Spaces:
Sleeping
Sleeping
Update start_mcp.py
Browse files- start_mcp.py +188 -166
start_mcp.py
CHANGED
|
@@ -1,166 +1,188 @@
|
|
| 1 |
-
#!/usr/bin/env python3
|
| 2 |
-
"""
|
| 3 |
-
MaTableGPT MCP Server Launcher
|
| 4 |
-
==============================
|
| 5 |
-
|
| 6 |
-
This script starts the MaTableGPT MCP service for extracting
|
| 7 |
-
table data from materials science literature.
|
| 8 |
-
|
| 9 |
-
Usage:
|
| 10 |
-
python start_mcp.py [--host HOST] [--port PORT] [--mode MODE]
|
| 11 |
-
|
| 12 |
-
Arguments:
|
| 13 |
-
--host Host address (default: 0.0.0.0)
|
| 14 |
-
--port Port number (default: 7860)
|
| 15 |
-
--mode Run mode: 'stdio' or 'sse' (default: sse for HuggingFace Space)
|
| 16 |
-
|
| 17 |
-
Environment Variables:
|
| 18 |
-
LLM_API_KEY / OPENAI_API_KEY - API key for LLM service
|
| 19 |
-
LLM_API_BASE / OPENAI_API_BASE - Custom API base URL (for third-party services)
|
| 20 |
-
LLM_MODEL / OPENAI_MODEL - Model name (default: gpt-4-turbo-preview)
|
| 21 |
-
MCP_HOST - Server host (default: 0.0.0.0)
|
| 22 |
-
MCP_PORT - Server port (default: 7860)
|
| 23 |
-
"""
|
| 24 |
-
|
| 25 |
-
import os
|
| 26 |
-
import sys
|
| 27 |
-
import argparse
|
| 28 |
-
import logging
|
| 29 |
-
|
| 30 |
-
# Add
|
| 31 |
-
sys.path.insert(0, os.path.dirname(os.path.
|
| 32 |
-
|
| 33 |
-
# Configure logging
|
| 34 |
-
logging.basicConfig(
|
| 35 |
-
level=logging.INFO,
|
| 36 |
-
format='%(asctime)s - %(name)s - %(levelname)s - %(message)s'
|
| 37 |
-
)
|
| 38 |
-
logger = logging.getLogger("matablgpt-mcp-launcher")
|
| 39 |
-
|
| 40 |
-
|
| 41 |
-
def check_environment():
|
| 42 |
-
"""Check if required environment variables are set."""
|
| 43 |
-
warnings = []
|
| 44 |
-
|
| 45 |
-
# Check for API key (support both naming conventions)
|
| 46 |
-
api_key = os.environ.get('LLM_API_KEY') or os.environ.get('OPENAI_API_KEY')
|
| 47 |
-
if not api_key:
|
| 48 |
-
warnings.append(
|
| 49 |
-
"LLM_API_KEY/OPENAI_API_KEY not set. GPT extraction features will not work. "
|
| 50 |
-
"Set it in HuggingFace Space secrets or environment variables."
|
| 51 |
-
)
|
| 52 |
-
|
| 53 |
-
# Check for API base (for third-party services)
|
| 54 |
-
api_base = os.environ.get('LLM_API_BASE') or os.environ.get('OPENAI_API_BASE')
|
| 55 |
-
if api_base:
|
| 56 |
-
logger.info(f"Using custom API base: {api_base}")
|
| 57 |
-
|
| 58 |
-
return warnings
|
| 59 |
-
|
| 60 |
-
|
| 61 |
-
def check_dependencies():
|
| 62 |
-
"""Check if required packages are installed."""
