Spaces:
Sleeping
Sleeping
| import io | |
| import re | |
| import streamlit as st | |
| import glob | |
| import os | |
| from PIL import Image | |
| import fitz | |
| from reportlab.lib.pagesizes import A4 | |
| from reportlab.platypus import SimpleDocTemplate, Paragraph, Spacer, Table, TableStyle | |
| from reportlab.lib.styles import getSampleStyleSheet, ParagraphStyle | |
| from reportlab.lib import colors | |
| from reportlab.pdfbase import pdfmetrics | |
| from reportlab.pdfbase.ttfonts import TTFont | |
| import unicodedata | |
| import asyncio | |
| import websockets | |
| import uuid | |
| from datetime import datetime | |
| import random | |
| import time | |
| import hashlib | |
| import base64 | |
| import streamlit.components.v1 as components | |
| import edge_tts | |
| from audio_recorder_streamlit import audio_recorder | |
| import nest_asyncio | |
| import pytz | |
| import shutil | |
| import anthropic | |
| import openai | |
| from PyPDF2 import PdfReader | |
| import threading | |
| import json | |
| import zipfile | |
| from gradio_client import Client | |
| from dotenv import load_dotenv | |
| from streamlit_marquee import streamlit_marquee | |
| from collections import defaultdict, Counter | |
| import pandas as pd | |
| nest_asyncio.apply() | |
| st.set_page_config(layout="wide", initial_sidebar_state="collapsed") | |
| icons = '๐ค๐ง ๐ฌ๐' | |
| Site_Name = '๐ค๐ง Chat & Quote Node๐๐ฌ' | |
| START_ROOM = "Sector ๐" | |
| FUN_USERNAMES = { | |
| "CosmicJester ๐": "en-US-AriaNeural", | |
| "PixelPanda ๐ผ": "en-US-JennyNeural", | |
| "QuantumQuack ๐ฆ": "en-GB-SoniaNeural", | |
| "StellarSquirrel ๐ฟ๏ธ": "en-AU-NatashaNeural", | |
| "GizmoGuru โ๏ธ": "en-CA-ClaraNeural", | |
| "NebulaNinja ๐ ": "en-US-GuyNeural", | |
| "ByteBuster ๐พ": "en-GB-RyanNeural", | |
| "GalacticGopher ๐": "en-AU-WilliamNeural", | |
| "RocketRaccoon ๐": "en-CA-LiamNeural", | |
| "EchoElf ๐ง": "en-US-AnaNeural", | |
| "PhantomFox ๐ฆ": "en-US-BrandonNeural", | |
| "WittyWizard ๐ง": "en-GB-ThomasNeural", | |
| "LunarLlama ๐": "en-AU-FreyaNeural", | |
| "SolarSloth โ๏ธ": "en-CA-LindaNeural", | |
| "AstroAlpaca ๐ฆ": "en-US-ChristopherNeural", | |
| "CyberCoyote ๐บ": "en-GB-ElliotNeural", | |
| "MysticMoose ๐ฆ": "en-AU-JamesNeural", | |
| "GlitchGnome ๐ง": "en-CA-EthanNeural", | |
| "VortexViper ๐": "en-US-AmberNeural", | |
| "ChronoChimp ๐": "en-GB-LibbyNeural" | |
| } | |
| EDGE_TTS_VOICES = list(set(FUN_USERNAMES.values())) | |
| FILE_EMOJIS = {"md": "๐", "mp3": "๐ต", "png": "๐ผ๏ธ", "mp4": "๐ฅ", "zip": "๐ฆ"} | |
| for d in ["chat_logs", "vote_logs", "audio_logs", "history_logs", "audio_cache", "paper_metadata"]: | |
| os.makedirs(d, exist_ok=True) | |
| CHAT_DIR = "chat_logs" | |
| VOTE_DIR = "vote_logs" | |
| MEDIA_DIR = "." | |
| AUDIO_CACHE_DIR = "audio_cache" | |
| AUDIO_DIR = "audio_logs" | |
| PAPER_DIR = "paper_metadata" | |
| STATE_FILE = "user_state.txt" | |
| CHAT_FILE = os.path.join(CHAT_DIR, "global_chat.md") | |
| QUOTE_VOTES_FILE = os.path.join(VOTE_DIR, "quote_votes.md") | |
| IMAGE_VOTES_FILE = os.path.join(VOTE_DIR, "image_votes.md") | |
| HISTORY_FILE = os.path.join(VOTE_DIR, "vote_history.md") | |
| load_dotenv() | |
| anthropic_key = os.getenv('ANTHROPIC_API_KEY', st.secrets.get('ANTHROPIC_API_KEY', "")) | |
| openai_api_key = os.getenv('OPENAI_API_KEY', st.secrets.get('OPENAI_API_KEY', "")) | |
| openai_client = openai.OpenAI(api_key=openai_api_key) | |
| def format_timestamp_prefix(username=""): | |
| central = pytz.timezone('US/Central') | |
| now = datetime.now(central) | |
| return f"{now.strftime('%Y%m%d_%H%M%S')}-by-{username}" | |
| class PerformanceTimer: | |
| def __init__(self, name): | |
| self.name, self.start = name, None | |
| def __enter__(self): | |
| self.start = time.time() | |
| return self | |
| def __exit__(self, *args): | |
| duration = time.time() - self.start | |
| st.session_state['operation_timings'][self.name] = duration | |
| st.session_state['performance_metrics'][self.name].append(duration) | |
| def init_session_state(): | |
| defaults = { | |
| 'server_running': False, 'server_task': None, 'active_connections': {}, | |
| 'media_notifications': [], 'last_chat_update': 0, 'displayed_chat_lines': [], | |
| 'message_text': "", 'audio_cache': {}, 'pasted_image_data': None, | |
| 'quote_line': None, 'refresh_rate': 10, 'base64_cache': {}, | |
| 'transcript_history': [], 'last_transcript': "", 'image_hashes': set(), | |
| 'tts_voice': "en-US-AriaNeural", 'chat_history': [], 'marquee_settings': { | |
| "background": "#1E1E1E", "color": "#FFFFFF", "font-size": "14px", | |
| "animationDuration": "20s", "width": "100%", "lineHeight": "35px" | |
| }, 'operation_timings': {}, 'performance_metrics': defaultdict(list), | |
| 'enable_audio': True, 'download_link_cache': {}, 'username': None, | |
| 'autosend': True, 'autosearch': True, 'last_message': "", 'last_query': "", | |
| 'mp3_files': {}, 'timer_start': time.time(), 'quote_index': 0, | |
| 'quote_source': "famous", 'last_sent_transcript': "", 'old_val': None, | |
| 'last_refresh': time.time(), 'paper_metadata': {}, 'paste_image_base64': "", | |
| 'use_arxiv': True, 'use_arxiv_audio': False, 'speech_processed': False, | |
| 'auto_refresh': True | |
| } | |
| for k, v in defaults.items(): | |
| if k not in st.session_state: | |
| st.session_state[k] = v | |
| def update_marquee_settings_ui(): | |
| st.sidebar.markdown("### ๐ฏ Marquee Settings") | |
| cols = st.sidebar.columns(2) | |
| with cols[0]: | |
| st.session_state['marquee_settings']['background'] = st.color_picker("๐จ Background", "#1E1E1E") | |
| st.session_state['marquee_settings']['color'] = st.color_picker("โ๏ธ Text", "#FFFFFF") | |
| with cols[1]: | |
| st.session_state['marquee_settings']['font-size'] = f"{st.slider('๐ Size', 10, 24, 14)}px" | |
| st.session_state['marquee_settings']['animationDuration'] = f"{st.slider('โฑ๏ธ Speed', 1, 20, 20)}s" | |
| def display_marquee(text, settings, key_suffix=""): | |
| truncated = text[:280] + "..." if len(text) > 280 else text | |
| streamlit_marquee(content=truncated, **settings, key=f"marquee_{key_suffix}") | |
| st.write("") | |
| def clean_text_for_tts(text): | |
| return re.sub(r'[#*!\[\]]+', '', ' '.join(text.split()))[:200] or "No text" | |
| def clean_text_for_filename(text): | |
| return '_'.join(re.sub(r'[^\w\s-]', '', text.lower()).split())[:50] | |
| def get_high_info_terms(text, top_n=10): | |
| stop_words = {'the', 'a', 'an', 'and', 'or', 'but', 'in', 'on', 'at', 'to', 'for', 'of', 'with'} | |
| words = re.findall(r'\b\w+(?:-\w+)*\b', text.lower()) | |
| bi_grams = [' '.join(pair) for pair in zip(words, words[1:])] | |
| filtered = [t for t in words + bi_grams if t not in stop_words and len(t.split()) <= 2] | |
| return [t for t, _ in Counter(filtered).most_common(top_n)] | |
| def generate_filename(prompt, username, file_type="md", title=None): | |
| timestamp = format_timestamp_prefix(username) | |
| if title: | |
| high_info = '-'.join(get_high_info_terms(title, 5)) | |
| return f"{timestamp}-{clean_text_for_filename(prompt[:20])}-{high_info}.{file_type}" | |
| hash_val = hashlib.md5(prompt.encode()).hexdigest()[:8] | |
| return f"{timestamp}-{hash_val}.{file_type}" | |
| def create_file(prompt, username, file_type="md", title=None): | |
| filename = generate_filename(prompt, username, file_type, title) | |
| with open(filename, 'w', encoding='utf-8') as f: | |
| f.write(prompt) | |
| return filename | |
| def get_download_link(file, file_type="mp3"): | |
| cache_key = f"dl_{file}" | |
| if cache_key not in st.session_state['download_link_cache']: | |
| with open(file, "rb") as f: | |
| b64 = base64.b64encode(f.read()).decode() | |
