fix(deps): drop a few unnecessary dependencies (#643) #654
Conversation
| return ( | ||
| literal: TemplateStringsArray, | ||
| ...placeholders: string[] | ||
| ): [string, Map<string, string[]>] => { | ||
| const code = dedent(literal, ...placeholders); | ||
| const [shaken, deps] = shake( | ||
| babel, |
There was a problem hiding this comment.
what's the benefit of injecting Babel here? I don't think we'll inject various versions of Babel and it stays as dependency, but peer. I think we should be fine by leaving these functions as they were – dependent on @babel/core. This would shrink the diff quite a bit :)
There was a problem hiding this comment.
Hm… We don't touch the original AST in evaluators, so it's probably safe to use imported babel here. I'll take a close look.
docs/HOW_IT_WORKS.md
Outdated
| @@ -174,6 +174,7 @@ Sometimes it can be useful to implement your own strategy (it can be just a mock | |||
|
|
|||
| ```typescript | |||
| type Evaluator = ( | |||
| babel: BabelCore, | |||
There was a problem hiding this comment.
no need for this change imho, as stated somewhere earlier
There was a problem hiding this comment.
Looks like this is the last thing to revert now :)
Anber
changed the title
Anber
changed the title
|
Why do we need type-only imports? |
|
@wereHamster mostly because of linter: eslint doesn't complain about type-only imports if dependencies aren't specified in package.json. Another reason is improved readability, that helps to manage dependencies. |
docs/HOW_IT_WORKS.md
Outdated
| @@ -174,6 +174,7 @@ Sometimes it can be useful to implement your own strategy (it can be just a mock | |||
|
|
|||
| ```typescript | |||
| type Evaluator = ( | |||
| babel: BabelCore, | |||
There was a problem hiding this comment.
Looks like this is the last thing to revert now :)
Motivation
We have some dependencies which we don't really use or can avoid using: webpack, @babel/core, etc.
Another problem here, that Linaria babel plugin uses an external version of babel instead of a version which launches the plugin, that can probably cause some tricky errors.
Summary
Despite the fact that PR touches a lot of files, the most changes are just
typeannotations for imports and linted files (update of eslint was required for supportingtype). The real change here is passingbabelthrough call stacks in the plugin, but it doesn't affect the logic.There are also three minor fixes:
expressionsToEvaluateandoriginalLazyExpressions;Results:
Test plan
Since the logic wasn't affected, all existed tests are still working.