|
| 63 |
-
missing = []
|
| 64 |
-
|
| 65 |
-
required = [
|
| 66 |
-
('mcp', 'mcp[cli]'),
|
| 67 |
-
('openai', 'openai'),
|
| 68 |
-
('bs4', 'beautifulsoup4'),
|
| 69 |
-
('pandas', 'pandas'),
|
| 70 |
-
('lxml', 'lxml')
|
| 71 |
-
]
|
| 72 |
-
|
| 73 |
-
for module, package in required:
|
| 74 |
-
try:
|
| 75 |
-
__import__(module)
|
| 76 |
-
except ImportError:
|
| 77 |
-
missing.append(package)
|
| 78 |
-
|
| 79 |
-
return missing
|
| 80 |
-
|
| 81 |
-
|
| 82 |
-
def
|
| 83 |
-
"""
|
| 84 |
-
|
| 85 |
-
|
| 86 |
-
|
| 87 |
-
|
| 88 |
-
|
| 89 |
-
|
| 90 |
-
)
|
| 91 |
-
|
| 92 |
-
|
| 93 |
-
|
| 94 |
-
|
| 95 |
-
|
| 96 |
-
|
| 97 |
-
|
| 98 |
-
|
| 99 |
-
|
| 100 |
-
|
| 101 |
-
|
| 102 |
-
|
| 103 |
-
|
| 104 |
-
|
| 105 |
-
|
| 106 |
-
|
| 107 |
-
|
| 108 |
-
|
| 109 |
-
|
| 110 |
-
|
| 111 |
-
|
| 112 |
-
)
|
| 113 |
-
|
| 114 |
-
|
| 115 |
-
|
| 116 |
-
|
| 117 |
-
|
| 118 |
-
|
| 119 |
-
|
| 120 |
-
|
| 121 |
-
|
| 122 |
-
|
| 123 |
-
|
| 124 |
-
|
| 125 |
-
|
| 126 |
-
|
| 127 |
-
|
| 128 |
-
|
| 129 |
-
|
| 130 |
-
|
| 131 |
-
|
| 132 |
-
|
| 133 |
-
|
| 134 |
-
|
| 135 |
-
|
| 136 |
-
|
| 137 |
-
|
| 138 |
-
|
| 139 |
-
|
| 140 |
-
|
| 141 |
-
|
| 142 |
-
|
| 143 |
-
|
| 144 |
-
|
| 145 |
-
|
| 146 |
-
|
| 147 |
-
|
| 148 |
-
|
| 149 |
-
|
| 150 |
-
|
| 151 |
-
|
| 152 |
-
|
| 153 |
-
|
| 154 |
-
|
| 155 |
-
|
| 156 |
-
logger.
|
| 157 |
-
|
| 158 |
-
|
| 159 |
-
|
| 160 |
-
|
| 161 |
-
|
| 162 |
-
|
| 163 |
-
|
| 164 |
-
|
| 165 |
-
|
| 166 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
#!/usr/bin/env python3
|
| 2 |
+
"""
|
| 3 |
+
MaTableGPT MCP Server Launcher
|
| 4 |
+
==============================
|
| 5 |
+
|
| 6 |
+
This script starts the MaTableGPT MCP service for extracting
|
| 7 |
+
table data from materials science literature.
|
| 8 |
+
|
| 9 |
+
Usage:
|
| 10 |
+
python start_mcp.py [--host HOST] [--port PORT] [--mode MODE]
|
| 11 |
+
|
| 12 |
+
Arguments:
|
| 13 |
+
--host Host address (default: 0.0.0.0)
|
| 14 |
+
--port Port number (default: 7860)
|
| 15 |
+
--mode Run mode: 'stdio' or 'sse' (default: sse for HuggingFace Space)
|
| 16 |
+
|
| 17 |
+
Environment Variables:
|
| 18 |
+
LLM_API_KEY / OPENAI_API_KEY - API key for LLM service
|
| 19 |
+
LLM_API_BASE / OPENAI_API_BASE - Custom API base URL (for third-party services)
|
| 20 |
+
LLM_MODEL / OPENAI_MODEL - Model name (default: gpt-4-turbo-preview)
|
| 21 |
+
MCP_HOST - Server host (default: 0.0.0.0)
|
| 22 |
+
MCP_PORT - Server port (default: 7860)
|
| 23 |
+
"""
|
| 24 |
+
|
| 25 |
+
import os
|
| 26 |
+
import sys
|
| 27 |
+
import argparse
|
| 28 |
+
import logging
|
| 29 |
+
|
| 30 |
+
# Add current directory to path for imports
|
| 31 |
+
sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
|
| 32 |
+
|
| 33 |
+
# Configure logging
|
| 34 |
+
logging.basicConfig(
|
| 35 |
+
level=logging.INFO,
|
| 36 |
+
format='%(asctime)s - %(name)s - %(levelname)s - %(message)s'
|
| 37 |
+
)
|
| 38 |
+
logger = logging.getLogger("matablgpt-mcp-launcher")
|
| 39 |
+
|
| 40 |
+
|
| 41 |
+
def check_environment():
|
| 42 |
+
"""Check if required environment variables are set."""