| mime_types = {"mp3": "audio/mpeg", "png": "image/png", "mp4": "video/mp4", "md": "text/markdown", "zip": "application/zip"} | |
| st.session_state['download_link_cache'][cache_key] = f'<a href="data:{mime_types.get(file_type, "application/octet-stream")};base64,{b64}" download="{os.path.basename(file)}">{FILE_EMOJIS.get(file_type, "Download")} Download {os.path.basename(file)}</a>' | |
| return st.session_state['download_link_cache'][cache_key] | |
| def save_username(username): | |
| try: | |
| with open(STATE_FILE, 'w') as f: | |
| f.write(username) | |
| except Exception as e: | |
| print(f"Failed to save username: {e}") | |
| def load_username(): | |
| if os.path.exists(STATE_FILE): | |
| try: | |
| with open(STATE_FILE, 'r') as f: | |
| return f.read().strip() | |
| except Exception as e: | |
| print(f"Failed to load username: {e}") | |
| return None | |
| def concatenate_markdown_files(exclude_files=["README.md"]): | |
| md_files = sorted([f for f in glob.glob("*.md") if os.path.basename(f) not in exclude_files], key=os.path.getmtime) | |
| all_md_content = "" | |
| for i, md_file in enumerate(md_files, 1): | |
| with open(md_file, 'r', encoding='utf-8') as f: | |
| content = f.read().strip() | |
| all_md_content += f"{i}. {content}\n" | |
| return all_md_content.rstrip() | |
| def get_chat_text_only(exclude_files=["README.md"]): | |
| md_files = sorted([f for f in glob.glob("*.md") if os.path.basename(f) not in exclude_files], key=os.path.getmtime) | |
| chat_text = "" | |
| for i, md_file in enumerate(md_files, 1): | |
| with open(md_file, 'r', encoding='utf-8') as f: | |
| content = f.read().strip() | |
| lines = content.split('\n') | |
| for line in lines: | |
| if line.strip() and not line.startswith('#'): | |
| match = re.match(r'\[(.*?)\]\s(.*?)\s\((.*?)\):\s*(.*)', line) | |
| if match: | |
| message = match.group(4).strip() | |
| if message.startswith('```markdown'): | |
| message = message.replace('```markdown', '').replace('```', '').strip() | |
| chat_text += f"{message}\n" | |
| return chat_text.rstrip() | |
| async def async_edge_tts_generate(text, voice, username, rate=0, pitch=0, file_format="mp3"): | |
| cache_key = f"{text[:100]}_{voice}_{rate}_{pitch}_{file_format}" | |
| if cache_key in st.session_state['audio_cache']: | |
| return st.session_state['audio_cache'][cache_key], 0 | |
| start_time = time.time() | |
| text = clean_text_for_tts(text) | |
| if not text or text == "No text": | |
| print(f"Skipping audio generation for empty/invalid text: '{text}'") | |
| return None, 0 | |
| filename = f"{format_timestamp_prefix(username)}-{hashlib.md5(text.encode()).hexdigest()[:8]}.{file_format}" | |
| try: | |
| communicate = edge_tts.Communicate(text, voice, rate=f"{rate:+d}%", pitch=f"{pitch:+d}Hz") | |
| await communicate.save(filename) | |
| if os.path.exists(filename) and os.path.getsize(filename) > 0: | |
| st.session_state['audio_cache'][cache_key] = filename | |
| return filename, time.time() - start_time | |
| else: | |
| print(f"Audio file {filename} was not created or is empty.") | |
| return None, 0 | |
| except edge_tts.exceptions.NoAudioReceived as e: | |
| print(f"No audio received for text: '{text}' with voice: {voice}. Error: {e}") | |
| return None, 0 | |
| except Exception as e: | |
| print(f"Error generating audio for text: '{text}' with voice: {voice}. Error: {e}") | |
| return None, 0 | |
| def play_and_download_audio(file_path): | |
| if file_path and os.path.exists(file_path): | |
| st.audio(file_path) | |
| st.markdown(get_download_link(file_path), unsafe_allow_html=True) | |
| else: | |
| st.warning(f"Audio file not found: {file_path}") | |
| def load_mp3_viewer(): | |
| mp3_files = sorted(glob.glob("*.mp3"), key=os.path.getmtime) | |
| for i, mp3 in enumerate(mp3_files, 1): | |
| filename = os.path.basename(mp3) | |
| if filename not in st.session_state['mp3_files']: | |
| st.session_state['mp3_files'][filename] = (i, mp3) | |
| async def save_chat_entry(username, message, voice, is_markdown=False): | |
| if not message.strip() or message == st.session_state.last_transcript: | |
| return None, None | |
| central = pytz.timezone('US/Central') | |
| timestamp = datetime.now(central).strftime("%Y-%m-%d %H:%M:%S") | |
| entry = f"[{timestamp}] {username} ({voice}): {message}" if not is_markdown else f"[{timestamp}] {username} ({voice}):\n```markdown\n{message}\n```" | |
| md_file = create_file(entry, username, "md") | |
| with open(CHAT_FILE, 'a') as f: | |
| f.write(f"{entry}\n") | |
| audio_file, _ Distance = await async_edge_tts_generate(message, voice, username) | |
| if audio_file: | |
| with open(HISTORY_FILE, 'a') as f: | |
| f.write(f"[{timestamp}] {username}: Audio - {audio_file}\n") | |
| st.session_state['mp3_files'][os.path.basename(audio_file)] = (len(st.session_state['chat_history']) + 1, audio_file) | |
| if st.session_state.get('speech_processed', False) and st.session_state.get('message_input', '') == message: | |
| st.session_state['message_input'] = "" | |
| st.session_state['speech_processed'] = False | |
| else: | |
| st.warning(f"Failed to generate audio for: {message}") | |
| await broadcast_message(f"{username}|{message}", "chat") | |
| st.session_state.last_chat_update = time.time() | |
| st.session_state.chat_history.append(entry) | |
| st.session_state.last_transcript = message | |
| return md_file, audio_file | |
| async def load_chat(): | |
| if not os.path.exists(CHAT_FILE): | |
| with open(CHAT_FILE, 'a') as f: | |
| f.write(f"# {START_ROOM} Chat\n\nWelcome to the cosmic hub! ๐ค\n") | |
| with open(CHAT_FILE, 'r') as f: | |
| content = f.read().strip() | |
| lines = content.split('\n') | |
| unique_lines = list(dict.fromkeys(line for line in lines if line.strip())) | |
| return unique_lines | |
| async def perform_claude_search(query, username, image=None): | |
| if not query.strip() or query == st.session_state.last_transcript: | |
| return None, None, None | |
| client = anthropic.Anthropic(api_key=anthropic_key) | |
| message_content = [{"type": "text", "text": query}] | |
| if image: | |
| buffered = io.BytesIO() | |
| image.save(buffered, format="PNG") | |
| img_base64 = base64.b64encode(buffered.getvalue()).decode('utf-8') | |
| message_content.append({ | |
| "type": "image", | |
| "source": { | |
| "type": "base64", | |
| "media_type": "image/png", | |
| "data": img_base64 | |
| } | |
| }) | |
| try: | |
| response = client.messages.create( | |
| model="claude-3-sonnet-20240229", | |
| max_tokens=1000, | |
| messages=[{"role": "user", "content": message_content}] | |
| ) | |
| result = response.content[0].text | |
| st.markdown(f"### Claude's Reply ๐ง \n{result}") | |
| except Exception as e: | |
| st.error(f"Claude processing failed: {e}") | |
| return None, None, None | |
| voice = FUN_USERNAMES.get(username, "en-US-AriaNeural") | |
| full_text = f"Prompt: {query}\nResponse: {result}" | |
| md_file, audio_file = await save_chat_entry(username, full_text, voice, True) | |
| return md_file, audio_file, result | |
| async def perform_arxiv_search(query, username, claude_result=None): | |
| if not query.strip() or query == st.session_state.last_transcript: | |
| return None, None | |
| if claude_result is None: | |
| client = anthropic.Anthropic(api_key=anthropic_key) | |
| claude_response = client.messages.create( | |
| model="claude-3-sonnet-20240229", | |
| max_tokens=1000, | |
| messages=[{"role": "user", "content": query}] | |
| ) | |
| claude_result = claude_response.content[0].text | |
| st.markdown(f"### Claude's Reply ๐ง \n{claude_result}") | |
| enhanced_query = f"{query}\n\n{claude_result}" | |
| gradio_client = Client("awacke1/Arxiv-Paper-Search-And-QA-RAG-Pattern") | |
| refs = gradio_client.predict( | |
