dotenv for shells with support for POSIX-compliant and multiple .env file syntax

Project Status: Almost complete. Major features have been implemented and v1.0.0 will be released in the near future.

Important Notes: Incompatible changes were made in Version 0.12.0. If a definition with the same name exists in the .env file when --overload is specified, the later definition takes precedence. Also, it has been changed to default to an error for undesirable usage. We believe this change will not affect many cases, but if you have a problem, please open an issue.

Quoting bkeepers/dotenv:

Storing configuration in the environment is one of the tenets of a twelve-factor app. Anything that is likely to change between deployment environments–such as resource handles for databases or credentials for external services–should be extracted from the code into environment variables.

It is not safe. There is no formal specification for the .env file syntax, and different languages, libraries, and tools use different syntaxes. If you load a .env file syntax that is incompatible with the POSIX shell syntax, you will get unexpected results and may even result in the execution of scripts.

shdotenv safely loads the syntax of .env files that are compatible with POSIX shell syntax. There is no possibility that the script will be executed. And also, for interoperability, .env files with other syntaxes are supported whenever possible.

1. Provide language-independent CLI utilities
2. Provide a library that can safely load .env file from shell scripts
3. Define POSIX shell compatible .env file syntax specification
4. Support for .env file syntax dialects for interoperation

shdotenv is a single file shell script with embedded awk script. It uses only the following commands which can be found anywhere.

• POSIX shell (dash, bash, ksh, zsh, etc)
• awk (gawk, nawk, mawk, busybox awk)

Download shdotenv (shell script) from releases.

$ wget https://github.com/ko1nksm/shdotenv/releases/latest/download/shdotenv -O $HOME/bin/shdotenv
$ chmod +x $HOME/bin/shdotenv

Build and install only

$ git clone https://github.com/ko1nksm/shdotenv.git
$ cd shdotenv
$ make build
$ make install PREFIX=$HOME

Full build

A full build requires requires shfmt, shellcheck and shellspec.

$ git clone https://github.com/ko1nksm/shdotenv.git
$ cd shdotenv
$ make MINIFY=true
$ make install PREFIX=$HOME

Note for developers: shdotenv can be run in source code without building. Please run src/shdotenv.

Usage: shdotenv [OPTION]... [--] [[COMMAND | export] [ARG]...]

  If the COMMAND is specified, it will load .env files and run the command.
  If the COMMAND is omitted, it will output the result of interpreting .env
  files. It can be safely loaded into the shell (For example, using eval).

Options:
  -d, --dialect DIALECT     Specify the .env dialect [default: posix]
                                posix, ruby, node, python,
                                php, go, rust, docker
  -f, --format FORMAT       Output in the specified format [default: sh]
                                sh, csh, fish, json, jsonl, yaml, name
  -e, --env ENV_PATH        Location of the .env file [default: .env]
                              Multiple -e options are allowed
                              If the ENV_PATH is "-", read from stdin
  -i, --ignore-environment  Ignore the current environment variables
      --overload            Overload predefined variables
      --no-allexport        Disable all variable export
                              Same as deprecated --noexport
      --no-nounset          Allow references to undefined variables
      --grep PATTERN        Output only names that match the regexp pattern
  -s, --sort                Sort variable names
  -q, --quiet               Suppress all output (useful for test .env files)
  -v, --version             Show the version and exit
  -h, --help                Show this message and exit

Usage: shdotenv export [-n | -p] [--] [NAME]...
  Exports environment variables in posix-compliant .env format.

  -n  List only environment variable names
  -p  Append "export" prefix to environment variable names

  This will be output after the .env files is loaded. If you do not want
  to load it, specify "-e /dev/null". This is similar to "export", "env"
  and "printenv" commands, but quoting correctly and exports only portable
  environment variable name that are valid as identifier for posix shell.

Set environment variables and execute the specified command.

shdotenv [OPTION]... <COMMAND> [ARGUMENTS]...

shdotenv --quiet --env .env

Load the .env file into the shell script. When run on the shell, it exports to the current shell.

eval "$(shdotenv [OPTION]...)"

You may want to abort the program when the .env file fails to parse. In that case, do the following

eval "$(shdotenv [OPTION]... || echo "exit $?")"

set newline='\
'
eval "`shdotenv -f csh [OPTION]...`"

eval (shdotenv -f fish [OPTION]...)

