How do i read parquet files on s3? #25
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| name: HugChat Issue Response | |
| on: | |
| issues: | |
| types: [opened, edited, closed, deleted] | |
| issue_comment: | |
| types: [created, edited] | |
| jobs: | |
| cleanup-conversation: | |
| if: github.event.action == 'closed' || github.event.action == 'deleted' | |
| runs-on: ubuntu-latest | |
| permissions: | |
| issues: read | |
| steps: | |
| - name: Checkout repository | |
| uses: actions/checkout@v4 | |
| - name: Set up Python | |
| uses: actions/setup-python@v5 | |
| with: | |
| python-version: '3.10' | |
| - name: Install dependencies | |
| run: | | |
| python -m pip install --upgrade pip | |
| pip install hugchat | |
| - name: Create HugChat cleanup script | |
| run: | | |
| cat > cleanup_hugchat.py << 'EOL' | |
| #!/usr/bin/env python3 | |
| import argparse | |
| import os | |
| import sys | |
| import re | |
| from hugchat import hugchat | |
| from hugchat.login import Login | |
| def login_to_hugchat(email, password): | |
| if not email or not password: | |
| print("Error: Email and password are required", file=sys.stderr) | |
| sys.exit(1) | |
| try: | |
| print(f"Logging in with email: {email[:3]}***", file=sys.stderr) | |
| os.makedirs('./cookies/', exist_ok=True) | |
| sign = Login(email, password) | |
| cookies = sign.login(cookie_dir_path='./cookies/', save_cookies=True) | |
| return cookies.get_dict() | |
| except Exception as e: | |
| print(f"Login failed: {e}", file=sys.stderr) | |
| sys.exit(1) | |
| def cleanup_conversation(conversation_id): | |
| email = os.environ.get('HUGCHAT_EMAIL') | |
| password = os.environ.get('HUGCHAT_PASSWORD') | |
| cookies = login_to_hugchat(email, password) | |
| try: | |
| chatbot = hugchat.ChatBot(cookies=cookies) | |
| conversations = chatbot.get_remote_conversations(replace_conversation_list=True) | |
| print(f"Looking for conversation {conversation_id} to delete...", file=sys.stderr) | |
| for conv in conversations: | |
| if conv.id == conversation_id: | |
| print(f"Found conversation, attempting deletion...", file=sys.stderr) | |
| try: | |
| chatbot.delete_conversation(conv) | |
| print(f"✓ Successfully deleted conversation {conversation_id}", file=sys.stderr) | |
| return True | |
| except Exception as delete_error: | |
| print(f"Failed to delete conversation: {delete_error}", file=sys.stderr) | |
| return False | |
| print(f"Conversation {conversation_id} not found (may already be deleted)", file=sys.stderr) | |
| return True # Consider this success since the goal is achieved | |
| except Exception as e: | |
| print(f"Error during cleanup: {e}", file=sys.stderr) | |
| return False | |
| if __name__ == "__main__": | |
| parser = argparse.ArgumentParser(description='Cleanup HugChat conversation') | |
| parser.add_argument('conversation_id', help='Conversation ID to delete') | |
| args = parser.parse_args() | |
| success = cleanup_conversation(args.conversation_id) | |
| sys.exit(0 if success else 1) | |
| EOL | |
| chmod +x cleanup_hugchat.py | |
| - name: Extract and cleanup conversation | |
| env: | |
| HUGCHAT_EMAIL: ${{ secrets.HUGCHAT_EMAIL }} | |
| HUGCHAT_PASSWORD: ${{ secrets.HUGCHAT_PASSWORD }} | |
| run: | | |
| echo "Issue closed/deleted: #${{ github.event.issue.number }}" | |
| echo "Looking for HugChat conversation to cleanup..." | |
| # Get all comments for this issue and look for HugChat Session ID | |
| COMMENTS=$(gh api repos/${{ github.repository }}/issues/${{ github.event.issue.number }}/comments --jq '.[].body' || echo "") | |
| # Also check the issue body itself | |
| ISSUE_BODY="${{ github.event.issue.body }}" | |
| ALL_CONTENT="$COMMENTS$ISSUE_BODY" | |
| # Extract the HugChat Session ID | |
| if echo "$ALL_CONTENT" | grep -q "HugChat Session ID:"; then | |
| CONVERSATION_ID=$(echo "$ALL_CONTENT" | grep "HugChat Session ID:" | tail -1 | sed -n 's/.*HugChat Session ID: `\([^`]*\)`.*/\1/p') | |
