app.py

import gradio as gr
from huggingface_hub import InferenceClient, HfApi
from datetime import datetime
import uuid
import os
import json

# ---- Configuration ----
MODEL_NAME = "HuggingFaceH4/zephyr-7b-beta"
DATASET_REPO = "frimelle/companion-chat-logs"
HF_TOKEN = os.environ.get("HF_TOKEN")

# ---- Load system prompt ----
with open("system_prompt.txt", "r") as f:
    SYSTEM_PROMPT = f.read()

client = InferenceClient(MODEL_NAME)
api = HfApi()

# ---- Chatbot Class with per-user session ID ----
class SessionChatBot:
    def __init__(self):
        self.session_id = str(uuid.uuid4())
        self.today_date = datetime.now().strftime("%Y-%m-%d")
        self.local_log_path = f"chatlog_{self.today_date}_{self.session_id}.jsonl"
        self.remote_log_path = f"sessions/{self.today_date}/{self.session_id}.jsonl"

    def append_to_session_log(self, user_message, assistant_message):
        row = {
            "timestamp": datetime.now().isoformat(),
            "user": user_message,
            "assistant": assistant_message,
            "system_prompt": SYSTEM_PROMPT,
            "session_id": self.session_id
        }
        with open(self.local_log_path, "a", encoding="utf-8") as f:
            f.write(json.dumps(row) + "\n")
        api.upload_file(
            path_or_fileobj=self.local_log_path,
            path_in_repo=self.remote_log_path,
            repo_id=DATASET_REPO,
            repo_type="dataset",
            token=HF_TOKEN
        )

    def respond(self, message, history):
        messages = [{"role": "system", "content": SYSTEM_PROMPT}]
        for user_msg, bot_msg in history:
            if user_msg:
                messages.append({"role": "user", "content": user_msg})
            if bot_msg:
                messages.append({"role": "assistant", "content": bot_msg})
        messages.append({"role": "user", "content": message})

        response = ""
        for chunk in client.chat_completion(
            messages,
            max_tokens=512,
            stream=True,
            temperature=0.7,
            top_p=0.95,
        ):
            token = chunk.choices[0].delta.content
            if token:
                response += token
                yield response

        # Save log after full response
        self.append_to_session_log(message, response)

    def report_interaction(self):
        if not os.path.exists(self.local_log_path):
            return "No session log found."

        with open(self.local_log_path, "r", encoding="utf-8") as f:
            lines = f.readlines()

        if not lines:
            return "No conversation to report."

        # Mark last interaction as reported
        last_entry = json.loads(lines[-1])
        last_entry["reported"] = True
        lines[-1] = json.dumps(last_entry) + "\n"

        # Overwrite file
        with open(self.local_log_path, "w", encoding="utf-8") as f:
            f.writelines(lines)

        # Upload updated log to Hugging Face
        api.upload_file(
            path_or_fileobj=self.local_log_path,
            path_in_repo=self.remote_log_path,
            repo_id=DATASET_REPO,
            repo_type="dataset",
            token=HF_TOKEN
        )
        return "Interaction reported successfully."

# ---- Instantiate and Share Bot Session Globally ----
chatbot_instance = SessionChatBot()

def create_chatbot():
    return chatbot_instance.respond

# ---- Build Gradio Interface ----
with gr.Blocks() as demo:
    chatbot = gr.ChatInterface(fn=create_chatbot(), title="BoundrAI")
    report_btn = gr.Button("Report Companion Interaction")
    status_box = gr.Textbox(label="Report Status", interactive=False)

    def report():
        return chatbot_instance.report_interaction()

    report_btn.click(fn=report, outputs=status_box)

if __name__ == "__main__":
    demo.launch()