add: add command 'add' that adds dependents to dependents.txt

[amishhaa]

Part of #1528.

Add command 'add' to write dependents to dependents.txt of workspace.
Usage:
grater add foo/bar --file <file_path>
Update related usage mentions in rootUsage and README to mention this command.

Add AddDependent method in workspace.go as the simplest method to write a raw string to dependents.txt in the workspace.
Add AddDependents method in internal/add.go to write each dependent using ws instance and AddDependentsFromFile method to parse a file and pass it to AddDependents.
Call AddDependentsFromFile and AddDependents from cmdAdd to write dependents.
Add GetDependents to retrieve the written dependents from ws instance.
Add internal.go file to write usage comment for header package internal.

[codecov]

Codecov Report

❌ Patch coverage is 82.60870% with 8 lines in your changes missing coverage. Please review.
✅ Project coverage is 64.43%. Comparing base (ce4cbf8) to head (6f5b90f).
⚠️ Report is 1 commits behind head on main.

Files with missing lines	Patch %	Lines
grater/cmd/add.go	76.47%	2 Missing and 2 partials ⚠️
grater/internal/add.go	75.00%	2 Missing and 2 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1702      +/-   ##
==========================================
+ Coverage   64.19%   64.43%   +0.24%     
==========================================
  Files          69       71       +2     
  Lines        3463     3509      +46     
==========================================
+ Hits         2223     2261      +38     
- Misses        998     1002       +4     
- Partials      242      246       +4     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[mx-psi]

Nit: Cobra's examples seem to move this to a top-level variable, see e.g. https://cobra.dev/docs/how-to-guides/working-with-flags/#make-a-flag-required

This pattern is fine, but I would rather follow what Cobra does in its examples

[mx-psi]

If you want to be a bit more compact you can do the following

Suggested change

	err := internal.AddDependentsFromFile(path)
	if err != nil {
	return err
	}
	if err := internal.AddDependentsFromFile(path); err != nil {
	return err
	}

[mx-psi]

Same here!

Suggested change

	err := internal.AddDependents(args)
	if err != nil {
	return err
	}
	if err := internal.AddDependents(args); err != nil {
	return err
	}

[mx-psi]

You can define the variables implicitly and Go will figure the types out

Suggested change

	var err error
	var out string
	out, err = runCobra(t, "add", "--help")
	out, err := runCobra(t, "add", "--help")

[mx-psi]

Same here! (only use := on the first out, err definition, then you use =)

[mx-psi]

I think it would be more convenient if you add the dependents to an internal list and then have a CommitToFile method that writes them to the file. Or you could even pass a dependents []string directly to this function.

There are two reasons for this:

1. It is more performant: you avoid closing and re-opening the file every time and in general doing bigger writes rather than many smaller writes is more performant. The programming language and the operating system may do some of this batching into big writes for you, but it is good practice to do it yourself when it makes sense.
2. If at some point you need to add more structure to your list of dependents, it is much easier and more performant to do so with some Go struct that you modify than reading the file and writing to it again. Reading from your program's struct is much more performant than reading from the disk.

[mx-psi]

I think it would be better to write your list of dependents on JSON format. Right now you only need to store each dependent on the list, but if you need to add multiple data about each dependent, it will become harder.

For example, instead of writing each item into a new line, you can define a struct like:

type Dependent struct {
   Path string `json:"path"`
}

and use the encoding/json/v2 package to marshal and unmarshal it. You can see some examples on how to use it here.

This will become easier to do if you address my comment above.