| if [ ! -z "$CONVERSATION_ID" ]; then | |
| echo "Found HugChat conversation to cleanup: $CONVERSATION_ID" | |
| if python cleanup_hugchat.py "$CONVERSATION_ID"; then | |
| echo "✓ Successfully cleaned up HugChat conversation" | |
| else | |
| echo "⚠️ Failed to cleanup HugChat conversation" | |
| fi | |
| else | |
| echo "Could not extract conversation ID from comments" | |
| fi | |
| else | |
| echo "No HugChat session found for this issue - nothing to cleanup" | |
| fi | |
| respond-to-issue: | |
| if: github.event.action != 'closed' && github.event.action != 'deleted' | |
| runs-on: ubuntu-latest | |
| permissions: | |
| issues: write | |
| steps: | |
| - name: Checkout repository | |
| uses: actions/checkout@v4 | |
| - name: Set up Python | |
| uses: actions/setup-python@v5 | |
| with: | |
| python-version: '3.10' | |
| - name: Install dependencies | |
| run: | | |
| python -m pip install --upgrade pip | |
| pip install hugchat | |
| - name: Create HugChat CLI script | |
| run: | | |
| mkdir -p cookies | |
| cat > hugchat_cli.py << 'EOL' | |
| #!/usr/bin/env python3 | |
| import argparse | |
| import os | |
| import sys | |
| import re | |
| from hugchat import hugchat | |
| from hugchat.login import Login | |
| def setup_argparse(): | |
| parser = argparse.ArgumentParser(description='Interact with HugChat from command line') | |
| parser.add_argument('prompt', help='The prompt to send to HugChat') | |
| parser.add_argument('--email', help='HuggingFace email', | |
| default=os.environ.get('HUGCHAT_EMAIL')) | |
| parser.add_argument('--password', help='HuggingFace password', | |
| default=os.environ.get('HUGCHAT_PASSWORD')) | |
| parser.add_argument('--cookie-dir', help='Cookie directory path', | |
| default='./cookies/') | |
| parser.add_argument('--assistant-id', help='Assistant ID', | |
| default=os.environ.get('HUGCHAT_ASSISTANTID', '673e290837ec25016921608f')) | |
| parser.add_argument('--web-search', action='store_true', | |
| help='Enable web search') | |
| parser.add_argument('--conversation-id', help='Existing conversation ID to continue', | |
| default=None) | |
| parser.add_argument('--create-new', action='store_true', | |
| help='Force create a new conversation') | |
| return parser | |
| def login_to_hugchat(email, password, cookie_dir): | |
| if not email or not password: | |
| print("Error: Email and password are required. Set them via arguments or environment variables " | |
| "HUGCHAT_EMAIL and HUGCHAT_PASSWORD", file=sys.stderr) | |
| sys.exit(1) | |
| try: | |
| print(f"Attempting to login with email: {email[:3]}***", file=sys.stderr) | |
| os.makedirs(cookie_dir, exist_ok=True) | |
| sign = Login(email, password) | |
| try: | |
| cookies = sign.login(cookie_dir_path=cookie_dir, save_cookies=True) | |
| print("Login successful with password", file=sys.stderr) | |
| return cookies.get_dict() | |
| except Exception as password_error: | |
| print(f"Password login failed: {password_error}", file=sys.stderr) | |
| if len(password) > 20 and password.startswith(('hf_', 'token')): | |
| print("Attempting login with access token method...", file=sys.stderr) | |
| try: | |
| sign = Login(email, None) | |
| sign.token = password | |
| cookies = sign.login(cookie_dir_path=cookie_dir, save_cookies=True) | |
| print("Login successful with token", file=sys.stderr) | |
| return cookies.get_dict() | |
| except Exception as token_error: | |
| print(f"Token login also failed: {token_error}", file=sys.stderr) | |
| raise password_error | |
| except Exception as e: | |
| print(f"Error logging in: {str(e)}", file=sys.stderr) | |
| print("", file=sys.stderr) | |
| print("TROUBLESHOOTING TIPS:", file=sys.stderr) | |
| print("1. Make sure HUGCHAT_PASSWORD is your actual HuggingFace password, NOT an access token", file=sys.stderr) | |
| print("2. Try logging into https://huggingface.co/chat manually to verify credentials", file=sys.stderr) | |
