Fix #389: Allow secondary alphabetical sorting #629

randallreedjr · 2016-10-17T22:25:44Z

After sorting imports by group, allow sorting alphabetically within a group
import/order rule now accepts sort-order option that may be 'ignore' (default) or 'alphabetical'
Fixes [import/order] sort order within groups #389

randallreedjr · 2016-10-17T22:26:32Z

This is my first attempt at trying to implement a secondary alphabetical sort on imports. I would appreciate any feedback!

coveralls · 2016-10-18T00:54:14Z

Coverage decreased (-0.04%) to 94.86% when pulling aa7aca1 on randallreedjr:import-sort-order into d9605a0 on benmosher:master.

coveralls · 2016-10-18T02:51:04Z

Coverage decreased (-0.05%) to 94.85% when pulling a2a80b5 on randallreedjr:import-sort-order into d9605a0 on benmosher:master.

coveralls · 2016-10-18T03:20:19Z

Coverage increased (+0.04%) to 94.935% when pulling 4cbdbb8 on randallreedjr:import-sort-order into d9605a0 on benmosher:master.

randallreedjr · 2016-10-18T03:26:26Z

I resolved the test coverage decrease. Not sure why appveyor is failing, but it appears to be happening when trying to run coveralls.

jfmengels

Hey @randallreedjr! Thanks a lot for working on this.

This is a pretty good start, but I have quite a few request changes. Let me know if you disagree, need explanations or guidance :)

