Skip to main content

Read variables from environment or files into dataclasses

Project description


Build status

Read variables from the environment or files into dataclasses.

While it is convenient to configure applications using environment variables during development, it is advised not to store sensitive information such as passwords in environment variables in production environments. The envee library allows reading variables either from the environment variables or files (which are typically used by e.g. docker secrets), thus keeping code used for development and production environments as close to each other as possible.


Variables to read from the environment are declared using classes annotated with the @envee.environment decorator. Using the fields of the classes are filled using the environment.


from typing import Optional
import envee

class Environment:
    username: str
    debug: Optional[str]
    workers: int = 5

env =

Environment variables names and file paths

For each field, per default envee looks for environment variables with the upper case name of the field. The corresponding file is looked in the directory /run/secrets and has the lower case field name as filename. If a corresponding file is found, the file will take precedence over an environment variable. For the example above, the read() method looks for the environment variables USERNAME, DEBUG, and WORKERS, respectively the files /run/secrets/username, /run/secrets/debug, and /run/secrets/workers.


Variables are typed using the dataclasses. Primitive types such as int, float, or str are automatically converted while reading. For more complex types, a conversion function needs to be provided. If fields are typed as Optional, their value will be set to None if no variable is found. If a default value is defined, this value will be used when no corresponding environment variable or file is found. Otherwise, when no environment variable is found and the field is not typed as Optional, a RuntimeError is raised.

dotenv (.env) support

envee provides rudimentary support for .env files. Currently, the path to the .env file must be specified in the read() method. The name of the variable name must be the upper case name of the field name. Comments and multiline variables are supported. Variables found in a .env file take precedence over environment variables but not files.

The following .env file can be read using the following Python code:

DEBUG="True" # a comment
class Environment:
    debug: str
    workers: int
    multiline: str

env =, dotenv_path="/path/to/.env/file")

Advanced usage

Override environment variable names

In the following example the field debug is filled using the environment variable PROJECT_DEBUG.

import envee

class Environment:
    debug: str = envee.field(env_name="PROJECT_DEBUG")

env =

Override file locations

The default location can be changed by passing a different location to the read() method.

import envee

class Environment:
    debug: str

env =, default_files_location="/path/to/a/directory")

Alternatively, the fields metadata can specify the file_location or file_name. The parameter file_location overrides default_files_location. file_name overrides the default file name. The direct path to a file can be specified using file_path. file_path will take precedence over file_location or file_name.

import envee

class Environment:
    debug: str = envee.field(file_location="/other/dir", file_name="DEBUG.txt")

env =

Complex datatypes

For complex datatypes, a conversion function needs to be passed to the field.

import envee

class Environment:
    timestamp: datetime.datetime = envee.field(
        conversion_func=lambda x: datetime.datetime.fromisoformat(x)

env =

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

envee-0.1.2.tar.gz (6.7 kB view hashes)

Uploaded source

Built Distribution

envee-0.1.2-py3-none-any.whl (6.9 kB view hashes)

Uploaded py3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page