An alphabetical order rule?

[fa93hws]

Hi!
I found that the alphabetical order import rule is very useful.
Do you mind I add a new rule for that?
Or an option to existed order rule?

[ljharb]

Why is it useful? What rule are you referring to?

Either way it’d be an option to the existing rule, not a new one.

[fa93hws]

Hi, we are a typescript team and are migrating to eslint from the tslint.
While the import order rule is quite different between that in eslint and the tslint.
I think it might be useful for other team who migrated from tslint as well.

Thanks, I will make such a rule when I have time.

[ljharb]

Closing as a duplicate of #1311.

[gustavopch]

Closing as a duplicate of #1311.

How is this is a duplicate of #1311? Alphabetical order is only mentioned there, but it's not the point of the issue.

[sveyret]

I also like very much the alphabetical order of the tslint rule, because when there are lot of imports in a file, it is far easier to find if one needs to be added or not. Probably should it be an option of the already existing order rule. What would be great also, to conform to tslint rule, would be to be able to sort imported members from a library by alphabetical order, and to have the ability to put lowercase starting members at the end. See https://palantir.github.io/tslint/rules/ordered-imports/

Would you accept PR for that?

[caribou-code]

This would be a useful addition to the package. For example in a React project we can already sort props, propTypes, defaultProps and more alphabetically with rules such as react/jsx-sort-default-props, react/jsx-sort-props and react/sort-prop-types.

For unordered lists of code such as imports, exports and objects, it makes sense to group and sort these so there's some consistent order throughout a project. import/order already handles the grouping, but being able to sort alphabetically within those groups would be beneficial.

[ljharb]

One challenge is that bindingless imports (import 'path' versus import x from 'path') are by convention side-effecting, so nothing should be reordered across them, and they should not be reordered themselves.

However, yes, after some thought I think it does make sense to add this as an option to the order rule. A PR would be appreciated.