|
| 43 |
+
warnings = []
|
| 44 |
+
|
| 45 |
+
# Check for API key (support both naming conventions)
|
| 46 |
+
api_key = os.environ.get('LLM_API_KEY') or os.environ.get('OPENAI_API_KEY')
|
| 47 |
+
if not api_key:
|
| 48 |
+
warnings.append(
|
| 49 |
+
"LLM_API_KEY/OPENAI_API_KEY not set. GPT extraction features will not work. "
|
| 50 |
+
"Set it in HuggingFace Space secrets or environment variables."
|
| 51 |
+
)
|
| 52 |
+
|
| 53 |
+
# Check for API base (for third-party services)
|
| 54 |
+
api_base = os.environ.get('LLM_API_BASE') or os.environ.get('OPENAI_API_BASE')
|
| 55 |
+
if api_base:
|
| 56 |
+
logger.info(f"Using custom API base: {api_base}")
|
| 57 |
+
|
| 58 |
+
return warnings
|
| 59 |
+
|
| 60 |
+
|
| 61 |
+
def check_dependencies():
|
| 62 |
+
"""Check if required packages are installed."""
|
| 63 |
+
missing = []
|
| 64 |
+
|
| 65 |
+
required = [
|
| 66 |
+
('mcp', 'mcp[cli]'),
|
| 67 |
+
('openai', 'openai'),
|
| 68 |
+
('bs4', 'beautifulsoup4'),
|
| 69 |
+
('pandas', 'pandas'),
|
| 70 |
+
('lxml', 'lxml')
|
| 71 |
+
]
|
| 72 |
+
|
| 73 |
+
for module, package in required:
|
| 74 |
+
try:
|
| 75 |
+
__import__(module)
|
| 76 |
+
except ImportError:
|
| 77 |
+
missing.append(package)
|
| 78 |
+
|
| 79 |
+
return missing
|
| 80 |
+
|
| 81 |
+
|
| 82 |
+
def run_sse_server(host: str, port: int):
|
| 83 |
+
"""Run MCP server in SSE mode using uvicorn."""
|
| 84 |
+
import uvicorn
|
| 85 |
+
from mcp_service import mcp
|
| 86 |
+
|
| 87 |
+
# Get the SSE app from FastMCP
|
| 88 |
+
sse_app = mcp.sse_app()
|
| 89 |
+
|
| 90 |
+
logger.info(f"Starting MCP SSE server on {host}:{port}")
|
| 91 |
+
logger.info(f"SSE Endpoint: http://{host}:{port}/sse")
|
| 92 |
+
|
| 93 |
+
# Run with uvicorn
|
| 94 |
+
uvicorn.run(
|
| 95 |
+
sse_app,
|
| 96 |
+
host=host,
|
| 97 |
+
port=port,
|
| 98 |
+
log_level="info"
|
| 99 |
+
)
|
| 100 |
+
|
| 101 |
+
|
| 102 |
+
def run_stdio_server():
|
| 103 |
+
"""Run MCP server in stdio mode."""
|
| 104 |
+
from mcp_service import mcp
|
| 105 |
+
logger.info("Starting MCP server in stdio mode...")
|
| 106 |
+
mcp.run()
|
| 107 |
+
|
| 108 |
+
|
| 109 |
+
def main():
|
| 110 |
+
"""Main entry point."""