| print("3. Check if your account has 2FA enabled (may cause issues)", file=sys.stderr) | |
| print("4. Ensure your account has access to HugChat", file=sys.stderr) | |
| sys.exit(1) | |
| def manage_conversation(chatbot, conversation_id, assistant_id, create_new): | |
| """Manage conversation - either create new or switch to existing""" | |
| try: | |
| if create_new or not conversation_id: | |
| print("Creating new conversation...", file=sys.stderr) | |
| chatbot.new_conversation(assistant=assistant_id, switch_to=True) | |
| # Get the new conversation ID | |
| current_conv = chatbot.get_conversation_info() | |
| new_conv_id = current_conv.id | |
| print(f"Created new conversation with ID: {new_conv_id}", file=sys.stderr) | |
| return new_conv_id | |
| else: | |
| print(f"Attempting to switch to existing conversation: {conversation_id}", file=sys.stderr) | |
| # First, let's get all conversations to verify the ID exists | |
| print("Fetching conversations...", file=sys.stderr) | |
| conversations = chatbot.get_remote_conversations(replace_conversation_list=True) | |
| # Only show counts and success/failure - don't leak IDs unless there's an error | |
| print(f"Found {len(conversations)} total conversations", file=sys.stderr) | |
| found_target = False | |
| for conv in conversations: | |
| if conv.id == conversation_id: | |
| found_target = True | |
| print(f"✓ Found target conversation in account", file=sys.stderr) | |
| break | |
| if not found_target: | |
| print(f"⚠️ Target conversation not found in {len(conversations)} available conversations", file=sys.stderr) | |
| print(f"Looking for conversation ID: {conversation_id}", file=sys.stderr) | |
| print("This may indicate the conversation was deleted or is inaccessible", file=sys.stderr) | |
| print("Creating new conversation instead...", file=sys.stderr) | |
| chatbot.new_conversation(assistant=assistant_id, switch_to=True) | |
| current_conv = chatbot.get_conversation_info() | |
| new_conv_id = current_conv.id | |
| print(f"Created new conversation with ID: {new_conv_id}", file=sys.stderr) | |
| return new_conv_id | |
| try: | |
| # Try to switch to the conversation | |
| # Fix: Some hugchat versions need the conversation object, not just the ID | |
| target_conversation = None | |
| for conv in conversations: | |
| if conv.id == conversation_id: | |
| target_conversation = conv | |
| break | |
| if target_conversation: | |
| print(f"Switching using conversation object instead of ID", file=sys.stderr) | |
| chatbot.change_conversation(target_conversation) | |
| else: | |
| # Fallback to ID-based switching | |
| chatbot.change_conversation(conversation_id) | |
| # Give it a moment to load (API might need time) | |
| import time | |
| time.sleep(1) | |
| # Force refresh the conversation data | |
| try: | |
| # Some versions might need to explicitly load conversation history | |
| if hasattr(chatbot, 'load_conversation'): | |
| chatbot.load_conversation(conversation_id) | |
| print("Explicitly loaded conversation history", file=sys.stderr) | |
| except Exception as load_error: | |
| print(f"Could not explicitly load conversation: {load_error}", file=sys.stderr) | |
| # Verify we actually switched by checking current conversation | |
| current_conv = chatbot.get_conversation_info() | |
| if current_conv.id == conversation_id: | |
| print(f"✓ Successfully switched to conversation: {conversation_id}", file=sys.stderr) | |
| print(f" Current conversation model: {current_conv.model}", file=sys.stderr) | |
| # Show conversation history count only (not content) | |
| if hasattr(current_conv, 'history') and current_conv.history: | |
| print(f" Conversation has {len(current_conv.history)} messages in history", file=sys.stderr) | |
| else: | |
| print(f" ⚠️ No conversation history found!", file=sys.stderr) | |
| return conversation_id | |
| else: | |