jfmengels · 2016-10-18T21:26:39Z

docs/rules/order.md

@@ -141,6 +141,42 @@ import index from './';
 import sibling from './foo';
 ```

+### `sort-order: [ignore|alphabetical]`:


Bikeshedding: maybe simply call this sort?

jfmengels · 2016-10-18T21:39:28Z

src/rules/order.js

@@ -205,6 +269,7 @@ module.exports = {
      },
      'Program:exit': function reportAndReset() {
        makeOutOfOrderReport(context, imported)
+        makeOutOfAlphabeticalOrderReport(context, imported, sortOrder)


Wrap this in a if (sortOrder === 'alphabetical') {} and remove the latter comparison to ignore. This will make it clearer when this will be called, and comparing it to 'alphabetical' will prevent surprises when we add other sort cases later on. Also, you won't have to pass this parameter around.

Another separate thing:
I think it would be nice to to not re-report the modules that were reported in makeOutOfOrderReport.

Let's assume lowercase to be part of group 1, uppercase of group 2, and group 1 should appear before group 2.

import 'B'; import 'A'; import 'b'; import 'a';

I haven't tried it yet, but I wouldn't be surprised if there were two error reports for moving B to a latter location or something. I would love some tests on that too, though the problem might not be shown with this example.

The reason I'm worried about this, is when you introduce the order rule to a new codebase, where imports are sorted in random ways, you'll get a lot of errors for the import section. I'd like for it to be the least confusing people and avoid having 20 errors when you required 10 modules. Note that the code might actually do this already, but without tests I have trouble being sure.

By not re-reporting modules, I think findOutOfAlphabeticalOrder could be simplified a bit too. I'm a bit confused about its inner workings at the moment.

Also, with only the makeOutOfOrderReport report, you have a consistent should occur before/should occur after message. I think it would be nice if that was true for all of the reports. Maybe compile the whole list with the group and the sort options one way, then do it the other way and take the shorter one.

Sorry to add a lot of work 😅

jfmengels · 2016-10-18T21:40:53Z

src/rules/order.js

@@ -56,6 +101,21 @@ function makeOutOfOrderReport(context, imported) {
  reportOutOfOrder(context, imported, outOfOrder, 'before')
 }

+function makeOutOfAlphabeticalOrderReport(context, imported, sortOrder) {
+  const outOfOrder = findOutOfAlphabeticalOrder(imported, sortOrder, 'a-z')
+  if (!outOfOrder.length) {


if (outOfOrder.length === 0) {}

jfmengels · 2016-10-18T21:43:07Z

src/rules/order.js

+    reportOutOfAlphabeticalOrder(context, reversedImported, reversedOrder, 'after', 'z-a')
+    return
+  }
+  reportOutOfAlphabeticalOrder(context, imported, outOfOrder, 'before', 'a-z')


Is there maybe a way to refactor this function and makeOutOfOrderReport, they are really similar. Forget it if you deem that it then becomes harder to read.

jfmengels · 2016-10-18T21:47:00Z

src/rules/order.js

+    return firstName > secondName
+  } else {
+    throw new Error('Invalid alphabetical direction' + direction)
+  }


function alphabetical(firstName, secondName, direction) { if (direction === 'a-z') { return firstName < secondName } return firstName > secondName }

You won't ever enter the last case, so no need to account for it.

jfmengels · 2016-10-18T22:07:08Z

src/rules/order.js

@@ -17,6 +17,16 @@ function reverse(array) {
  }).reverse()
 }

+function alphabetical(firstName, secondName, direction) {
+  if (direction === 'a-z') {
+    return firstName < secondName


The same module could be imported twice, so this should be <= and >=. Please add a test for that case :)

jfmengels · 2016-10-18T22:07:44Z

src/rules/order.js

@@ -17,6 +17,16 @@ function reverse(array) {
  }).reverse()
 }

+function alphabetical(firstName, secondName, direction) {


Name it something like compareAlphabetically? Or isBeforeAlphabetically (berk). They're ugly names, so if you have better ideas, go for it, but at the moment, without looking at the code and by just looking at places where it's called, it's hard to know what exactly this function does.

isAlphabetized?

jfmengels · 2016-10-18T22:16:47Z

src/rules/order.js

+      direction
+    )
+    const reversedDirection = direction.split('').reverse().join('')
+    if (alphabetical(importedModule.name, maxSeenAlphabeticalNode.name, reversedDirection)) {


Unless the two names are the same, isn't this just equal to resAlphabetical?

coveralls · 2016-10-19T02:18:27Z

Coverage increased (+0.04%) to 94.94% when pulling 8eeb761 on randallreedjr:import-sort-order into d9605a0 on benmosher:master.

coveralls · 2016-10-21T17:02:48Z

Coverage decreased (-1.3%) to 93.588% when pulling ba90260 on randallreedjr:import-sort-order into 2cbd801 on benmosher:master.

coveralls · 2016-10-23T22:22:49Z

Coverage increased (+0.04%) to 94.949% when pulling 58da14f on randallreedjr:import-sort-order into edbb570 on benmosher:master.

