exports-last #632

k15a · 2016-10-19T15:17:03Z

Hey,

I am currently working on #620 and would like some feedback.

Currently my implementation only checks for ES Modules export declaration but I would also like to support CommonJS modules as well. Does anybody has an idea how to check if an ExpressionStatement is an CommonJS export?

coveralls · 2016-10-19T15:53:05Z

Coverage increased (+0.05%) to 94.943% when pulling 7e00bb2 on k15a:master into 2cbd801 on benmosher:master.

benmosher · 2016-11-02T08:57:14Z

The ES module part looks good to me. 👍

I'd leave out CommonJS for now and add it later behind a flag (unless it's a use case you're looking forward to linting).

k15a · 2016-11-03T07:01:37Z

I've just added documentation for this rule. 📖

coveralls · 2016-11-03T07:05:29Z

Coverage increased (+0.1%) to 94.991% when pulling 037d32b on k15a:master into 2cbd801 on benmosher:master.

coveralls · 2016-11-12T12:29:59Z

Coverage increased (+0.05%) to 94.907% when pulling dc65882 on k15a:master into 90ef48b on benmosher:master.

k15a · 2016-11-12T12:50:42Z

I've resolved the merge conflicts but now travis ci is failing and I couldn't figure out why.

k15a · 2016-12-14T19:54:01Z

I've removed the CommonJS part. Let me know if I can do something else to get this merged.

coveralls · 2016-12-14T20:56:44Z

Coverage increased (+0.04%) to 94.904% when pulling 3fcdec1 on k15a:master into 90ef48b on benmosher:master.

benmosher

looks good to me! sorry this has been parked for so long. bad timing 😅

benmosher · 2016-12-15T12:04:29Z

LGTM, but would like to have one of @jfmengels or @ljharb take a quick look and sign off as well before merging 👍🏻

jfmengels

Hey @k15a! Thanks for the PR, and sorry for the delay in answering this.

This looks pretty good. I made a few comments, and would love a few more tests.

Support for the export all declaration (export * from 'foo'). Which, by the way, might be more interesting to allow that to be on top as it's also kind of an import (happy to discuss that further). It needs to be either tested or documented as an exception IMO.
More test cases:
- complex cases with plenty (3-5) of exports and some normal case in between
- empty file with no statements (trying to make sure this doesn't crash now or in the future)
- file with only exports

jfmengels · 2016-12-15T14:35:57Z

src/rules/exports-last.js

@@ -0,0 +1,35 @@
+function isExportStatement({ type }) {
+  // ES Module export statements
+  if (type === 'ExportDefaultDeclaration' || type === 'ExportNamedDeclaration') {


Simpler:

function isExportStatement({ type }) { return type === 'ExportDefaultDeclaration' || type === 'ExportNamedDeclaration'; }

And you should probably also test for ExportAllDeclaration (and add tests for that case), but happy to discuss that as an exception

jfmengels · 2016-12-15T14:36:49Z

src/rules/exports-last.js

+  create(context) {
+    return {
+      Program({ body }) {
+        const lastNonExportStatement = body.reduce((acc, node, index) => {


rename to lastNonExportStatementIndex

jfmengels · 2016-12-15T14:37:48Z

src/rules/exports-last.js

+
+        body.forEach((node, index) => {
+          if (isExportStatement(node) && index < lastNonExportStatement) {
+


remove this blank line (sorry about the nitpicking 😅)

jfmengels · 2016-12-15T14:41:22Z

src/rules/exports-last.js

+        }, 0)
+
+        body.forEach((node, index) => {
+          if (isExportStatement(node) && index < lastNonExportStatement) {


It might be a bit faster to do the index < lastNonExportStatement before the isExportStatement.

Or even better (?): only run the forEach on the statements after lastNonExportStatement.

// maybe add +/-1 to lastNonExportStatement to make it work right body.slice(lastNonExportStatement).forEach(node => { if (isExportStatement(node)) { // report } })

jfmengels · 2016-12-15T14:48:35Z

docs/rules/exports-last.md

+
+## When Not To Use It
+
+If you don't mind exports being sprinkled throughout a file, you may not want to enable this rule.


It would be nice to indicate that this is only for the export statements using the export keyword, and not for CommonJS module.exports/exports

k15a · 2016-12-15T16:23:03Z

Hey @jfmengels, I've added some more tests unfortunately one test is breaking but I think that's not an issue with my rule. I've also refactored the code. Let me know if I can do something else.

k15a · 2016-12-15T16:23:43Z

tests/src/rules/exports-last.js

+  valid: [
+    // Empty file
+    test({
+      code: '',


This line is breaking the tests.