| enhanced_query, 10, "Semantic Search", "mistralai/Mixtral-8x7B-Instruct-v0.1", api_name="/update_with_rag_md" | |
| )[0] | |
| result = f"๐ {enhanced_query}\n\n{refs}" | |
| voice = FUN_USERNAMES.get(username, "en-US-AriaNeural") | |
| md_file, audio_file = await save_chat_entry(username, result, voice, True) | |
| return md_file, audio_file | |
| async def perform_ai_lookup(q, vocal_summary=True, extended_refs=False, titles_summary=True, full_audio=False, useArxiv=True, useArxivAudio=False): | |
| start = time.time() | |
| client = anthropic.Anthropic(api_key=anthropic_key) | |
| response = client.messages.create( | |
| model="claude-3-sonnet-20240229", | |
| max_tokens=1000, | |
| messages=[{"role": "user", "content": q}] | |
| ) | |
| st.write("Claude's reply ๐ง :") | |
| st.markdown(response.content[0].text) | |
| result = response.content[0].text | |
| md_file = create_file(result, "System", "md") | |
| audio_file, _ = await async_edge_tts_generate(result, st.session_state['tts_voice'], "System") | |
| st.subheader("๐ Main Response Audio") | |
| play_and_download_audio(audio_file) | |
| papers = [] | |
| if useArxiv: | |
| q = q + result | |
| st.write('Running Arxiv RAG with Claude inputs.') | |
| gradio_client = Client("awacke1/Arxiv-Paper-Search-And-QA-RAG-Pattern") | |
| refs = gradio_client.predict( | |
| q, 20, "Semantic Search", "mistralai/Mixtral-8x7B-Instruct-v0.1", api_name="/update_with_rag_md" | |
| )[0] | |
| papers = parse_arxiv_refs(refs, q) | |
| for paper in papers: | |
| filename = create_file(generate_5min_feature_markdown(paper), "System", "md", paper['title']) | |
| paper['md_file'] = filename | |
| st.session_state['paper_metadata'][paper['title']] = filename | |
| if papers and useArxivAudio: | |
| await create_paper_audio_files(papers, q) | |
| elapsed = time.time() - start | |
| st.write(f"**Total Elapsed:** {elapsed:.2f} s") | |
| return result, papers | |
| async def websocket_handler(websocket, path): | |
| client_id = str(uuid.uuid4()) | |
| room_id = "chat" | |
| if room_id not in st.session_state.active_connections: | |
| st.session_state.active_connections[room_id] = {} | |
| st.session_state.active_connections[room_id][client_id] = websocket | |
| username = st.session_state.get('username', random.choice(list(FUN_USERNAMES.keys()))) | |
| chat_content = await load_chat() | |
| if not any(f"Client-{client_id}" in line for line in chat_content): | |
| await save_chat_entry("System ๐", f"{username} has joined {START_ROOM}!", "en-US-AriaNeural") | |
| try: | |
| async for message in websocket: | |
| if '|' in message: | |
| username, content = message.split('|', 1) | |
| voice = FUN_USERNAMES.get(username, "en-US-AriaNeural") | |
| await save_chat_entry(username, content, voice) | |
| else: | |
| await websocket.send("ERROR|Message format: username|content") | |
| except websockets.ConnectionClosed: | |
| await save_chat_entry("System ๐", f"{username} has left {START_ROOM}!", "en-US-AriaNeural") | |
| finally: | |
| if room_id in st.session_state.active_connections and client_id in st.session_state.active_connections[room_id]: | |
| del st.session_state.active_connections[room_id][client_id] | |
| async def broadcast_message(message, room_id): | |
| if room_id in st.session_state.active_connections: | |
| disconnected = [] | |
| for client_id, ws in st.session_state.active_connections[room_id].items(): | |
| try: | |
| await ws.send(message) | |
| except websockets.ConnectionClosed: | |
| disconnected.append(client_id) | |
| for client_id in disconnected: | |
| if client_id in st.session_state.active_connections[room_id]: | |
| del st.session_state.active_connections[room_id][client_id] | |
| async def run_websocket_server(): | |
| if not st.session_state.get('server_running', False): | |
| server = await websockets.serve(websocket_handler, '0.0.0.0', 8765) | |
| st.session_state['server_running'] = True | |
| await server.wait_closed() | |
| def start_websocket_server(): | |
| loop = asyncio.new_event_loop() | |
| asyncio.set_event_loop(loop) | |
| loop.run_until_complete(run_websocket_server()) | |
| class AudioProcessor: | |
| def __init__(self): | |
| self.cache_dir = AUDIO_CACHE_DIR | |
| os.makedirs(self.cache_dir, exist_ok=True) | |
| self.metadata = json.load(open(f"{self.cache_dir}/metadata.json")) if os.path.exists(f"{self.cache_dir}/metadata.json") else {} | |
| def _save_metadata(self): | |
| with open(f"{self.cache_dir}/metadata.json", 'w') as f: | |
| json.dump(self.metadata, f) | |
| async def create_audio(self, text, voice='en-US-AriaNeural'): | |
| cache_key = hashlib.md5(f"{text}:{voice}".encode()).hexdigest() | |
| cache_path = f"{self.cache_dir}/{cache_key}.mp3" | |
| if cache_key in self.metadata and os.path.exists(cache_path): | |
| return cache_path | |
| text = clean_text_for_tts(text) | |
| if not text: | |
| return None | |
| communicate = edge_tts.Communicate(text, voice) | |
| await communicate.save(cache_path) | |
| self.metadata[cache_key] = {'timestamp': datetime.now().isoformat(), 'text_length': len(text), 'voice': voice} | |
| self._save_metadata() | |
| return cache_path | |
| def process_pdf(pdf_file, max_pages, voice, audio_processor): | |
| reader = PdfReader(pdf_file) | |
| total_pages = min(len(reader.pages), max_pages) | |
| texts, audios = [], {} | |
| async def process_page(i, text): | |
| audio_path = await audio_processor.create_audio(text, voice) | |
| if audio_path: | |
| audios[i] = audio_path | |
| for i in range(total_pages): | |
| text = reader.pages[i].extract_text() | |
| texts.append(text) | |
| threading.Thread(target=lambda: asyncio.run(process_page(i, text))).start() | |
| return texts, audios, total_pages | |
| def parse_arxiv_refs(ref_text, query): | |
| if not ref_text: | |
| return [] | |
| papers = [] | |
| current = {} | |
| for line in ref_text.split('\n'): | |
| if line.count('|') == 2: | |
| if current: | |
| papers.append(current) | |
| date, title, *_ = line.strip('* ').split('|') | |
| url = re.search(r'(https://arxiv.org/\S+)', line).group(1) if re.search(r'(https://arxiv.org/\S+)', line) else f"paper_{len(papers)}" | |
| current = {'date': date, 'title': title, 'url': url, 'authors': '', 'summary': '', 'full_audio': None, 'download_base64': '', 'query': query} | |
| elif current: | |
| if not current['authors']: | |
| current['authors'] = line.strip('* ') | |
| else: | |
| current['summary'] += ' ' + line.strip() if current['summary'] else line.strip() | |
| if current: | |
| papers.append(current) | |
| return papers[:20] | |
| def generate_5min_feature_markdown(paper): | |
| title, summary, authors, date, url = paper['title'], paper['summary'], paper['authors'], paper['date'], paper['url'] | |
| pdf_url = url.replace("abs", "pdf") + (".pdf" if not url.endswith(".pdf") else "") | |
| wct, sw = len(title.split()), len(summary.split()) | |
| terms = get_high_info_terms(summary, 15) | |
| rouge = round((len(terms) / max(sw, 1)) * 100, 2) | |
| mermaid = "```mermaid\nflowchart TD\n" + "\n".join(f' T{i+1}["{terms[i]}"] --> T{i+2}["{terms[i+1]}"]' for i in range(len(terms)-1)) + "\n```" | |
| return f""" | |
| ## ๐ {title} | |
| **Authors:** {authors} | |
| **Date:** {date} | |
| **Words:** Title: {wct}, Summary: {sw} | |
| **Links:** [Abstract]({url}) | [PDF]({pdf_url}) | |
| **Terms:** {', '.join(terms)} | |
| **ROUGE:** {rouge}% | |
| ### ๐ค TTF Read Aloud | |
| - **Title:** {title} | |
| - **Terms:** {', '.join(terms)} | |
| - **ROUGE:** {rouge}% | |
| #### Concepts Graph | |
| {mermaid} | |
| --- | |
| """ | |
| async def create_paper_audio_files(papers, query): | |
| for p in papers: | |
| audio_text = clean_text_for_tts(f"{p['title']} by {p['authors']}. {p['summary']}") | |