|
| 111 |
+
# Get default values from environment variables
|
| 112 |
+
default_host = os.environ.get('MCP_HOST', '0.0.0.0')
|
| 113 |
+
default_port = int(os.environ.get('MCP_PORT', '7860'))
|
| 114 |
+
|
| 115 |
+
parser = argparse.ArgumentParser(
|
| 116 |
+
description="MaTableGPT MCP Server - Table Data Extraction from Materials Science Literature"
|
| 117 |
+
)
|
| 118 |
+
parser.add_argument(
|
| 119 |
+
'--host',
|
| 120 |
+
default=default_host,
|
| 121 |
+
help=f'Host address (default: {default_host})'
|
| 122 |
+
)
|
| 123 |
+
parser.add_argument(
|
| 124 |
+
'--port',
|
| 125 |
+
type=int,
|
| 126 |
+
default=default_port,
|
| 127 |
+
help=f'Port number (default: {default_port})'
|
| 128 |
+
)
|
| 129 |
+
parser.add_argument(
|
| 130 |
+
'--mode',
|
| 131 |
+
choices=['stdio', 'sse'],
|
| 132 |
+
default='sse',
|
| 133 |
+
help='Run mode: stdio for standard I/O, sse for Server-Sent Events (default: sse)'
|
| 134 |
+
)
|
| 135 |
+
parser.add_argument(
|
| 136 |
+
'--debug',
|
| 137 |
+
action='store_true',
|
| 138 |
+
help='Enable debug logging'
|
| 139 |
+
)
|
| 140 |
+
|
| 141 |
+
args = parser.parse_args()
|
| 142 |
+
|
| 143 |
+
if args.debug:
|
| 144 |
+
logging.getLogger().setLevel(logging.DEBUG)
|
| 145 |
+
|
| 146 |
+
# Check dependencies
|
| 147 |
+
missing = check_dependencies()
|
| 148 |
+
if missing:
|
| 149 |
+
logger.error(f"Missing required packages: {', '.join(missing)}")
|
| 150 |
+
logger.error(f"Install with: pip install {' '.join(missing)}")
|
| 151 |
+
sys.exit(1)
|
| 152 |
+
|
| 153 |
+
# Check environment
|
| 154 |
+
warnings = check_environment()
|
| 155 |
+
for warning in warnings:
|
| 156 |
+
logger.warning(warning)
|
| 157 |
+
|
| 158 |
+
# Display startup info
|
| 159 |
+
logger.info("=" * 60)
|
| 160 |
+
logger.info("MaTableGPT MCP Server")
|
| 161 |
+
logger.info("=" * 60)
|
| 162 |
+
logger.info(f"Mode: {args.mode}")
|
| 163 |
+
if args.mode == 'sse':
|
| 164 |
+
logger.info(f"Host: {args.host}")
|
| 165 |
+
logger.info(f"Port: {args.port}")
|
| 166 |
+
logger.info(f"SSE Endpoint: http://{args.host}:{args.port}/sse")
|
| 167 |
+
logger.info("=" * 60)
|
| 168 |
+
|
| 169 |
+
# Run the server
|
| 170 |
+
try:
|
| 171 |
+
if args.mode == 'stdio':
|
| 172 |
+
run_stdio_server()
|
| 173 |
+
else:
|
| 174 |
+
run_sse_server(args.host, args.port)
|
| 175 |
+
|
| 176 |
+
except ImportError as e:
|
| 177 |
+
logger.error(f"Failed to import required module: {e}")
|
| 178 |
+
logger.error("Make sure all dependencies are installed: pip install -r requirements.txt")
|
| 179 |
+
sys.exit(1)
|
| 180 |
+
except Exception as e:
|
| 181 |
+
logger.error(f"Error starting MCP server: {e}")
|
| 182 |
+
import traceback
|
| 183 |
+
traceback.print_exc()
|
| 184 |
+
sys.exit(1)
|
| 185 |
+
|
| 186 |
+
|
| 187 |
+
if __name__ == "__main__":
|
| 188 |
+
main()
|