our tweets
     mentions = client.get_users_mentions(
-        id=1612106815096999936,
-        since_id=since_id,
+        id=bot_user_id,
         expansions=["author_id", "in_reply_to_user_id", "referenced_tweets.id"],
+        tweet_fields=["conversation_id"],
+        since_id=last_tweet_id,
     )
+
+    # if there are no new mentions, return
     if mentions.data is None:
+        # log it
         logger.info("No new mentions found")
         return
 
+    data_to_add = {"id": [], "conversation_id": []}
+    # otherwise, iterate through the mentions and respond to them
+    # we iterate through the mentions in reverse order so that we respond to the oldest mentions first
     for mention in reversed(mentions.data):
-        try:
-            if mention.author_id == 1612106815096999936:
-                continue
-            
-            if mention.referenced_tweets is None:
-                logger.info(f"Skipping {mention.id} as it is not a reply")
-            tweet_to_roast_id = mention.referenced_tweets[0].id
-            tweet_to_roast = client.get_tweet(id=tweet_to_roast_id)
-            text_to_roast = tweet_to_roast.data.text
-
-            text_out = None
-            text_in = mention.text.lower()
-            text_in = text_in.replace('@roastortoastgpt', '')
-            text_in = text_in.replace('roastortoastgpt','')
-
-            logger.info(f"In Text: {text_in}")
-            if "roast" in text_in:
-                logger.info(f"Roasting!")
-                text_out = roast(text_to_roast)
-            elif "toast" in text_in:
-                logger.info("Toasting!")
-                text_out = toast(text_to_roast)
-
-            if text_out is None:
-                continue
-
-            logger.info(f"Out Text: {text_out}")
-        except Exception as e:
-            logger.error(e)
+
+        if mention.author_id == bot_user_id:
+            # don't respond to our own tweets
+            logger.info(f"Skipping {mention.id} as it is from the bot")
             continue
 
-        try:
-            logger.info(f"Responding to: {mention.id}")
-            client.create_tweet(text=text_out, in_reply_to_tweet_id=mention.id)
-        except Exception as e:
-            logger.error(e)
+        if mention.in_reply_to_user_id == bot_user_id:
+            # don't respond to our own tweets
+            logger.info(f"Skipping {mention.id} as the tweet to roast is from the bot")
+            continue
+
+        if not mention.referenced_tweets:
+            logger.info(f"Skipping {mention.id} as it is not a reply")
+            continue
+
+        # if we've already responded to this conversation, skip it
+        # also should catch the case where we've already responded to this tweet (though that shouldn't happen)
+        if mention.conversation_id in all_convo_ids:
+            logger.info(f"Skipping {mention.id} as we've already responded to this conversation")
+            continue
+
+        logger.info(f"Responding to {mention.id}, which said {mention.text}")
+
+        tweet_to_roast_id = mention.referenced_tweets[0].id
+        tweet_to_roast = client.get_tweet(tweet_to_roast_id)
+        text_to_roast = tweet_to_roast.data.text
+
+        mention_text = mention.text
+        mention_text = pattern_at_mention.sub("", mention_text)
+        logger.info(f"Mention Text: {mention_text}")
+
+        if "roast" in mention_text.lower():
+            logger.info(f"Roasting {mention.id}")
+            text_out = roast(text_to_roast)
+        elif "toast" in mention_text.lower():
+            logger.info(f"Toasting {mention.id}")
+            text_out = toast(text_to_roast)
+        else:
+            logger.info(f"Skipping {mention.id} as it is not a roast or toast")
             continue
 
-    put_last_tweet(last_id_filepath, mention.id)
+        # Quote tweet the tweet to roast
+        logger.info(f"Quote tweeting {tweet_to_roast_id} with response: {text_out}")
+        quote_tweet_response = client.create_tweet(
+            text=text_out,
+            quote_tweet_id=tweet_to_roast_id,
+        )
+        print("QUOTE TWEET RESPONSE", quote_tweet_response.data)
+        response_quote_tweet_id = quote_tweet_response.data.get("id")
+        logger.info(f"Response Quote Tweet ID: {response_quote_tweet_id}")
+        response_quote_tweet_url = f"https://twitter.com/{bot_user_name}/status/{response_quote_tweet_id}"
+        logger.info(f"Response Quote Tweet URL: {response_quote_tweet_url}")
+
+        # reply to the mention with the link to the response tweet
+        logger.info(f"Responding to: {mention.id}")
+        response_reply = client.create_tweet(
+            text=f"Here's my response: {response_quote_tweet_url}",
+            in_reply_to_tweet_id=mention.id,
+        )
+        response_reply_id = response_reply.data.get("id")
+        logger.info(f"Response Reply ID: {response_reply_id}")
+
+        # add the mention to the history
+        data_to_add["id"].append(mention.id)
+        data_to_add["conversation_id"].append(mention.conversation_id)
+
+        # add a line break to the log
+        logger.info("-" * 100)
+
+    # update the history df and upload it to the persistent storage repo
+    if len(data_to_add["id"]) == 0:
+        logger.info("No new mentions to add to the history")
+        return
+
+    logger.info(f"Adding {len(data_to_add['id'])} new mentions to the history")
+
+    df_to_add = pd.DataFrame(data_to_add)
+    df = pd.concat([df, df_to_add], ignore_index=True)
+    df.to_csv(persistent_storage_file_name, index=False)
+    upload_file(
+        repo_id=persistent_storage_repo_id,
+        path_or_fileobj=persistent_storage_file_name,
+        path_in_repo=persistent_storage_file_name,
+        repo_type="dataset",
+        token=hf_token,
+    )
 
 
 def main():
+    logger.info("Starting up...")
+
     while True:
         try:
+            # Dummy request to keep the Hugging Face Space awake
+            requests.get("https://team6-roast.hf.space/")
             reply_to_mentions()
         except Exception as e:
             logger.error(e)
-            # Print more helpful error messages with line number
-            import traceback
-
             traceback.print_exc()
 
-        time.sleep(60)
+        logger.info("Sleeping for 30 seconds...")
+        time.sleep(30)
 
 
-# if __name__ == "__main__":
-#     main()
+if __name__ == "__main__":
+    main()

requirements.txt

@@ -1,6 +1,5 @@
-fastapi==0.74.*
-requests==2.27.*
-sentencepiece==0.1.*
-uvicorn[standard]==0.17.*
 openai
-tweepy
+pandas
+requests
+tweepy
+huggingface_hub