| p['full_audio'], _ = await async_edge_tts_generate(audio_text, st.session_state['tts_voice'], p['authors']) | |
| if p['full_audio']: | |
| p['download_base64'] = get_download_link(p['full_audio']) | |
| def save_vote(file, item, user_hash): | |
| timestamp = datetime.now().strftime("%Y-%m-%d %H:%M:%S") | |
| entry = f"[{timestamp}] {user_hash} voted for {item}" | |
| try: | |
| with open(file, 'a') as f: | |
| f.write(f"{entry}\n") | |
| with open(HISTORY_FILE, 'a') as f: | |
| f.write(f"- {timestamp} - User {user_hash} voted for {item}\n") | |
| return True | |
| except Exception as e: | |
| print(f"Vote save flop: {e}") | |
| return False | |
| def load_votes(file): | |
| if not os.path.exists(file): | |
| with open(file, 'w') as f: | |
| f.write("# Vote Tally\n\nNo votes yet - get clicking! ๐ฑ๏ธ\n") | |
| try: | |
| with open(file, 'r') as f: | |
| lines = f.read().strip().split('\n') | |
| votes = {} | |
| for line in lines[2:]: | |
| if line.strip() and 'voted for' in line: | |
| item = line.split('voted for ')[1] | |
| votes[item] = votes.get(item, 0) + 1 | |
| return votes | |
| except Exception as e: | |
| print(f"Vote load oopsie: {e}") | |
| return {} | |
| def generate_user_hash(): | |
| if 'user_hash' not in st.session_state: | |
| session_id = str(random.getrandbits(128)) | |
| hash_object = hashlib.md5(session_id.encode()) | |
| st.session_state['user_hash'] = hash_object.hexdigest()[:8] | |
| return st.session_state['user_hash'] | |
| async def save_pasted_image(image, username, prompt=""): | |
| img_hash = hashlib.md5(image.tobytes()).hexdigest()[:8] | |
| if img_hash in st.session_state.image_hashes: | |
| return None | |
| context = prompt if prompt else st.session_state.get('last_message', "pasted_image") | |
| timestamp = format_timestamp_prefix(username) | |
| filename = f"{timestamp}-{clean_text_for_filename(context)}-{img_hash}.png" | |
| filepath = filename | |
| try: | |
| image.save(filepath, "PNG") | |
| st.session_state.image_hashes.add(img_hash) | |
| await save_chat_entry(username, f"Pasted image saved: {filepath}", FUN_USERNAMES.get(username, "en-US-AriaNeural")) | |
| return filepath | |
| except Exception as e: | |
| st.error(f"Failed to save image: {e}") | |
| return None | |
| def create_zip_of_files(files, prefix="All", query="latest"): | |
| if not files: | |
| return None | |
| terms = get_high_info_terms(" ".join([open(f, 'r', encoding='utf-8').read() if f.endswith('.md') else os.path.splitext(os.path.basename(f))[0].replace('_', ' ') for f in files] + [query]), 5) | |
| zip_name = f"{prefix}_{format_timestamp_prefix()}_{'-'.join(terms)[:20]}.zip" | |
| with zipfile.ZipFile(zip_name, 'w') as z: | |
| [z.write(f) for f in files] | |
| return zip_name | |
| def delete_files(file_type, exclude_files=["README.md"]): | |
| files = glob.glob(f"*.{file_type}") | |
| if file_type == "md": | |
| files = [f for f in files if os.path.basename(f) not in exclude_files] | |
| for f in files: | |
| try: | |
| os.remove(f) | |
| st.session_state['mp3_files'] = {k: v for k, v in st.session_state['mp3_files'].items() if not k.endswith(f".{file_type}")} | |
| except Exception as e: | |
| st.error(f"Failed to delete {f}: {e}") | |
| if file_type in ["md", "mp3", "png", "mp4"]: | |
| st.session_state['download_link_cache'] = {} | |
| def paste_image_component(): | |
| with st.form(key="paste_form"): | |
| paste_input = st.text_area("Paste Image Data Here", key="paste_input", height=100) | |
| submit_button = st.form_submit_button("Paste Image ๐") | |
| if submit_button and paste_input: | |
| try: | |
| if paste_input.startswith('data:image'): | |
| mime_type = paste_input.split(';')[0].split(':')[1] | |
| base64_str = paste_input.split(',')[1] | |
| img_bytes = base64.b64decode(base64_str) | |
| img = Image.open(io.BytesIO(img_bytes)) | |
| st.image(img, caption=f"Pasted Image ({mime_type.split('/')[1].upper()})", use_column_width=True) | |
| return img, mime_type.split('/')[1] | |
| else: | |
| st.warning("Pasted data is not a recognized image format.") | |
| return None, None | |
| except Exception as e: | |
| st.error(f"Error decoding pasted image: {e}") | |
| return None, None | |
| return None, None | |
| def create_pdf_tab(default_markdown): | |
| font_files = glob.glob("*.ttf") | |
| if not font_files: | |
| st.error("No .ttf font files found in the current directory. Please add some, e.g., NotoEmoji-Bold.ttf and DejaVuSans.ttf.") | |
| return | |
| available_fonts = {os.path.splitext(os.path.basename(f))[0]: f for f in font_files} | |
| md_files = [f for f in glob.glob("*.md") if os.path.basename(f) != "README.md"] | |
| md_options = [os.path.splitext(os.path.basename(f))[0] for f in md_files] | |
| with st.sidebar: | |
| selected_md = st.selectbox("Select Markdown File", options=md_options, index=0 if md_options else -1) | |
| selected_font_name = st.selectbox("Select Emoji Font", options=list(available_fonts.keys()), index=0 if "NotoEmoji-Bold" in available_fonts else 0) | |
| selected_font_path = available_fonts[selected_font_name] | |
| base_font_size = st.slider("Font Size (points)", min_value=6, max_value=16, value=8, step=1) | |
| plain_text_mode = st.checkbox("Render as Plain Text (Preserve Bold Only)", value=False) | |
| auto_bold_numbers = st.checkbox("Auto-Bold Numbered Lines", value=False) | |
| enlarge_font_size = st.checkbox("Enlarge Font Size for Numbered Lines", value=True) | |
| num_columns = st.selectbox("Number of Columns", options=[1, 2, 3, 4, 5, 6], index=3) | |
| if 'markdown_content' not in st.session_state or not md_options: | |
| st.session_state.markdown_content = default_markdown | |
| if md_options and selected_md: | |
| with open(f"{selected_md}.md", "r", encoding="utf-8") as f: | |
| st.session_state.markdown_content = f.read() | |
| edited_markdown = st.text_area("Modify the markdown content below:", value=st.session_state.markdown_content, height=300, key=f"markdown_{selected_md}_{selected_font_name}_{num_columns}") | |
| if st.button("Update PDF"): | |
| st.session_state.markdown_content = edited_markdown | |
| with open(f"{selected_md}.md", "w", encoding="utf-8") as f: | |
| f.write(edited_markdown) | |
| st.rerun() | |
| st.download_button(label="Save Markdown", data=st.session_state.markdown_content, file_name=f"{selected_md}.md", mime="text/markdown") | |
| st.subheader("Voice Settings") | |
| voice = st.selectbox("Select Voice", options=EDGE_TTS_VOICES, index=0) | |
| if st.button("Generate MP3"): | |
| audio_file, _ = asyncio.run(async_edge_tts_generate(edited_markdown, voice, "System")) | |
| if audio_file: | |
| st.audio(audio_file) | |
| st.markdown(get_download_link(audio_file), unsafe_allow_html=True) | |
| if not md_options: | |
| st.warning("No .md files found in the directory (excluding README.md). Using default content.") | |
| return | |
| try: | |
| pdfmetrics.registerFont(TTFont(selected_font_name, selected_font_path)) | |
| pdfmetrics.registerFont(TTFont("DejaVuSans", "DejaVuSans.ttf")) | |
| except Exception as e: | |
| st.error(f"Failed to register fonts: {e}. Ensure both {selected_font_name}.ttf and DejaVuSans.ttf are in the directory.") | |
| return | |
| def apply_emoji_font(text, emoji_font): | |
| emoji_pattern = re.compile( | |
| r"([\U0001F300-\U0001F5FF" | |
| r"\U0001F600-\U0001F64F" | |
| r"\U0001F680-\U0001F6FF" | |
| r"\U0001F700-\U0001F77F" | |
| r"\U0001F780-\U0001F7FF" | |
| r"\U0001F800-\U0001F8FF" | |
| r"\U0001F900-\U0001F9FF" | |
| r"\U0001FA00-\U0001FA6F" | |
| r"\U0001FA70-\U0001FAFF" | |
| r"\u2600-\u26FF" | |
| r"\u2700-\u27BF]+)" | |
| ) | |
| def replace_emoji(match): | |
| emoji = match.group(1) | |
