app.py

import os
import sys
import tempfile
import time
import itertools
import streamlit as st
import pandas as pd
from threading import Thread
from io import StringIO

# Add 'src' to Python path
sys.path.append(os.path.join(os.path.dirname(__file__), 'src'))
from main import run_pipeline

st.set_page_config(page_title="📰 AI News Analyzer", layout="wide")
st.title("🧠 AI-Powered Investing News Analyzer")

# === API Key Input ===
st.subheader("🔐 API Keys")
openai_api_key = st.text_input("OpenAI API Key", type="password").strip()
tavily_api_key = st.text_input("Tavily API Key", type="password").strip()

# === Topic Input ===
st.subheader("📈 Topics of Interest")
topics_data = []

with st.form("topics_form"):
    topic_count = st.number_input("How many topics?", min_value=1, max_value=10, value=1, step=1)

    for i in range(topic_count):
        col1, col2 = st.columns(2)
        with col1:
            topic = st.text_input(f"Topic {i+1}", key=f"topic_{i}")
        with col2:
            days = st.number_input(f"Timespan (days)", min_value=1, max_value=30, value=7, key=f"days_{i}")
        topics_data.append({"topic": topic, "timespan_days": days})

    submitted = st.form_submit_button("Run Analysis")

# === Submission logic ===
if submitted:
    if not openai_api_key or not tavily_api_key or not all([td['topic'] for td in topics_data]):
        st.warning("Please fill in all fields.")
    else:
        os.environ["OPENAI_API_KEY"] = openai_api_key
        os.environ["TAVILY_API_KEY"] = tavily_api_key

        df = pd.DataFrame(topics_data)
        with tempfile.NamedTemporaryFile(delete=False, suffix=".csv") as tmp_csv:
            df.to_csv(tmp_csv.name, index=False)
            csv_path = tmp_csv.name

        # Placeholders
        spinner_box = st.empty()
        log_box = st.empty()

        # Rotating spinner text
        rotating = True
        logs = []

        def log(msg):
            logs.append(msg)
            log_box.code("\n".join(logs))

        def rotating_messages():
            messages = itertools.cycle([
                "🔍 Searching financial news...",
                "🧠 Running language models...",
                "📊 Generating investment reports...",
                "📝 Compiling markdown...",
                "🔎 Looking for value signals..."
            ])
            while rotating:
                spinner_box.markdown(f"⏳ {next(messages)}")
                time.sleep(1.4)

        rotator_thread = Thread(target=rotating_messages)
        rotator_thread.start()

        try:
            # Log: API checks
            log("🔐 Checking API keys...")
            import openai
            openai_client = openai.OpenAI(api_key=openai_api_key)
            openai_client.models.list()  # triggers check
            log("✅ OpenAI API key is valid.")

            import requests
            tavily_check = requests.post(
                "https://api.tavily.com/search",
                headers={"Authorization": f"Bearer {tavily_api_key}"},
                json={"query": "test", "days": 1, "max_results": 1}
            )
            if tavily_check.status_code == 200:
                log("✅ Tavily API key is valid.")
            else:
                raise ValueError(f"Tavily key failed: {tavily_check.status_code} - {tavily_check.text}")

            # Run main pipeline
            log("🚀 Running analysis pipeline...")
            output_path = run_pipeline(csv_path, tavily_api_key, progress_callback=log)

            rotating = False
            rotator_thread.join()
            spinner_box.success("✅ Analysis complete!")

            if output_path and isinstance(output_path, list):
                for path in output_path:
                    if os.path.exists(path):
                        with open(path, 'r', encoding='utf-8') as file:
                            html_content = file.read()
                            filename = os.path.basename(path)

                            st.download_button(
                                label=f"📥 Download {filename}",
                                data=html_content,
                                file_name=filename,
                                mime="text/html"
                            )
                            st.components.v1.html(html_content, height=600, scrolling=True)
            else:
                st.error("❌ No reports were generated.")

        except Exception as e:
            rotating = False
            rotator_thread.join()
            spinner_box.error("❌ Failed.")
            log_box.error(f"❌ Error: {e}")

##################################################################################################
##################################################################################################
# import os
# import sys
# import tempfile
# import streamlit as st
# import pandas as pd
# from io import StringIO

# # Add 'src' to Python path so we can import main.py
# sys.path.append(os.path.join(os.path.dirname(__file__), 'src'))
# from main import run_pipeline

# st.set_page_config(page_title="📰 AI News Analyzer", layout="wide")
# st.title("🧠 AI-Powered Investing News Analyzer")

# # === API Key Input ===
# st.subheader("🔐 API Keys")
# openai_api_key = st.text_input("OpenAI API Key", type="password").strip()
# tavily_api_key = st.text_input("Tavily API Key", type="password").strip()

# # === Topic Input ===
# st.subheader("📈 Topics of Interest")
# topics_data = []

# with st.form("topics_form"):
#     topic_count = st.number_input("How many topics?", min_value=1, max_value=10, value=1, step=1)

#     for i in range(topic_count):
#         col1, col2 = st.columns(2)
#         with col1:
#             topic = st.text_input(f"Topic {i+1}", key=f"topic_{i}")
#         with col2:
#             days = st.number_input(f"Timespan (days)", min_value=1, max_value=30, value=7, key=f"days_{i}")
#         topics_data.append({"topic": topic, "timespan_days": days})

#     submitted = st.form_submit_button("Run Analysis")

# # === Submission logic ===
# if submitted:
#     if not openai_api_key or not tavily_api_key or not all([td['topic'] for td in topics_data]):
#         st.warning("Please fill in all fields.")
#     else:
#         os.environ["OPENAI_API_KEY"] = openai_api_key
#         os.environ["TAVILY_API_KEY"] = tavily_api_key

#         df = pd.DataFrame(topics_data)
#         with tempfile.NamedTemporaryFile(delete=False, suffix=".csv") as tmp_csv:
#             df.to_csv(tmp_csv.name, index=False)
#             csv_path = tmp_csv.name

#         progress_box = st.empty()

#         def show_progress(msg):
#             progress_box.markdown(f"⏳ {msg}")

#         try:
#             output_path = run_pipeline(csv_path, tavily_api_key, progress_callback=show_progress)
#             progress_box.success("✅ Analysis complete!")

#             if output_path and isinstance(output_path, list):
#                 for path in output_path:
#                     if os.path.exists(path):
#                         with open(path, 'r', encoding='utf-8') as file:
#                             html_content = file.read()
#                             filename = os.path.basename(path)

#                             st.download_button(
#                                 label=f"📥 Download {filename}",
#                                 data=html_content,
#                                 file_name=filename,
#                                 mime="text/html"
#                             )
#                             st.components.v1.html(html_content, height=600, scrolling=True)
#             else:
#                 st.error("❌ No reports were generated.")
#         except Exception as e:
#             progress_box.error(f"❌ Error: {e}")