randallreedjr · 2016-10-24T00:14:37Z

@jfmengels I reworked the approach to solving this problem, please have a look and let me know what you think!

coveralls · 2016-10-24T00:17:08Z

Coverage increased (+0.01%) to 94.919% when pulling 15289cc on randallreedjr:import-sort-order into edbb570 on benmosher:master.

randallreedjr · 2016-11-03T16:48:57Z

Anything I can do to help with the review of this PR?

coveralls · 2016-11-03T18:05:15Z

Coverage increased (+0.01%) to 94.963% when pulling d2b0921 on randallreedjr:import-sort-order into 4744468 on benmosher:master.

ctavan · 2016-11-25T16:13:09Z

This is definitely an addition I've been looking for! It would be great if it could get merged at some point.

@randallreedjr just one remark: I suppose the yarn.lock was commited by accident, wasn't it?

randallreedjr · 2016-12-05T16:11:30Z

@ctavan Good catch, thanks!

coveralls · 2016-12-05T17:06:14Z

Coverage increased (+0.01%) to 94.875% when pulling 2233e97 on randallreedjr:import-sort-order into bfdc2bb on benmosher:master.

coveralls · 2016-12-05T17:12:45Z

Coverage increased (+0.01%) to 94.875% when pulling 2233e97 on randallreedjr:import-sort-order into bfdc2bb on benmosher:master.

coveralls · 2016-12-05T18:18:18Z

Coverage increased (+0.01%) to 94.875% when pulling c0dea5e on randallreedjr:import-sort-order into bfdc2bb on benmosher:master.

randallreedjr · 2016-12-05T19:29:55Z

@benmosher @jfmengels This PR has been open for 7 weeks now. Is there anything I can do to help get it merged in?

snoob · 2018-06-28T23:12:48Z

@ljharb too late :p i have already rebased the @randallreedjr on my repo branch https://github.com/snoob/eslint-plugin-import/tree/import-sort-order you can pull them :) i changed my mind i ll be more reactive if this PR is synchronized to my repo

janpaul123 · 2018-07-10T22:24:35Z

This PR adds an optional configuration that a lot of people want. I don't quite understand the controversy?

ljharb · 2018-07-10T22:27:56Z

@janpaul123 there's no controversy; it's just that the number of people that want an option is irrelevant if that option can't be provided correctly.

ljharb

To reiterate, the changes requested here are still needed for this to go in.

ljharb · 2018-07-10T22:32:23Z

@snoob thanks for the help, i've used your rebase to update this PR.

janpaul123 · 2018-07-12T17:32:09Z

@ljharb cool, makes sense. I think it is good to fix problems with this rule before adding more functionality to it, thanks for reiterating that. 👍

I unfortunately really needed this functionality though, and also the autofixing (which this PR doesn't include), so for anyone who comes across this thread and needs that too, I made a small package based on another PR over here: https://github.com/janpaul123/eslint-plugin-import-order-alphabetical It's not perfect but it suited my needs well enough. 🤷‍♂️

snoob · 2018-07-13T07:31:03Z

@ljharb : What do you think about @janpaul123 work ?

Because we have to fix test on this branch and we don't have autofix also
So what do we do ? Do we provide a native solution on this repo ? or we assume that you need an additional plugin for alphabetical sorting ?

ljharb · 2018-07-16T06:11:22Z

@snoob any time you have commits you want me to pull into this PR, please ping me here and I'll do so.

cdreve · 2018-11-13T13:56:58Z

Hi guys,
Any updates on this PR? I would really like to use this feature.
Thank you.

@janpaul123 @ljharb @snoob @randallreedjr

janpaul123 · 2018-11-13T23:27:05Z

@cdreve you can use https://github.com/janpaul123/eslint-plugin-import-order-alphabetical

snoob · 2018-11-14T08:25:58Z

Sorry i haven't time to continue my work on this PR :(

itajaja · 2019-02-13T19:52:31Z

hi @ljharb, I'll be happy to pick up the torch from where it was left. by following the conversation, it seems like the main problem is interleaved code between import groups, but as far as I can tell, that seems resolved. is it something in here specifically that adds a regression? happy to open a new PR for the additional work, or for you to cherry pick from my branch.

ljharb · 2019-02-23T00:27:58Z

@itajaja posting a link here would be preferable to opening a new PR; indeed, if the interleaving concerns mentioned upthread are covered by tests, this PR, rebased, might be sufficient.

It's a bit tricky right now, though, because tests are failing on master.

jeffshaver · 2019-03-11T01:29:05Z

@ljharb also interested in this. let me know if any help is needed getting this done

VincentLanglet · 2019-04-19T14:26:51Z

I think #1105 will do the same job

Glinkis · 2019-08-19T09:58:44Z

Has there been any update on this? I would love if this was thing.

dismissing

rjhilgefort · 2019-12-07T03:59:01Z

Whoa! It happened! Pumped to be getting this feature- everyone that contributed, thanks!

randallreedjr · 2019-12-07T06:52:28Z

Wow, I can't believe it! This was one of my earliest open source pull requests. Thanks @ljharb for merging and everyone else for helping to get this in!!!