| print(f"⚠️ Failed to switch - current ID is {current_conv.id}, expected {conversation_id}", file=sys.stderr) | |
| raise Exception(f"Conversation switch verification failed") | |
| except Exception as switch_error: | |
| print(f"Failed to switch to conversation {conversation_id}: {switch_error}", file=sys.stderr) | |
| print("Creating new conversation instead...", file=sys.stderr) | |
| chatbot.new_conversation(assistant=assistant_id, switch_to=True) | |
| current_conv = chatbot.get_conversation_info() | |
| new_conv_id = current_conv.id | |
| print(f"Created new conversation with ID: {new_conv_id}", file=sys.stderr) | |
| return new_conv_id | |
| except Exception as e: | |
| print(f"Error managing conversation: {e}", file=sys.stderr) | |
| print("Falling back to new conversation", file=sys.stderr) | |
| chatbot.new_conversation(assistant=assistant_id, switch_to=True) | |
| current_conv = chatbot.get_conversation_info() | |
| return current_conv.id | |
| def main(): | |
| parser = setup_argparse() | |
| args = parser.parse_args() | |
| print(f"Starting HugChat CLI with prompt: {args.prompt[:50]}...", file=sys.stderr) | |
| cookies = login_to_hugchat(args.email, args.password, args.cookie_dir) | |
| try: | |
| print("Creating chatbot...", file=sys.stderr) | |
| chatbot = hugchat.ChatBot(cookies=cookies) | |
| # Manage conversation (create new or switch to existing) | |
| conversation_id = manage_conversation(chatbot, args.conversation_id, args.assistant_id, args.create_new) | |
| # Additional verification: Since conversation switching is working, | |
| # let's trust HugChat's internal context and just verify we're in the right conversation | |
| if args.conversation_id and not args.create_new: | |
| try: | |
| print("Verifying conversation context...", file=sys.stderr) | |
| current_conv = chatbot.get_conversation_info() | |
| if current_conv.id == args.conversation_id: | |
| print("✓ Successfully maintaining conversation context!", file=sys.stderr) | |
| if hasattr(current_conv, 'history') and current_conv.history: | |
| print(f"✓ Conversation has {len(current_conv.history)} messages in history", file=sys.stderr) | |
| else: | |
| print("⚠️ Conversation ID mismatch - context may be lost", file=sys.stderr) | |
| except Exception as context_error: | |
| print(f"Could not verify conversation: {context_error}", file=sys.stderr) | |
| print("Sending message to HugChat...", file=sys.stderr) | |
| # Debug: Verify conversation before sending (without showing content) | |
| try: | |
| current_conv_before = chatbot.get_conversation_info() | |
| print(f"Sending message in conversation: {current_conv_before.id}", file=sys.stderr) | |
| if hasattr(current_conv_before, 'history') and current_conv_before.history: | |
| print(f"Conversation has {len(current_conv_before.history)} messages before sending", file=sys.stderr) | |
| except Exception as debug_error: | |
| print(f"Could not get conversation info before sending: {debug_error}", file=sys.stderr) | |
| message_result = chatbot.chat(args.prompt, web_search=args.web_search) | |
| response = message_result.wait_until_done() | |
| # Debug: Check conversation after sending (without showing content) | |
| try: | |
| current_conv_after = chatbot.get_conversation_info() | |
| print(f"After sending, conversation: {current_conv_after.id}", file=sys.stderr) | |
| if hasattr(current_conv_after, 'history') and current_conv_after.history: | |
| print(f"Conversation now has {len(current_conv_after.history)} messages", file=sys.stderr) | |
| except Exception as debug_error: | |
| print(f"Could not get conversation info after sending: {debug_error}", file=sys.stderr) | |
| # Output the response and conversation ID | |
| print(f"{response}") | |
| print(f"\n---\n*HugChat Session ID: `{conversation_id}`*") | |
| except Exception as e: | |
| print(f"Error during chat: {str(e)}", file=sys.stderr) | |
| import traceback | |