This is similar to export, env and printenv commands, but quoting correctly and exports only portable environment variable name that are valid as identifier for POSIX shell.

shdotenv export [-n | -p] [NAME]...

The docker command has the --env-file option, but it only supports setting simple values.

• docker cannot pass newlines from variables in --env-file files

This tool makes the files read by --env-file compatible with the .env format, and supports variable expansion and newlines.

Example: (Use dockerenv instead of docker)

dockerenv run --env-file .env -it debian

# dotenv posix
# This line is a comment, The above line is a directive
COMMENT=This-#-is-a-character # This is a comment

UNQUOTED=value1 # Spaces and some special characters cannot be used
SINGLE_QUOTED='value 2' # Cannot use single quote
DOUBLE_QUOTED="value 3" # Some special characters need to be escaped

MULTILINE="line1
line2: \n is not a newline
line3"
LONGLINE="https://github.com/ko1nksm\
/shdotenv/blob/main/README.md"

ENDPOINT="http://${HOST}/api" # Variable expansion requires braces

export EXPORT1="value"
export EXPORT2 # Equivalent to: export EXPORT2="${EXPORT2:-}"

• The syntax is a subset of the POSIX shell.
• The first line is an optional directive that specifies the dialect of the .env syntax
• No spaces are allowed before or after the = separating the name and value
• ANSI-C style escapes are not available (i.e., \n is not a newline)
• Unquoted value
  • The special characters that can be used are # % + , - . / : = @ ^ _
• Single-quoted value
  • The disallowed character is: '
  • It can contain newline characters.
• Double-quoted value
  • Variable expansion is available (only ${VAR} style is supported)
  • The following values should be escaped with a backslash (\): $ ` " \
  • The \ at the end of a line value means line continuation
  • It can contain newline characters.
• An optional export prefix can be added to the name
• Comments at the end of a line need to be preceded by spaces before the #

Detailed POSIX-compliant .env syntax specification

Specifies the dotenv syntanx dialect that this .env file.

# dotenv <DIALECT>

Example:

# dotenv ruby

The formal .env syntax for this project is posix only. The posix is a subset of the POSIX shell and is compatible with shell scripts. Support for other .env syntax dialects is for interoperability purposes. Compatibility will be improved gradually, but is not fully compatible. Reports of problems are welcome.

• docker: docker
• ruby: dotenv
• node: dotenv + dotenv-expand
• python: python-dotenv
• php: phpdotenv
• go: godotenv
• rust: dotenv

Comparing Dialects

Specifies options for shdotenv. Currently, only dialect is supported. It is recommended that the dotenv dialect be specified with the dotenv directive. The .shdotenv setting is for personal use in projects where it is not allowed.

dialect: <DIALECT>

Example:

dialect: ruby

Note and reference: The FAQs present on motdotla's dotenv node project page and cdimascio's dotenv-java project page are so well done that I've included those that are relevant in the FAQs above.

A: Tenant III of the 12 factor app methodology states "The twelve-factor app stores config in environment variables". Thus, it is not recommended to provide the .env file to such environments. dotenv, however, is super useful in e.g a local development environment as it enables a developer to manage the environment via a file which is more convenient.

Using dotenv in production would be cheating. This type of usage, however is an anti-pattern.

No. We strongly recommend against committing your .env file to version control. It should only include environment-specific values such as database passwords or API keys. Your production database should have a different password than your development database.

By default, we will never modify any environment variables that have already been set. In particular, if there is a variable in your .env file which collides with one that already exists in your environment, then that variable will be skipped.

If instead, you want to override environment variables use the --overload option.

shdotenv --overload

We allows multiple .env files for convenience and interoperability with other dotenv tools. However, we believe that being able to use the same name in different .env files will lead to environment variables that are not "fully orthogonal" as The Twelve-Factor App outlines.

In a twelve-factor app, env vars are granular controls, each fully orthogonal to other env vars. They are never grouped together as “environments”, but instead are independently managed for each deploy. This is a model that scales up smoothly as the app naturally expands into more deploys over its lifetime.

– The Twelve-Factor App

If you want to override a previous definition, use the --overload option.