Throw an error or output a warning when `--env-file` can not be opened.

[virtuallyunknown]

What is the problem this feature will solve?

Greetings! Thanks a lot for implementing --env-file directly into Node.

I am currently using Node v20.9.0 on Linux, and if you run node with the --env-file option that points to a file that can't be reached, it still will execute the program as if everything went fine. For example:

node --env-file=non_existing_file.env my_app.js

What is the feature you are proposing to solve the problem?

This is not an uncommon scenario that you can find yourself in, and it can occur for various reasons. For example if you are working in a monorepository and the context of the current working directory changes, or if you lack sufficient permissions to open the file.

I think it would be beneficial for developers if an error or a warning was shown if the env file can't be opened.

What alternatives have you considered?

I have been using the dotenv package and currently in the process of migrating to native --env-file instead.

[sunil-kumarr]

@anonrig can you assign this issue to me will like to work on this as my first OSS contribution

[arxngr]

hello there,

I want to take this issue :) , and i would like to discuss an issue and some concerns I have. Specifically, I am wondering what my approach should be if the --env-file is not present. we should return an error because the environment file is important. wdyt?

Additionally, I have another question. What if the user wants to use multiple environment files but one of them is missing? Should we provide a warning and continue the process or stop the process? Any ideas are welcome. :)

cc @anonrig

[virtuallyunknown]

In my personal opinion, it should throw an error. This will be consistent with how the FileSystem module works, for example if you fs.readFile a non-existing file, it will throw. Another example would be the Node CLI, if you call node --watch non_existing_file.js it will also throw.

[arxngr]

Seems good, I will create the PR soon for this issue

[Simplisticbubble]

Hi, I would like to be involved in solving this issue! (my first issue)

[arxngr]

Cool! Thanks for being interested in this issue, but unfortunately I am already working on this :)

[Simplisticbubble]

Sorry, this was my first time looking at making a contribution. mb

…

On Mon, 13 Nov 2023 at 12:54 pm, Ardi Nugraha ***@***.***> wrote: Cool! Thanks for being interested in this issue, but unfortunately I am already working on this :) — Reply to this email directly, view it on GitHub <#50536 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AZPH7EEGNMZBFITFI4QUHMTYEF4XFAVCNFSM6AAAAAA64FR27KVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQMBXGM3DEMZTGY> . You are receiving this because you commented.Message ID: ***@***.***>

[dimonachi]

Hi, do you need some help with this?

[40oleg]

@virtuallyunknown
I suppose this issue can be closed because of the problem was solved in pr above