no-duplicates: fix false positive

[steved-stripe]

fixes #231

This PR removes the auto fix for:

import {AValue, type x, BValue} from './foo';
import {type y} from './foo'

as our heuristic for type vs non-type imports has changed slightly.

The second import would be considered a type-only import, and the first import would be considered a value import, so we aren't combining them.

We could make this more sophisticated if we wanted to but this fixes the bad autofix that would change:

import {
  type bar,
  buzz,
} from 'foo';
import type {bizz} from 'foo';

into:

import type {
  type bar,
  buzz,
  bizz,
} from 'foo';

related: import-js#2697

[changeset-bot]

⚠️ No Changeset found

Latest commit: 098ee4c

Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.

This PR includes no changesets

When changesets are added to this PR, you'll see the packages that this PR includes changesets for and the associated semver types

Click here to learn what changesets are, and how to add one.

Click here if you're a maintainer who wants to add a changeset to this PR

[codesandbox-ci]

This pull request is automatically built and testable in CodeSandbox.

To see build info of the built libraries, click here or the icon next to each commit SHA.

[SukkaW]

Just wondering, will import-x/consistent-type-specifier-style (along with its autofixes) work for you? This means, after you get a "not-ideal" from the no-duplicates output you can then get the right one from import-x/consistent-type-specifier-style.

[steved-stripe]

hmm I'm not sure, but it would be a significant amount of work to migrate the codebase to a consistent import style, there's a lot of various import styles currently

I think the current setup of the autofix resulting in broken builds is what I'm trying to avoid!

[JounQin]

But this PR is not a correct fix, it's just ignoring issues.

[steved-stripe]

Yeah fair enough, need to think about this more. I've updated the test cases with a new failing test that exercise the issue I ran into