| emoji = unicodedata.normalize('NFC', emoji) | |
| return f'<font face="{emoji_font}">{emoji}</font>' | |
| segments = [] | |
| last_pos = 0 | |
| for match in emoji_pattern.finditer(text): | |
| start, end = match.span() | |
| if last_pos < start: | |
| segments.append(f'<font face="DejaVuSans">{text[last_pos:start]}</font>') | |
| segments.append(replace_emoji(match)) | |
| last_pos = end | |
| if last_pos < len(text): | |
| segments.append(f'<font face="DejaVuSans">{text[last_pos:]}</font>') | |
| return ''.join(segments) | |
| def markdown_to_pdf_content(markdown_text, plain_text_mode, auto_bold_numbers): | |
| lines = markdown_text.strip().split('\n') | |
| pdf_content = [] | |
| number_pattern = re.compile(r'^\d+\.\s') | |
| if plain_text_mode: | |
| for line in lines: | |
| line = line.strip() | |
| if not line or line.startswith('# '): | |
| continue | |
| bold_pattern = re.compile(r'\*\*(.*?)\*\*') | |
| line = bold_pattern.sub(r'<b>\1</b>', line) | |
| line = re.sub(r'\*\*', '', line) | |
| pdf_content.append(line) | |
| else: | |
| for line in lines: | |
| line = line.strip() | |
| if not line or line.startswith('# '): | |
| continue | |
| bold_pattern = re.compile(r'\*\*(.*?)\*\*') | |
| if bold_pattern.search(line): | |
| line = bold_pattern.sub(r'<b>\1</b>', line) | |
| line = re.sub(r'\*\*', '', line) | |
| if line.startswith('## ') or line.startswith('### '): | |
| text = line.replace('## ', '').replace('### ', '').strip() | |
| pdf_content.append(f"<b>{text}</b>") | |
| elif auto_bold_numbers and number_pattern.match(line): | |
| pdf_content.append(f"<b>{line}</b>") | |
| else: | |
| pdf_content.append(line.strip()) | |
| total_lines = len(pdf_content) | |
| return pdf_content, total_lines | |
| def create_pdf(markdown_text, base_font_size, plain_text_mode, num_columns, auto_bold_numbers, enlarge_font_size): | |
| buffer = io.BytesIO() | |
| page_width = A4[0] * 2 | |
| page_height = A4[1] | |
| doc = SimpleDocTemplate(buffer, pagesize=(page_width, page_height), leftMargin=36, rightMargin=36, topMargin=36, bottomMargin=36) | |
| styles = getSampleStyleSheet() | |
| story = [] | |
| spacer_height = 10 | |
| section_spacer_height = 15 | |
| pdf_content, total_lines = markdown_to_pdf_content(markdown_text, plain_text_mode, auto_bold_numbers) | |
| item_font_size = base_font_size | |
| section_font_size = base_font_size * 1.1 | |
| numbered_font_size = base_font_size + 1 if enlarge_font_size else base_font_size | |
| section_style = ParagraphStyle( | |
| 'SectionStyle', parent=styles['Heading2'], fontName="DejaVuSans", | |
| textColor=colors.darkblue, fontSize=section_font_size, leading=section_font_size * 1.2, spaceAfter=2 | |
| ) | |
| item_style = ParagraphStyle( | |
| 'ItemStyle', parent=styles['Normal'], fontName="DejaVuSans", | |
| fontSize=item_font_size, leading=item_font_size * 1.15, spaceAfter=1 | |
| ) | |
| numbered_style = ParagraphStyle( | |
| 'NumberedStyle', parent=styles['Normal'], fontName="DejaVuSans", | |
| fontSize=numbered_font_size, leading=numbered_font_size * 1.15, spaceAfter=1 | |
| ) | |
| story.append(Spacer(1, spacer_height)) | |
| columns = [[] for _ in range(num_columns)] | |
| lines_per_column = total_lines / num_columns if num_columns > 0 else total_lines | |
| current_line_count = 0 | |
| current_column = 0 | |
| number_pattern = re.compile(r'^\d+\.\s') | |
| for i, item in enumerate(pdf_content): | |
| if i > 0 and number_pattern.match(item.replace('<b>', '').replace('</b>', '')): | |
| columns[current_column].append(Spacer(1, section_spacer_height)) | |
| if current_line_count >= lines_per_column and current_column < num_columns - 1: | |
| current_column += 1 | |
| current_line_count = 0 | |
| columns[current_column].append(item) | |
| current_line_count += 1 | |
| column_cells = [[] for _ in range(num_columns)] | |
| for col_idx, column in enumerate(columns): | |
| for item in column: | |
| if isinstance(item, Spacer): | |
| column_cells[col_idx].append(item) | |
| elif isinstance(item, str) and item.startswith('<b>'): | |
| text = item.replace('<b>', '').replace('</b>', '') | |
| column_cells[col_idx].append(Paragraph(apply_emoji_font(text, selected_font_name), section_style)) | |
| elif number_pattern.match(item): | |
| column_cells[col_idx].append(Paragraph(apply_emoji_font(item, selected_font_name), numbered_style)) | |
| else: | |
| column_cells[col_idx].append(Paragraph(apply_emoji_font(item, selected_font_name), item_style)) | |
| max_cells = max(len(cells) for cells in column_cells) if column_cells else 0 | |
| for cells in column_cells: | |
| cells.extend([Paragraph("", item_style)] * (max_cells - len(cells))) | |
| col_width = (page_width - 72) / num_columns if num_columns > 0 else page_width - 72 | |
| table_data = list(zip(*column_cells)) if column_cells else [[]] | |
| table = Table(table_data, colWidths=[col_width] * num_columns, hAlign='CENTER') | |
| table.setStyle(TableStyle([ | |
| ('VALIGN', (0, 0), (-1, -1), 'TOP'), ('ALIGN', (0, 0), (-1, -1), 'LEFT'), | |
| ('BACKGROUND', (0, 0), (-1, -1), colors.white), ('GRID', (0, 0), (-1, -1), 0, colors.white), | |
| ('LINEAFTER', (0, 0), (num_columns-1, -1), 0.5, colors.grey), | |
| ('LEFTPADDING', (0, 0), (-1, -1), 2), ('RIGHTPADDING', (0, 0), (-1, -1), 2), | |
| ('TOPPADDING', (0, 0), (-1, -1), 1), ('BOTTOMPADDING', (0, 0), (-1, -1), 1), | |
| ])) | |
| story.append(table) | |
| doc.build(story) | |
| buffer.seek(0) | |
| return buffer.getvalue() | |
| def pdf_to_image(pdf_bytes): | |
| try: | |
| doc = fitz.open(stream=pdf_bytes, filetype="pdf") | |
| images = [] | |
| for page in doc: | |
| pix = page.get_pixmap(matrix=fitz.Matrix(2.0, 2.0)) | |
| img = Image.frombytes("RGB", [pix.width, pix.height], pix.samples) | |
| images.append(img) | |
| doc.close() | |
| return images | |
| except Exception as e: | |
| st.error(f"Failed to render PDF preview: {e}") | |
| return None | |
| with st.spinner("Generating PDF..."): | |
| pdf_bytes = create_pdf(st.session_state.markdown_content, base_font_size, plain_text_mode, num_columns, auto_bold_numbers, enlarge_font_size) | |
| with st.container(): | |
| pdf_images = pdf_to_image(pdf_bytes) | |
| if pdf_images: | |
| for img in pdf_images: | |
| st.image(img, use_container_width=True) | |
| else: | |
| st.info("Download the PDF to view it locally.") | |
| with st.sidebar: | |
| st.download_button(label="Download PDF", data=pdf_bytes, file_name="deities_guide.pdf", mime="application/pdf") | |
| def main(): | |
| init_session_state() | |
| load_mp3_viewer() | |
| saved_username = load_username() | |
| if saved_username and saved_username in FUN_USERNAMES: | |
| st.session_state.username = saved_username | |
| if not st.session_state.username: | |
| available = [n for n in FUN_USERNAMES if not any(f"{n} has joined" in l for l in asyncio.run(load_chat()))] | |
| st.session_state.username = random.choice(available or list(FUN_USERNAMES.keys())) | |
| st.session_state.tts_voice = FUN_USERNAMES[st.session_state.username] | |
| asyncio.run(save_chat_entry("System ๐", f"{st.session_state.username} has joined {START_ROOM}!", "en-US-AriaNeural")) | |
| save_username(st.session_state.username) | |
| st.title(f"{Site_Name} for {st.session_state.username}") | |
| update_marquee_settings_ui() | |
| chat_text = get_chat_text_only() | |
| display_marquee(f"๐ Welcome to {START_ROOM} | ๐ค {st.session_state.username} | Chat: {chat_text}", st.session_state['marquee_settings'], "welcome") | |
| mycomponent = components.declare_component("mycomponent", path="mycomponent") | |
| val = mycomponent(my_input_value="", key=f"speech_{st.session_state.get('speech_processed', False)}") | |
| if val and val != st.session_state.last_transcript: | |