Ok, try replacing it with a comment

test({ code: '// comment', })

and if that doesn't work out, just remove the test case.

jfmengels

Hey @k15a !
Sorry for the late reply, totally my fault...
I still have a few comments about the rule. If you're still up for making changes, that'd be great!

jfmengels · 2017-01-21T12:03:42Z

docs/rules/exports-last.md

@@ -0,0 +1,48 @@
+# exports-last
+
+This rule reports all export declaration which come before any non-export statements.


This rule enforces that all exports are declared at the bottom of the file. This rule will report any export declarations that comes before any non-export statements.

jfmengels · 2017-01-21T12:06:57Z

tests/src/rules/exports-last.js

+        export const even = 'count'
+        export const how = 'many'
+      `,
+      errors,


Shouldn't there be 2 errors here? One for the export default and one for export const so = 'many'?

jfmengels · 2017-01-21T12:09:32Z

tests/src/rules/exports-last.js

+
+const ruleTester = new RuleTester()
+
+const errors = ['Export statements should appear at the end of the file']


const error = { ruleId: 'exports-last', message: 'Export statements should appear at the end of the file', };

You can make this into an array by default but I think you'll need to make an array of multiple errors for some tests.

Would be nice if you could add:

either the location of the reported node

the type of the Node : type: 'ImportDefaultSpecifier. This might be easier.

At the moment, you are not reporting the export statements, but non-export statements. Any would be fine (though I' prefer reporting the export statements), but the message mentions the export statement, so let's report those. That's why I'd like to add this additional verification.

It could look something like:

const error = type => { ruleId: 'exports-last', message: 'Export statements should appear at the end of the file', type }; // ------ test({ code: ` export * from './foo' const bar = true `, errors: [error('ExportAllDeclaration')], }),

jfmengels · 2017-01-21T12:12:26Z

src/rules/exports-last.js

+  },
+}
+
+export default rule


Would you mind making the file structure (exports and stuff) look a bit more like the others, for consistency? Here's an example: https://github.com/benmosher/eslint-plugin-import/blob/master/src/rules/no-dynamic-require.js

jfmengels · 2017-01-21T12:13:32Z

CHANGELOG.md

@@ -26,6 +26,9 @@ This change log adheres to standards from [Keep a CHANGELOG](http://keepachangel
 - Properly report [`newline-after-import`] when next line is a decorator
 - Fixed documentation for the default values for the [`order`] rule ([#601])

+### Added
+- [`exports-last`] lints that export statements are at the end of the file ([#620] + [#632])


You'll need to rebase and move this back into the Unreleased section. Things have changed ever since you mad your PR 😅

If you don't know how to rebase, don't mind it, I'll do it later for you.

jfmengels · 2017-01-21T12:17:34Z

tests/src/rules/exports-last.js

+  valid: [
+    // Empty file
+    test({
+      code: '',


Ok, try replacing it with a comment

test({ code: '// comment', })

and if that doesn't work out, just remove the test case.

jfmengels · 2017-01-21T12:27:45Z

docs/rules/exports-last.md

+## This will not be reported
+
+```JS
+export const bool = true


Add a simple non-exported variable declaration up top, to show that that is allowed

k15a · 2017-01-25T15:31:22Z

Hey @jfmengels,

I think I now solved all the requested changes. Let me know if I can do something else to get this merged.

coveralls · 2017-01-25T16:26:57Z

Coverage increased (+0.04%) to 94.952% when pulling 0f1c691 on k15a:master into e137b98 on benmosher:master.

coveralls · 2017-01-25T16:37:56Z

Coverage increased (+0.04%) to 94.952% when pulling 0f1c691 on k15a:master into e137b98 on benmosher:master.

jfmengels

LGTM, thanks @k15a!

I'm giving the others a bit time to review this :)

coveralls · 2017-01-25T19:07:44Z

Coverage increased (+0.04%) to 94.952% when pulling 26cc994 on k15a:master into e137b98 on benmosher:master.

ljharb · 2017-03-01T18:59:02Z

@k15a please rebase latest master, so as to remove all merge commits

exports-last will check that all export statements are at the end of the file Closes #620

k15a · 2017-03-01T19:12:28Z

I tried my best. Did this removed the merge commits?

ljharb

Rebase looks good except for one thing :-)

ljharb · 2017-03-01T19:14:26Z

CHANGELOG.md

@@ -6,8 +6,8 @@ This change log adheres to standards from [Keep a CHANGELOG](http://keepachangel
 ## [Unreleased]
 ### Added
 - [`no-anonymous-default-export`] rule: report anonymous default exports ([#712], thanks [@duncanbeevers]).
- Add new value to [`order`]'s `newlines-between` option to allow newlines inside import groups ([#627], [#628], thanks [@giodamelio])
- Add `count` option to the [`newline-after-import`] rule to allow configuration of number of newlines expected ([#742], thanks [@ntdb])


we lost one of the changelog entries - let's fix that :-)

k15a · 2017-03-01T19:22:08Z

Fixed that as well. Hope everything is ok now.

