EnvironmentsΒΆ
Choosing between development, testing, and production environments via environment variables.
(Example courtesy of Anthropic claude-3-7-sonnet.)
# ruff: noqa: F821,PLR2004
import os
import sys
import random
from enum import Enum
import ictruck
class Environment( Enum ):
DEV = 'development'
TEST = 'testing'
PROD = 'production'
def main( ):
env_name = os.environ.get( 'APP_ENV', 'development' ).lower( )
try: env = next( ev for ev in Environment if ev.value == env_name )
except StopIteration:
print(
f"Warning: Unknown environment '{env_name}', "
"defaulting to development", file = sys.stderr )
env = Environment.DEV
flavors: dict[ ictruck.Flavor, ictruck.FlavorConfiguration ] = {
'api': ictruck.FlavorConfiguration( prefix_emitter = 'API| ' ),
'db': ictruck.FlavorConfiguration( prefix_emitter = 'DATABASE| ' ),
'auth': ictruck.FlavorConfiguration( prefix_emitter = 'AUTH| ' ),
'cache': ictruck.FlavorConfiguration( prefix_emitter = 'CACHE| ' ),
'ui': ictruck.FlavorConfiguration( prefix_emitter = 'UI| ' ),
'error': ictruck.FlavorConfiguration( prefix_emitter = 'ERROR| ' ),
'warning': ictruck.FlavorConfiguration( prefix_emitter = 'WARNING| ' ),
'notice': ictruck.FlavorConfiguration( prefix_emitter = 'NOTICE| ' ),
**ictruck.produce_default_flavors( ),
}
trace_levels = -1 # Default to disabled
active_flavors = [ ]
match env:
case Environment.DEV:
# Development: Full debugging
trace_levels = 9 # All trace levels
active_flavors = [
'api', 'db', 'auth', 'cache', 'ui',
'error', 'warning', 'notice' ]
case Environment.TEST:
# Testing: Moderate debugging
trace_levels = 2 # Only important traces
active_flavors = [ 'api', 'db', 'auth', 'error', 'warning' ]
case Environment.PROD:
# Production: Minimal debugging (errors only)
trace_levels = 0 # Only critical traces
active_flavors = [ 'error' ]
# Allow trace levels to be overridden by environment variable
if 'DEBUG_LEVEL' in os.environ:
try: trace_levels = int( os.environ[ 'DEBUG_LEVEL' ] )
except ValueError:
print(
f"Warning: Invalid DEBUG_LEVEL '{os.environ['DEBUG_LEVEL']}', "
"using default", file = sys.stderr )
# Allow active flavors to be overridden by environment variable
if 'DEBUG_FLAVORS' in os.environ:
override_flavors = [
f.strip( ) for f in os.environ[ 'DEBUG_FLAVORS' ].split( "," ) ]
# Only use flavors that are registered
active_flavors = [ f for f in override_flavors if f in flavors ]
generalcfg = ictruck.VehicleConfiguration( flavors = flavors )
ictruck.install(
active_flavors = active_flavors,
generalcfg = generalcfg,
trace_levels = trace_levels )
run_application( )
def run_application( ):
''' Simulates application with various debug outputs. '''
ictr( 0 )( "Application starting" )
# API calls (level 1 trace)
ictr( 1 )( "Loading configuration" )
api_call( "/api/config", "GET" )
# Database operations (level 2 trace)
ictr( 2 )( "Initializing database connections" )
db_query( "SELECT version()" )
# Detailed operations (level 3 trace)
ictr( 3 )( "Detailed initialization steps" )
# Named flavors for specific subsystems
ictr( 'api' )( "API subsystem initialized" )
ictr( 'db' )( "Database connections established" )
ictr( 'auth' )( "Authentication service ready" )
# Always show error messages
ictr( 'error' )( "This error message should appear even in production" )
# Simulate a user operation with multiple debug levels
process_user_operation( )
# Application shutdown
ictr( 0 )( "Application shutting down" )
def api_call( endpoint, method, data = None ):
''' Simulates API call with debug tracking. '''
ictr( 'api' )( method, endpoint )
# Simulate success/failure
if random.random( ) < 0.9: # noqa: S311
ictr( 'api' )( "API call successful" )
return { 'status': 'success' }
ictr( 'warning' )( "API call failed" )
return { 'status': 'error' }
def db_query( query ):
''' Simulates database query with debug tracking. '''
ictr( 'db' )( query )
ictr( 3 )( "Establishing connection pool" )
ictr( 3 )( "Preparing statement" )
# Success/failure simulation
if random.random( ) < 0.95: # noqa: S311
ictr( 'db' )( "Query executed successfully" )
return [ { 'result': 'Sample data' } ]
ictr( 'error' )( "Database query failed" )
return None
def process_user_operation( ):
''' Simulates user operation with various debug outputs. '''
ictr( 1 )( "Processing user operation." )
ictr( 'auth' )( "Verifying user credentials" )
ictr( 'cache' )( "Checking cache for data" )
ictr( 2 )( "Creating user context." )
ictr( 2 )( "Validating input data." )
# Simulate error condition
if random.random( ) < 0.2: # noqa: S311
ictr( 'error' )( "Invalid user input detected." )
ictr( 'ui' )( "Updating user interface." )
ictr( 'notice' )( "User operation completed." )
if __name__ == '__main__': main( )
Running this will result in the following (or something similar):
TRACE0| 'Application starting'
TRACE1| 'Loading configuration'
API| method: 'GET', endpoint: '/api/config'
API| f"API call successful": 'API call successful'
TRACE2| 'Initializing database connections'
DATABASE| query: 'SELECT version()'
TRACE3| 'Establishing connection pool'
TRACE3| 'Preparing statement'
DATABASE| 'Query executed successfully'
TRACE3| 'Detailed initialization steps'
API| 'API subsystem initialized'
DATABASE| 'Database connections established'
AUTH| 'Authentication service ready'
ERROR| 'This error message should appear even in production'
TRACE1| 'Processing user operation.'
AUTH| 'Verifying user credentials'
CACHE| 'Checking cache for data'
TRACE2| 'Creating user context.'
TRACE2| 'Validating input data.'
ERROR| 'Invalid user input detected.'
UI| 'Updating user interface.'
NOTICE| 'User operation completed.'
TRACE0| 'Application shutting down'
Running this with APP_ENV set to testing will result in the following
(or something similar):
TRACE0| 'Application starting'
TRACE1| 'Loading configuration'
API| method: 'GET', endpoint: '/api/config'
API| f"API call successful": 'API call successful'
TRACE2| 'Initializing database connections'
DATABASE| query: 'SELECT version()'
DATABASE| 'Query executed successfully'
API| 'API subsystem initialized'
DATABASE| 'Database connections established'
AUTH| 'Authentication service ready'
ERROR| 'This error message should appear even in production'
TRACE1| 'Processing user operation.'
AUTH| 'Verifying user credentials'
TRACE2| 'Creating user context.'
TRACE2| 'Validating input data.'
ERROR| 'Invalid user input detected.'
TRACE0| 'Application shutting down'
And, running this with APP_ENV set to prod will result in the following
(or something similar):
TRACE0| 'Application starting'
ERROR| 'This error message should appear even in production'
TRACE0| 'Application shutting down'