| val_stripped = val.strip().replace('\n', ' ') | |
| if val_stripped: | |
| voice = FUN_USERNAMES.get(st.session_state.username, "en-US-AriaNeural") | |
| st.session_state['speech_processed'] = True | |
| md_file, audio_file = asyncio.run(save_chat_entry(st.session_state.username, val_stripped, voice)) | |
| if audio_file: | |
| play_and_download_audio(audio_file) | |
| st.rerun() | |
| tab_main = st.radio("Action:", ["๐ค Chat & Voice", "๐ ArXiv", "๐ PDF to Audio", "๐ PDF Output"], horizontal=True, key="tab_main") | |
| st.checkbox("Search ArXiv", key="use_arxiv") | |
| st.checkbox("ArXiv Audio", key="use_arxiv_audio") | |
| st.checkbox("Autosend Chat", key="autosend") | |
| st.checkbox("Autosearch ArXiv", key="autosearch") | |
| if tab_main == "๐ค Chat & Voice": | |
| st.subheader(f"{START_ROOM} Chat ๐ฌ") | |
| chat_content = asyncio.run(load_chat()) | |
| chat_container = st.container() | |
| with chat_container: | |
| numbered_content = "\n".join(f"{i+1}. {line}" for i, line in enumerate(chat_content)) | |
| st.code(numbered_content, language="python") | |
| message = st.text_input(f"Message as {st.session_state.username}", key="message_input") | |
| col_paste, col_upload = st.columns(2) | |
| with col_paste: | |
| pasted_image, img_type = paste_image_component() | |
| with col_upload: | |
| uploaded_files = st.file_uploader("Upload Files", accept_multiple_files=True, type=["mp3", "png", "mp4", "md"], key="file_upload") | |
| if pasted_image is not None: | |
| if st.session_state['paste_image_base64'] != base64.b64encode(pasted_image.tobytes()).decode('utf-8'): | |
| st.session_state['paste_image_base64'] = base64.b64encode(pasted_image.tobytes()).decode('utf-8') | |
| voice = FUN_USERNAMES.get(st.session_state.username, "en-USA-AriaNeural") | |
| image_prompt = st.text_input("Add a prompt for Claude (e.g., 'OCR this image')", key="image_prompt", value="") | |
| with st.spinner("Saving image..."): | |
| filename = asyncio.run(save_pasted_image(pasted_image, st.session_state.username, image_prompt)) | |
| if filename: | |
| st.success(f"Image saved as: {filename}") | |
| if image_prompt: | |
| with st.spinner("Processing with Claude..."): | |
| md_file_claude, audio_file_claude, claude_result = asyncio.run( | |
| perform_claude_search(image_prompt, st.session_state.username, pasted_image) | |
| ) | |
| if audio_file_claude: | |
| play_and_download_audio(audio_file_claude) | |
| if claude_result: | |
| with st.spinner("Searching ArXiv..."): | |
| md_file_arxiv, audio_file_arxiv = asyncio.run( | |
| perform_arxiv_search(image_prompt, st.session_state.username, claude_result) | |
| ) | |
| if audio_file_arxiv: | |
| play_and_download_audio(audio_file_arxiv) | |
| st.session_state.pasted_image_data = None | |
| st.session_state['paste_image_base64'] = "" | |
| st.session_state.timer_start = time.time() | |
| save_username(st.session_state.username) | |
| st.rerun() | |
| if uploaded_files: | |
| for uploaded_file in uploaded_files: | |
| file_type = uploaded_file.name.split('.')[-1].lower() | |
| if file_type in ["mp3", "png", "mp4", "md"]: | |
| timestamp = format_timestamp_prefix(st.session_state.username) | |
| filename = f"{timestamp}-{clean_text_for_filename(uploaded_file.name)}" | |
| with open(filename, "wb") as f: | |
| f.write(uploaded_file.getbuffer()) | |
| st.success(f"Uploaded {file_type.upper()} as: {filename}") | |
| if file_type == "png": | |
| img = Image.open(filename) | |
| st.image(img, caption=f"Uploaded Image: {filename}", use_column_width=True) | |
| elif file_type == "mp3": | |
| st.audio(filename) | |
| elif file_type == "mp4": | |
| st.video(filename) | |
| elif file_type == "md": | |
| with open(filename, 'r', encoding='utf-8') as f: | |
| st.markdown(f.read()) | |
| asyncio.run(save_chat_entry(st.session_state.username, f"Uploaded {file_type.upper()}: {filename}", FUN_USERNAMES.get(st.session_state.username, "en-US-AriaNeural"))) | |
| st.session_state.timer_start = time.time() | |
| save_username(st.session_state.username) | |
| st.rerun() | |
| if (message and message != st.session_state.last_message) or (st.session_state.pasted_image_data and not st.session_state['paste_image_base64']): | |
| st.session_state.last_message = message | |
| col_send, col_claude, col_arxiv = st.columns([1, 1, 1]) | |
| with col_send: | |
| if st.session_state.autosend or st.button("Send ๐", key="send_button"): | |
| voice = FUN_USERNAMES.get(st.session_state.username, "en-US-AriaNeural") | |
| if message.strip(): | |
| md_file, audio_file = asyncio.run(save_chat_entry(st.session_state.username, message, voice, True)) | |
| if audio_file: | |
| play_and_download_audio(audio_file) | |
| if st.session_state.pasted_image_data: | |
| asyncio.run(save_chat_entry(st.session_state.username, f"Pasted image: {st.session_state.pasted_image_data}", voice)) | |
| st.session_state.pasted_image_data = None | |
| st.session_state.timer_start = time.time() | |
| save_username(st.session_state.username) | |
| st.rerun() | |
| with col_claude: | |
| if st.button("๐ง Claude", key="claude_button"): | |
| voice = FUN_USERNAMES.get(st.session_state.username, "en-US-AriaNeural") | |
| if message.strip(): | |
| md_file, audio_file, _ = asyncio.run(perform_claude_search(message, st.session_state.username)) | |
| if audio_file: | |
| play_and_download_audio(audio_file) | |
| st.session_state.timer_start = time.time() | |
| save_username(st.session_state.username) | |
| st.rerun() | |
| with col_arxiv: | |
| if st.button("๐ ArXiv", key="arxiv_button"): | |
| voice = FUN_USERNAMES.get(st.session_state.username, "en-US-AriaNeural") | |
| if message.strip(): | |
| md_file, audio_file = asyncio.run(perform_arxiv_search(message, st.session_state.username)) | |
| if audio_file: | |
| play_and_download_audio(audio_file) | |
| st.session_state.timer_start = time.time() | |
| save_username(st.session_state.username) | |
| st.rerun() | |
| elif tab_main == "๐ ArXiv": | |
| st.subheader("๐ Query ArXiv") | |
| q = st.text_input("๐ Query:", key="arxiv_query") | |
| if q and q != st.session_state.last_query: | |
| st.session_state.last_query = q | |
| if st.session_state.autosearch or st.button("๐ Run", key="arxiv_run"): | |
| result, papers = asyncio.run(perform_ai_lookup(q, useArxiv=st.session_state['use_arxiv'], useArxivAudio=st.session_state['use_arxiv_audio'])) | |
| st.markdown(f"### Query: {q}") | |
| for i, p in enumerate(papers, 1): | |
| expander_label = f"{p['title']} | [arXiv Link]({p['url']})" | |
| with st.expander(expander_label): | |
| with open(p['md_file'], 'r', encoding='utf-8') as f: | |
| content = f.read() | |
| numbered_content = "\n".join(f"{j+1}. {line}" for j, line in enumerate(content.split('\n'))) | |
| st.code(numbered_content, language="python") | |
| elif tab_main == "๐ PDF to Audio": | |
| audio_processor = AudioProcessor() | |
| pdf_file = st.file_uploader("Choose PDF", "pdf", key="pdf_upload") | |
| max_pages = st.slider('Pages', 1, 100, 10, key="pdf_pages") | |
| if pdf_file: | |
| with st.spinner('Processing...'): | |
| texts, audios, total = process_pdf(pdf_file, max_pages, st.session_state['tts_voice'], audio_processor) | |
| for i, text in enumerate(texts): | |
| with st.expander(f"Page {i+1}"): | |
| st.markdown(text) | |
| while i not in audios: | |
| time.sleep(0.1) | |
| if audios.get(i): | |
| st.audio(audios[i]) | |
| st.markdown(get_download_link(audios[i], "mp3"), unsafe_allow_html=True) | |
| voice = FUN_USERNAMES.get(st.session_state.username, "en-US-AriaNeural") | |
| asyncio.run(save_chat_entry(st.session_state.username, f"PDF Page {i+1} converted to audio: {audios[i]}", voice)) | |
| elif tab_main == "๐ PDF Output": | |
| create_pdf_tab(default_markdown) | |