| traceback.print_exc(file=sys.stderr) | |
| sys.exit(1) | |
| if __name__ == "__main__": | |
| main() | |
| EOL | |
| chmod +x hugchat_cli.py | |
| - name: Get issue content and generate response | |
| env: | |
| GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }} | |
| HUGCHAT_EMAIL: ${{ secrets.HUGCHAT_EMAIL }} | |
| HUGCHAT_PASSWORD: ${{ secrets.HUGCHAT_PASSWORD }} | |
| HUGCHAT_ASSISTANTID: ${{ secrets.HUGCHAT_ASSISTANTID }} | |
| ISSUE_TITLE: ${{ github.event.issue.title }} | |
| ISSUE_BODY: ${{ github.event.comment.body || github.event.issue.body }} | |
| run: | | |
| echo "Checking environment variables..." | |
| if [ -z "$HUGCHAT_EMAIL" ]; then | |
| echo "Error: HUGCHAT_EMAIL is not set" | |
| exit 1 | |
| fi | |
| if [ -z "$HUGCHAT_PASSWORD" ]; then | |
| echo "Error: HUGCHAT_PASSWORD is not set" | |
| exit 1 | |
| fi | |
| if [ -z "$HUGCHAT_ASSISTANTID" ]; then | |
| echo "Warning: HUGCHAT_ASSISTANTID is not set, using default" | |
| fi | |
| if [ -z "$ISSUE_TITLE" ]; then | |
| echo "Error: ISSUE_TITLE is empty" | |
| exit 1 | |
| fi | |
| echo "Email is set: ${HUGCHAT_EMAIL:0:3}***" | |
| echo "Assistant ID is set: ${HUGCHAT_ASSISTANTID:-default}" | |
| echo "Issue title: $ISSUE_TITLE" | |
| echo "Issue body length: ${#ISSUE_BODY}" | |
| # Create the full prompt with title and body | |
| if [ -z "$ISSUE_BODY" ] || [ "$ISSUE_BODY" = "" ]; then | |
| echo "Issue body is empty, using title as the main question" | |
| FULL_PROMPT="GitHub Issue: $ISSUE_TITLE" | |
| IS_NEW_ISSUE=true | |
| else | |
| echo "Using both title and body for context" | |
| FULL_PROMPT="GitHub Issue: $ISSUE_TITLE | |
| Issue Description: | |
| $ISSUE_BODY" | |
| IS_NEW_ISSUE=false | |
| fi | |
| echo "Full prompt length: ${#FULL_PROMPT}" | |
| # Check if this is a follow-up comment by looking for existing HugChat session ID | |
| CONVERSATION_ID="" | |
| if [ "$IS_NEW_ISSUE" = false ]; then | |
| echo "Looking for existing HugChat session ID in issue comments..." | |
| # Get all comments for this issue and look for HugChat Session ID | |
| COMMENTS=$(gh api repos/${{ github.repository }}/issues/${{ github.event.issue.number }}/comments --jq '.[].body') | |
| # Extract the last HugChat Session ID found | |
| if echo "$COMMENTS" | grep -q "HugChat Session ID:"; then | |
| CONVERSATION_ID=$(echo "$COMMENTS" | grep "HugChat Session ID:" | tail -1 | sed -n 's/.*HugChat Session ID: `\([^`]*\)`.*/\1/p') | |
| if [ ! -z "$CONVERSATION_ID" ]; then | |
| echo "Found existing HugChat session: $CONVERSATION_ID" | |
| fi | |
| fi | |
| fi | |
| # Determine if we should create a new conversation | |
| if [ -z "$CONVERSATION_ID" ]; then | |
| echo "No existing session found, will create new conversation" | |
| HUGCHAT_ARGS="--create-new" | |
| else | |
| echo "Continuing existing conversation: $CONVERSATION_ID" | |
| HUGCHAT_ARGS="--conversation-id $CONVERSATION_ID" | |
| fi | |
| echo "Generating response using HugChat..." | |
| # Capture both stdout and stderr for debugging | |
| if python hugchat_cli.py --web-search $HUGCHAT_ARGS "$FULL_PROMPT" > /tmp/hugchat_output.log 2> /tmp/hugchat_error.log; then | |
| echo "HugChat response generated successfully" | |
| RESPONSE=$(cat /tmp/hugchat_output.log) | |
| # Show debugging info from stderr | |
| echo "=== HugChat Debug Information ===" | |
| cat /tmp/hugchat_error.log | |
| echo "=== End Debug Information ===" | |
| else | |
| echo "HugChat failed with exit code $?" | |
| echo "Error log:" | |
| cat /tmp/hugchat_error.log | |
| exit 1 | |
| fi | |
| if [ -z "$RESPONSE" ]; then | |
| echo "Error: Empty response from HugChat" | |
| exit 1 | |
| fi | |
| echo "Response length: ${#RESPONSE}" | |
| COMMENT="👋 Hello! | |
| $RESPONSE | |
| > I am a bot powered by Hugging Face. Please verify any information provided." | |
| echo "Posting comment to issue..." | |
| gh issue comment ${{ github.event.issue.number }} --body "$COMMENT" |