@benmosher

looks good; deferring to @benmosher and @jfmengels

alexkrolick · 2017-05-16T21:00:52Z

I'd like to use this rule - anyone want to take a look at it again? Looks reeaaallly close to being mergeable.

k15a mentioned this pull request Oct 22, 2016

Rule Proposal: export-style #637

Closed

benmosher approved these changes Dec 15, 2016

View reviewed changes

jfmengels requested changes Dec 15, 2016

View reviewed changes

k15a commented Dec 15, 2016

View reviewed changes

jfmengels mentioned this pull request Jan 21, 2017

New: group-exports rule #721

Merged

jfmengels requested changes Jan 21, 2017

View reviewed changes

jfmengels approved these changes Jan 25, 2017

View reviewed changes

k15a added 7 commits March 1, 2017 20:02

Add rule exports-last

84fd4f2

exports-last will check that all export statements are at the end of the file Closes #620

Add documentation for exports-last

76f1b84

Remove CommonJS exports

0de28bd

Refactor exports-last rule

0611e21

Add ES6 only hint to the docs

531d04d

Add more tests for exports-last

baa585d

Refactor exports-last

8c0e7d6

ljharb previously requested changes Mar 1, 2017

View reviewed changes

Fix Changelog

8d4e25e

removing changelog note, will re-add

1ba1c3a

benmosher merged commit 92dd662 into import-js:master Aug 27, 2017

benmosher added a commit that referenced this pull request Aug 27, 2017

changelog notes for #620 + #632

5f5fbf6

This was referenced Apr 15, 2018

Update eslint-plugin-import to 2.12.0 sbardian/DebtTrackerAPI#25

Merged

Update dependency eslint-plugin-import to v2.11.0 Urigo/tortilla#68

Merged

This comment has been minimized.

Sign in to view


		body.forEach((node, index) => {
		if (isExportStatement(node) && index < lastNonExportStatement) {


		## When Not To Use It

		If you don't mind exports being sprinkled throughout a file, you may not want to enable this rule.

		@@ -0,0 +1,48 @@
		# exports-last

		This rule reports all export declaration which come before any non-export statements.


		const ruleTester = new RuleTester()

		const errors = ['Export statements should appear at the end of the file']

exports-last #632

exports-last #632

Conversation

k15a commented Oct 19, 2016

coveralls commented Oct 19, 2016 • edited Loading

benmosher commented Nov 2, 2016

k15a commented Nov 3, 2016

coveralls commented Nov 3, 2016 • edited Loading

coveralls commented Nov 12, 2016 • edited Loading

k15a commented Nov 12, 2016

k15a commented Dec 14, 2016

coveralls commented Dec 14, 2016 • edited Loading

benmosher left a comment

Choose a reason for hiding this comment

benmosher commented Dec 15, 2016

jfmengels left a comment • edited Loading

Choose a reason for hiding this comment

jfmengels Dec 15, 2016 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

k15a commented Dec 15, 2016

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jfmengels left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

k15a commented Jan 25, 2017 • edited Loading

coveralls commented Jan 25, 2017 • edited Loading

coveralls commented Jan 25, 2017 • edited Loading

jfmengels left a comment

Choose a reason for hiding this comment

coveralls commented Jan 25, 2017 • edited Loading

ljharb commented Mar 1, 2017

k15a commented Mar 1, 2017

ljharb left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

k15a commented Mar 1, 2017

alexkrolick commented May 16, 2017 • edited Loading

This comment has been minimized.

coveralls commented Oct 19, 2016 •

edited

Loading

coveralls commented Nov 3, 2016 •

edited

Loading

coveralls commented Nov 12, 2016 •

edited

Loading

coveralls commented Dec 14, 2016 •

edited

Loading

jfmengels left a comment •

edited

Loading

jfmengels Dec 15, 2016 •

edited

Loading

k15a commented Jan 25, 2017 •

edited

Loading

coveralls commented Jan 25, 2017 •

edited

Loading

coveralls commented Jan 25, 2017 •

edited

Loading

coveralls commented Jan 25, 2017 •

edited

Loading

alexkrolick commented May 16, 2017 •

edited

Loading