| st.header("๐ธ Media Gallery") | |
| all_files = sorted(glob.glob("*.md") + glob.glob("*.mp3") + glob.glob("*.png") + glob.glob("*.mp4"), key=os.path.getmtime) | |
| md_files = [f for f in all_files if f.endswith('.md') and os.path.basename(f) != "README.md"] | |
| mp3_files = [f for f in all_files if f.endswith('.mp3')] | |
| png_files = [f for f in all_files if f.endswith('.png')] | |
| mp4_files = [f for f in all_files if f.endswith('.mp4')] | |
| st.subheader("All Submitted Text") | |
| all_md_content = concatenate_markdown_files() | |
| with st.expander("View All Markdown Content"): | |
| st.markdown(all_md_content) | |
| st.subheader("๐ต Audio (MP3)") | |
| for filename, (num, mp3) in sorted(st.session_state['mp3_files'].items(), key=lambda x: x[1][0]): | |
| with st.expander(f"{num}. {os.path.basename(mp3)}"): | |
| st.audio(mp3) | |
| st.markdown(get_download_link(mp3, "mp3"), unsafe_allow_html=True) | |
| st.subheader("๐ผ๏ธ Images (PNG)") | |
| for png in sorted(png_files, key=os.path.getmtime): | |
| with st.expander(os.path.basename(png)): | |
| st.image(png, use_container_width=True) | |
| st.markdown(get_download_link(png, "png"), unsafe_allow_html=True) | |
| st.subheader("๐ฅ Videos (MP4)") | |
| for mp4 in sorted(mp4_files, key=os.path.getmtime): | |
| with st.expander(os.path.basename(mp4)): | |
| st.video(mp4) | |
| st.markdown(get_download_link(mp4, "mp4"), unsafe_allow_html=True) | |
| st.sidebar.subheader("Voice Settings") | |
| new_username = st.sidebar.selectbox("Change Name/Voice", list(FUN_USERNAMES.keys()), index=list(FUN_USERNAMES.keys()).index(st.session_state.username), key="username_select") | |
| if new_username != st.session_state.username: | |
| asyncio.run(save_chat_entry("System ๐", f"{st.session_state.username} changed to {new_username}", "en-US-AriaNeural")) | |
| st.session_state.username, st.session_state.tts_voice = new_username, FUN_USERNAMES[new_username] | |
| st.session_state.timer_start = time.time() | |
| save_username(st.session_state.username) | |
| st.rerun() | |
| st.sidebar.markdown("### ๐ฌ Chat Dialog") | |
| chat_content = asyncio.run(load_chat()) | |
| with st.sidebar.expander("Chat History"): | |
| numbered_content = "\n".join(f"{i+1}. {line}" for i, line in enumerate(chat_content)) | |
| st.code(numbered_content, language="python") | |
| st.sidebar.markdown("### ๐ฌ Chat Text Only") | |
| chat_text_only = get_chat_text_only() | |
| with st.sidebar.expander("Text Only History"): | |
| numbered_text = "\n".join(f"{i+1}. {line}" for i, line in enumerate(chat_text_only.split('\n'))) | |
| st.code(numbered_text, language="python") | |
| st.sidebar.subheader("Vote Totals") | |
| chat_votes = load_votes(QUOTE_VOTES_FILE) | |
| image_votes = load_votes(IMAGE_VOTES_FILE) | |
| for item, count in chat_votes.items(): | |
| st.sidebar.write(f"{item}: {count} votes") | |
| for image, count in image_votes.items(): | |
| st.sidebar.write(f"{image}: {count} votes") | |
| st.sidebar.markdown("### ๐ File History") | |
| for f in all_files[:10]: | |
| st.sidebar.write(f"{FILE_EMOJIS.get(f.split('.')[-1], '๐')} {os.path.basename(f)}") | |
| st.sidebar.subheader("๐ฆ Zip & Delete") | |
| col_zip, col_del = st.sidebar.columns(2) | |
| with col_zip: | |
| if st.button("โฌ๏ธ Zip All", key="zip_all"): | |
| zip_name = create_zip_of_files(all_files, "All") | |
| if zip_name: | |
| st.session_state['download_link_cache'] = {} | |
| if st.button("โฌ๏ธ Zip All MD", key="zip_md"): | |
| zip_name = create_zip_of_files(md_files, "MD") | |
| if zip_name: | |
| st.session_state['download_link_cache'] = {} | |
| if st.button("โฌ๏ธ Zip All MP3", key="zip_mp3"): | |
| zip_name = create_zip_of_files(mp3_files, "MP3") | |
| if zip_name: | |
| st.session_state['download_link_cache'] = {} | |
| if st.button("โฌ๏ธ Zip All PNG", key="zip_png"): | |
| zip_name = create_zip_of_files(png_files, "PNG") | |
| if zip_name: | |
| st.session_state['download_link_cache'] = {} | |
| if st.button("โฌ๏ธ Zip All MP4", key="zip_mp4"): | |
| zip_name = create_zip_of_files(mp4_files, "MP4") | |
| if zip_name: | |
| st.session_state['download_link_cache'] = {} | |
| with col_del: | |
| if st.button("๐๏ธ Del All", key="del_all"): | |
| for ft in ["md", "mp3", "png", "mp4"]: | |
| delete_files(ft) | |
| st.rerun() | |
| if st.button("๐๏ธ Del All MD", key="del_md"): | |
| delete_files("md") | |
| st.rerun() | |
| if st.button("๐๏ธ Del All MP3", key="del_mp3"): | |
| delete_files("mp3") | |
| st.rerun() | |
| if st.button("๐๏ธ Del All PNG", key="del_png"): | |
| delete_files("png") | |
| st.rerun() | |
| if st.button("๐๏ธ Del All MP4", key="del_mp4"): | |
| delete_files("mp4") | |
| st.rerun() | |
| if st.button("๐๏ธ Del All Zip", key="del_zip"): | |
| delete_files("zip", exclude_files=[]) | |
| st.rerun() | |
| zip_files = sorted(glob.glob("*.zip"), key=os.path.getmtime, reverse=True) | |
| for zip_file in zip_files: | |
| st.sidebar.markdown(get_download_link(zip_file, "zip"), unsafe_allow_html=True) | |
| st.sidebar.subheader("Set Refresh Rate โณ") | |
| st.session_state['auto_refresh'] = st.sidebar.radio("Auto Refresh", ["On", "Off"], index=0 if st.session_state['auto_refresh'] else 1) == "On" | |
| st.markdown(""" | |
| <style> | |
| .timer { | |
| font-size: 24px; | |
| color: #ffcc00; | |
| text-align: center; | |
| animation: pulse 1s infinite; | |
| } | |
| @keyframes pulse { | |
| 0% { transform: scale(1); } | |
| 50% { transform: scale(1.1); } | |
| 100% { transform: scale(1); } | |
| } | |
| </style> | |
| """, unsafe_allow_html=True) | |
| refresh_rate = st.sidebar.slider("Refresh Rate (seconds)", min_value=1, max_value=300, value=st.session_state.refresh_rate, step=1) | |
| if refresh_rate != st.session_state.refresh_rate: | |
| st.session_state.refresh_rate = refresh_rate | |
| st.session_state.timer_start = time.time() | |
| save_username(st.session_state.username) | |
| col1, col2, col3 = st.sidebar.columns(3) | |
| with col1: | |
| if st.button("๐ Small (1s)"): | |
| st.session_state.refresh_rate = 1 | |
| st.session_state.timer_start = time.time() | |
| save_username(st.session_state.username) | |
| with col2: | |
| if st.button("๐ข Medium (10s)"): | |
| st.session_state.refresh_rate = 10 | |
| st.session_state.timer_start = time.time() | |
| save_username(st.session_state.username) | |
| with col3: | |
| if st.button("๐ Large (5m)"): | |
| st.session_state.refresh_rate = 300 | |
| st.session_state.timer_start = time.time() | |
| save_username(st.session_state.username) | |
| timer_placeholder = st.sidebar.empty() | |
| def update_timer(): | |
| start_time = st.session_state.timer_start | |
| remaining_time = max(0, int(st.session_state.refresh_rate - (time.time() - start_time))) | |
| timer_placeholder.markdown(f"<p class='timer'>โณ Next refresh in: {remaining_time} seconds</p>", unsafe_allow_html=True) | |
| if st.session_state['auto_refresh'] and remaining_time <= 0: | |
| st.session_state.timer_start = time.time() | |
| st.session_state.last_refresh = time.time() | |
| st.rerun() | |
| threading.Thread(target=lambda: [time.sleep(1) or update_timer() for _ in range(int(st.session_state.refresh_rate)+1)], daemon=True).start() | |
| update_timer() | |
| if not st.session_state.get('server_running', False) and not st.session_state.get('server_task', None): | |
| st.session_state.server_task = threading.Thread(target=start_websocket_server, daemon=True) | |
| st.session_state.server_task.start() | |
| default_markdown = """# Deities Guide: Mythology and Moral Lessons ๐โจ | |
| 1. ๐ **Introduction** | |
| - **Purpose**: Explore deities, spirits, saints, and beings with their epic stories and morals! ๐๐ | |
| - **Usage**: A guide for learning and storytelling across traditions. ๐ญโ๏ธ | |
