Split .env loading and process.env update

[djfo]

Hi everybody,

this is the first time I use the Discussions feature and also the first time I'm trying to participate in the development of the dotenv package.

I have a suggestion related to the interface/API of the package.

The package exports the functions parse and config, where parse parses the contents of a .env file and config updates process.env based on the key-value pairs read from the .env file.

Since config is implemented using parse, I hoped to also find a function that parses the .env file (taking a configuration as an argument like config) and returns a DotenvParseOutput, without updating process.env. Or alternatively, a function that loads the contents of .env (according to the configuration) and returns a string or buffer (essentially something that can be passed to parse in turn).

However, there is no such function.

I suggest to add this function, maybe named load, with the same interface as config, but not updating process.env:

export function load(options?: DotenvConfigOptions): DotenvConfigOutput;

It is easily implemented in code using the dotenv package, but I think it would make a nice addition to the package.

Rationale: I expect that a simple call to config is the main use case for this package. However, sometimes side-effect-free loading of .env (in the sense of only loading the key-value pairs without updating the environment) may be desirable.

Please let me know if such a change is welcome, I'd be happy to open a corresponding PR if so.

[motdotla]

@djfo this is a great suggestion. Thanks @djfo. Others have mentioned this as well.

It's something we will try to work on soon.

[djfo]

@motdotla I would like to open a PR if that's ok?

[motdotla]

Yes, please do. I will try to get to it in a reasonable time.

[motdotla]

@djfo this has been added to 16.1.0-rc2 and will soon be released as 16.1.0. See the populate function.

[motdotla]

here are docs: https://github.com/motdotla/dotenv#populate