Buckets:

download
raw
4.01 kB
#!/usr/bin/env python3
"""
Generic script to generate OpenAPI JSON specification for any FastAPI backend
without starting the server.
"""
import json
import os
import sys
from pathlib import Path
def main():
"""Generate OpenAPI JSON specification."""
try:
# Determine the backend directory from current working directory
backend_dir = Path.cwd()
app_name = '_'.join(sys.argv[1:])
app_dir = backend_dir / app_name
config_dir = backend_dir / "config"
# Validate that we're in a valid backend directory
if not app_dir.exists() or not (app_dir / "main.py").exists():
raise FileNotFoundError(f"Could not find main.py in {backend_dir}")
if not config_dir.exists() or not (config_dir / "configuration.yaml").exists():
raise FileNotFoundError(f"Could not find config/configuration.yaml in {backend_dir}")
# Add the app directory to the Python path
sys.path.insert(0, str(app_dir))
# Set required environment variables
os.environ.setdefault("ENV_FILE", str(config_dir / ".env"))
os.environ.setdefault("CONFIG_FILE", str(config_dir / "configuration.yaml"))
# Disable HuggingFace downloads during static OpenAPI generation (prevents network
# errors when generate_openapi imports ML models that try to download from HuggingFace)
os.environ.setdefault("HF_DATASETS_OFFLINE", "1")
os.environ.setdefault("HF_HUB_OFFLINE", "1")
os.environ.setdefault("TRANSFORMERS_OFFLINE", "1")
# Set dummy secrets for static generation (prevents validation errors).
# create_app() loads the full config but never connects to any backend here,
# so dummy values satisfy the fail-fast validators (OpenSearch / Postgres
# passwords) that otherwise refuse to start when no .env is present (e.g. CI).
os.environ.setdefault("OPENAI_API_KEY", "sk-dummy-key-for-static-generation")
os.environ.setdefault("OPENSEARCH_PASSWORD", "dummy-password-for-static-generation") # pragma: allowlist secret
os.environ.setdefault("FRED_POSTGRES_PASSWORD", "dummy-password-for-static-generation") # pragma: allowlist secret
# Import and create the FastAPI app
from main import create_app
app = create_app()
# Generate OpenAPI specification
openapi_spec = app.openapi()
# Output file in the backend directory
output_file = backend_dir / "openapi.json"
# Write to file
with open(output_file, 'w') as f:
json.dump(openapi_spec, f, indent=2)
backend_name = backend_dir.name
print(f"✅ OpenAPI specification generated for {backend_name}: {output_file}")
# Cleanup ApplicationContext resources (e.g. ThreadPoolExecutor) to ensure clean exit
try:
# Try to find the application_context module for this app
app_ctx_module_name = f"{app_name}.application_context"
if app_ctx_module_name in sys.modules:
ctx_module = sys.modules[app_ctx_module_name]
if hasattr(ctx_module, "get_app_context"):
try:
ctx = ctx_module.get_app_context()
if hasattr(ctx, "shutdown"):
import asyncio
asyncio.run(ctx.shutdown())
elif hasattr(ctx, "shutdown_io_executor"):
ctx.shutdown_io_executor()
except RuntimeError:
pass # Context might not be initialized
except Exception:
# Best-effort cleanup only; a failed context lookup/shutdown must not fail spec generation.
pass
return 0
except Exception as e:
print(f"❌ Error generating OpenAPI specification: {e}")
return 1
if __name__ == "__main__":
sys.exit(main())

Xet Storage Details

Size:
4.01 kB
·
Xet hash:
b7bc540ed10c6e57a73eb7ca7382fec43ddee33e9dd8aaaa99ddc7f850a6c612

Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.