| - **Themes**: Justice โ๏ธ, faith ๐, hubris ๐ค, redemption ๐, cosmic order ๐. | |
| 2. ๐ ๏ธ **Core Concepts of Divinity** | |
| - **Powers**: Creation ๐, omniscience ๐๏ธโ๐จ๏ธ, shapeshifting ๐ฆ across entities. | |
| - **Life Cycle**: Mortality ๐, immortality โจ, transitions like saints and avatars ๐. | |
| - **Communication**: Omens ๐ฉ๏ธ, visions ๐๏ธ, miracles โจ from gods and spirits. | |
| 3. โก **Standard Abilities** | |
| - **Creation**: Gods and spirits shape worlds, e.g., Allah ๐ and Vishnu ๐. | |
| - **Influence**: Saints and prophets intercede, like Muhammad ๐ and Paul โ๏ธ. | |
| - **Transformation**: Angels and avatars shift forms, e.g., Gabriel ๐ and Krishna ๐ฆ. | |
| - **Knowledge**: Foresight ๐ฎ or revelation ๐, as with the Holy Spirit ๐๏ธ and Brahma ๐ง . | |
| - **Judgment**: Divine authority ๐, e.g., Yahweh โ๏ธ and Yama ๐. | |
| 4. โณ **Mortality and Immortality** | |
| - **Gods**: Eternal โฐ, like Allah ๐ and Shiva ๐๏ธ. | |
| - **Spirits**: Realm-bound ๐ , e.g., jinn ๐ฅ and devas โจ. | |
| - **Saints/Prophets**: Mortal to divine ๐โก๏ธ๐, e.g., Moses ๐ and Rama ๐น. | |
| - **Beings**: Limbo states โ, like cherubim ๐ and rakshasas ๐น. | |
| - **Lessons**: Faith ๐ and duty โ๏ธ define transitions. | |
| 5. ๐ **Ascension and Signs** | |
| - **Paths**: Birth ๐ถ, deeds ๐ก๏ธ, revelation ๐, as with Jesus โ๏ธ and Arjuna ๐น. | |
| - **Signs**: Miracles โจ and prophecies ๐ฎ, like those in the Quran ๐ and Gita ๐. | |
| - **Morals**: Obedience ๐ง and devotion โค๏ธ shape destiny ๐. | |
| 6. ๐ฒ **Storytelling and Games** | |
| - **Portrayal**: Gods, spirits, and saints in narratives or RPGs ๐ฎ๐. | |
| - **Dynamics**: Clerics โช, imams ๐, and sadhus ๐ง serve higher powers. | |
| - **Balance**: Power ๐ช vs. personality ๐ for depth. | |
| 7. ๐ฎ **Dungeon Mastering Beings** | |
| - **Gods**: Epic scope ๐, e.g., Allah โจ and Vishnu ๐. | |
| - **Spirits**: Local influence ๐๏ธ, like jinn ๐ฅ and apsaras ๐. | |
| - **Saints**: Moral anchors โ, e.g., St. Francis ๐พ and Ali โ๏ธ. | |
| 8. ๐ **Devotee Relationships** | |
| - **Clerics**: Serve gods, e.g., Krishnaโs priests ๐ฆ. | |
| - **Mediums**: Channel spirits, like jinn whisperers ๐ฅ๐๏ธ. | |
| - **Faithful**: Venerate saints and prophets, e.g., Fatimaโs followers ๐น. | |
| 9. ๐ฆ **American Indian Traditions** | |
| - **Coyote, Raven, White Buffalo Woman**: Trickster kin ๐ฆ๐ฆ and wise mother ๐. | |
| - **Relation**: Siblings and guide teach balance โ๏ธ. | |
| - **Lesson**: Chaos ๐ช๏ธ breeds wisdom ๐ง . | |
| 10. โ๏ธ **Arthurian Legends** | |
| - **Merlin, Morgan le Fay, Arthur**: Mentor ๐ง, rival ๐งโโ๏ธ, son ๐. | |
| - **Relation**: Family tests loyalty ๐ค. | |
| - **Lesson**: Honor ๐ก๏ธ vs. betrayal ๐ก๏ธ. | |
| 11. ๐๏ธ **Babylonian Mythology** | |
| - **Marduk, Tiamat, Ishtar**: Son โ๏ธ, mother ๐, lover โค๏ธ. | |
| - **Relation**: Kinship drives order ๐ฐ. | |
| - **Lesson**: Power ๐ช reshapes chaos ๐ช๏ธ. | |
| 12. โ๏ธ **Christian Trinity** | |
| - **God (Yahweh), Jesus, Holy Spirit**: Father ๐, Son โ๏ธ, Spirit ๐๏ธ. | |
| - **Relation**: Divine family redeems ๐. | |
| - **Lesson**: Faith ๐ restores grace โจ. | |
| 13. ๐ **Christian Saints & Angels** | |
| - **St. Michael, Gabriel, Mary**: Warrior โ๏ธ, messenger ๐, mother ๐น. | |
| - **Relation**: Heavenly kin serve God ๐. | |
| - **Lesson**: Duty โ๏ธ upholds divine will ๐. | |
| 14. ๐ **Celtic Mythology** | |
| - **Lugh, Morrigan, Cernunnos**: Son โ๏ธ, mother ๐ฆ, father ๐ฆ. | |
| - **Relation**: Family governs cycles ๐. | |
| - **Lesson**: Courage ๐ช in fate ๐ฒ. | |
| 15. ๐ **Central American Traditions** | |
| - **Quetzalcoatl, Tezcatlipoca, Huitzilopochtli**: Brothers ๐๐ and war son โ๏ธ. | |
| - **Relation**: Sibling rivalry creates ๐. | |
| - **Lesson**: Sacrifice ๐ฉธ builds worlds ๐ฐ. | |
| 16. ๐ **Chinese Mythology** | |
| - **Jade Emperor, Nuwa, Sun Wukong**: Father ๐, mother ๐, rebel son ๐. | |
| - **Relation**: Family enforces harmony ๐ถ. | |
| - **Lesson**: Duty โ๏ธ curbs chaos ๐ช๏ธ. | |
| 17. ๐ **Cthulhu Mythos** | |
| - **Cthulhu, Nyarlathotep, Yog-Sothoth**: Elder kin ๐๐๏ธโ๐จ๏ธ๐. | |
| - **Relation**: Cosmic trio overwhelms ๐ฑ. | |
| - **Lesson**: Insignificance ๐ humbles ๐. | |
| 18. โฅ **Egyptian Mythology** | |
| - **Ra, Osiris, Isis**: Father โ๏ธ, son โฐ๏ธ, mother ๐. | |
| - **Relation**: Family ensures renewal ๐. | |
| - **Lesson**: Justice โ๏ธ prevails. | |
| 19. โ๏ธ **Finnish Mythology** | |
| - **Vรคinรคmรถinen, Louhi, Ukko**: Son ๐ถ, mother โ๏ธ, father โก. | |
| - **Relation**: Kinship tests wisdom ๐ง . | |
| - **Lesson**: Perseverance ๐๏ธ wins. | |
| 20. ๐๏ธ **Greek Mythology** | |
| - **Zeus, Hera, Athena**: Father โก, mother ๐, daughter ๐ฆ. | |
| - **Relation**: Family rules with tension โ๏ธ. | |
| - **Lesson**: Hubris ๐ค meets wisdom ๐ง . | |
| 21. ๐๏ธ **Hindu Trimurti** | |
| - **Brahma, Vishnu, Shiva**: Creator ๐, preserver ๐ก๏ธ, destroyer ๐ฅ. | |
| - **Relation**: Divine trio cycles existence ๐. | |
| - **Lesson**: Balance โ๏ธ sustains life ๐. | |
| 22. ๐บ **Hindu Avatars & Devis** | |
| - **Krishna, Rama, Durga**: Sons ๐ฆ๐น and fierce mother ๐ก๏ธ. | |
| - **Relation**: Avatars and goddess protect dharma โ๏ธ. | |
| - **Lesson**: Duty โ๏ธ defeats evil ๐น. | |
| 23. ๐ธ **Japanese Mythology** | |
| - **Amaterasu, Susanoo, Tsukuyomi**: Sister โ๏ธ, brothers ๐๐. | |
| - **Relation**: Siblings balance cosmos ๐. | |
| - **Lesson**: Harmony ๐ถ vs. chaos ๐ช๏ธ. | |
| 24. ๐ก๏ธ **Melnibonean Legends** | |
| - **Arioch, Xiombarg, Elric**: Lords ๐ and mortal son โ๏ธ. | |
| - **Relation**: Pact binds chaos ๐ช๏ธ. | |
| - **Lesson**: Power ๐ช corrupts ๐. | |
| 25. โช๏ธ **Muslim Divine & Messengers** | |
| - **Allah, Muhammad, Gabriel**: God ๐, prophet ๐, angel ๐. | |
| - **Relation**: Messenger reveals divine will ๐. | |
| - **Lesson**: Submission ๐ brings peace โฎ๏ธ. | |
| 26. ๐ป **Muslim Spirits & Kin** | |
| - **Jinn, Iblis, Khidr**: Spirits ๐ฅ๐ and guide ๐ฟ defy or aid. | |
| - **Relation**: Supernatural kin test faith ๐. | |
| - **Lesson**: Obedience ๐ง vs. rebellion ๐ก. | |
| 27. ๐ฐ **Nehwon Legends** | |
| - **Death, Ningauble, Sheelba**: Fateful trio ๐๐๏ธโ๐จ๏ธ๐ฟ. | |
| - **Relation**: Guides shape destiny ๐ฒ. | |
| - **Lesson**: Cunning ๐ง defies fate โฐ๏ธ. | |
| 28. ๐ง **Nonhuman Traditions** | |
| - **Corellon, Moradin, Gruumsh**: Elf ๐ง, dwarf โ๏ธ, orc ๐ก๏ธ fathers. | |
| - **Relation**: Rivals define purpose โ๏ธ. | |
| - **Lesson**: Community ๐ค endures. | |
| 29. แฑ **Norse Mythology** | |
| - **Odin, Frigg, Loki**: Father ๐๏ธ, mother ๐, trickster son ๐ฆ. | |
| - **Relation**: Family faces doom โก. | |
| - **Lesson**: Sacrifice ๐ฉธ costs. | |
| 30. ๐ฟ **Sumerian Mythology** | |
| - **Enki, Inanna, Anu**: Son ๐, daughter โค๏ธ, father ๐. | |
| - **Relation**: Kin wield knowledge ๐ง . | |
| - **Lesson**: Ambition ๐ shapes. | |
| 31. ๐ **Appendices** | |
| - **Planes**: Realms of gods, spirits, saints, e.g., Paradise ๐ and Svarga โจ. | |
| - **Symbols**: Rituals ๐๏ธ and artifacts ๐ฟ of faith. | |
| - **Charts**: Domains and duties for devotees ๐. | |
| 32. ๐ **Planes of Existence** | |
| - **Heaven/Paradise**: Christian/Muslim abode ๐. | |
| - **Svarga**: Hindu divine realm โจ. | |
| - **Underworld**: Spirits linger, e.g., Sheol โฐ๏ธ and Naraka ๐ฅ. | |
| 33. ๐ **Temple Trappings** | |
| - **Cross/Crescent**: Christian/Muslim faith โ๏ธโช๏ธ. | |
| - **Mandalas**: Hindu devotion ๐. | |
| - **Relics**: Saintsโ and prophetsโ legacy ๐๏ธ. | |
| 34. ๐ **Clerical Chart** | |
| - **Gods**: Domains, e.g., creation ๐ and mercy โค๏ธ. | |
| - **Spirits**: Influence, like guidance ๐ฟ and mischief ๐. | |
| - **Saints/Prophets**: Virtues, e.g., justice โ๏ธ and prophecy ๐ฎ. | |
| """ | |
| if __name__ == "__main__": | |
| main() |