…n groups Fixes import-js#1406. Fixes import-js#389. Closes import-js#629. Closes import-js#1105. Closes import-js#1360. Co-Authored-By: dannysindra <als478@cornell.edu> Co-Authored-By: Radim Svoboda <radim.svoboda@socialbakers.com> Co-Authored-By: Soma Lucz <luczsoma@gmail.com> Co-Authored-By: Randall Reed, Jr <randallreedjr@gmail.com> Co-Authored-By: Jordan Harband <ljharb@gmail.com>

techsin · 2019-12-19T16:59:45Z

Just to be sure, this only sorts based on path of import NOT on variable being imported.. So import { prop }..., import Abc..., import *... , etc will have no effect.

randallreedjr force-pushed the import-sort-order branch from aa7aca1 to a2a80b5 Compare October 18, 2016 02:23

randallreedjr force-pushed the import-sort-order branch from a2a80b5 to 4cbdbb8 Compare October 18, 2016 03:02

jfmengels requested changes Oct 18, 2016

View reviewed changes

randallreedjr force-pushed the import-sort-order branch from 8eeb761 to ba90260 Compare October 21, 2016 16:54

randallreedjr force-pushed the import-sort-order branch 2 times, most recently from 8eeb761 to 58da14f Compare October 23, 2016 22:18

randallreedjr force-pushed the import-sort-order branch from 58da14f to 15289cc Compare October 24, 2016 00:12

randallreedjr force-pushed the import-sort-order branch from 15289cc to d2b0921 Compare November 3, 2016 16:56

randallreedjr force-pushed the import-sort-order branch from d2b0921 to bca28f3 Compare December 5, 2016 16:10

randallreedjr force-pushed the import-sort-order branch 2 times, most recently from 2233e97 to c0dea5e Compare December 5, 2016 16:18

randallreedjr mentioned this pull request Dec 9, 2016

[import/order] sort order within groups #389

Closed

ljharb force-pushed the import-sort-order branch from 940939c to 6a4156d Compare July 10, 2018 22:31

ljharb requested changes Jul 10, 2018

View reviewed changes

randycoulman mentioned this pull request Mar 1, 2019

Use eslint-plugin-import-order-alphabetical in place of import/order rule CodingZeal/eslint-config-zeal#125

Closed

luczsoma mentioned this pull request Aug 23, 2019

Alphabetize imports within groups #1360

Merged

ljharb force-pushed the import-sort-order branch from 6a4156d to 8224e51 Compare December 7, 2019 00:29

ljharb requested a review from jfmengels December 7, 2019 00:32

ljharb approved these changes Dec 7, 2019

View reviewed changes

ljharb closed this in 8224e51 Dec 7, 2019

ljharb merged commit 8224e51 into import-js:master Dec 7, 2019

ljharb mentioned this pull request Dec 9, 2019

import/order: alphabetize doesn't work correctly with group & newlines-between #1561

Closed

Fix #389: Allow secondary alphabetical sorting #629

Fix #389: Allow secondary alphabetical sorting #629

Conversation

randallreedjr commented Oct 17, 2016

randallreedjr commented Oct 17, 2016

coveralls commented Oct 18, 2016 • edited

coveralls commented Oct 18, 2016 • edited

coveralls commented Oct 18, 2016 • edited

randallreedjr commented Oct 18, 2016

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

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

coveralls commented Oct 19, 2016 • edited

coveralls commented Oct 21, 2016 • edited

coveralls commented Oct 23, 2016 • edited

randallreedjr commented Oct 24, 2016

coveralls commented Oct 24, 2016 • edited

randallreedjr commented Nov 3, 2016

coveralls commented Nov 3, 2016 • edited

ctavan commented Nov 25, 2016

randallreedjr commented Dec 5, 2016

coveralls commented Dec 5, 2016 • edited

coveralls commented Dec 5, 2016 • edited

coveralls commented Dec 5, 2016 • edited

randallreedjr commented Dec 5, 2016

snoob commented Jun 28, 2018 • edited

janpaul123 commented Jul 10, 2018

ljharb commented Jul 10, 2018

ljharb left a comment

Choose a reason for hiding this comment

ljharb commented Jul 10, 2018

janpaul123 commented Jul 12, 2018

snoob commented Jul 13, 2018 • edited

ljharb commented Jul 16, 2018

cdreve commented Nov 13, 2018

janpaul123 commented Nov 13, 2018

snoob commented Nov 14, 2018 • edited

itajaja commented Feb 13, 2019

ljharb commented Feb 23, 2019

jeffshaver commented Mar 11, 2019

VincentLanglet commented Apr 19, 2019

Glinkis commented Aug 19, 2019

rjhilgefort commented Dec 7, 2019 • edited

randallreedjr commented Dec 7, 2019

techsin commented Dec 19, 2019

coveralls commented Oct 18, 2016 •

edited

coveralls commented Oct 18, 2016 •

edited

coveralls commented Oct 18, 2016 •

edited

coveralls commented Oct 19, 2016 •

edited

coveralls commented Oct 21, 2016 •

edited

coveralls commented Oct 23, 2016 •

edited

coveralls commented Oct 24, 2016 •

edited

coveralls commented Nov 3, 2016 •

edited

coveralls commented Dec 5, 2016 •

edited

coveralls commented Dec 5, 2016 •

edited

coveralls commented Dec 5, 2016 •

edited

snoob commented Jun 28, 2018 •

edited

snoob commented Jul 13, 2018 •

edited

snoob commented Nov 14, 2018 •

edited

rjhilgefort commented Dec 7, 2019 •

edited