lib: port errors to new system #19034
Conversation
| fail('windowsHide', __dirname, invalidArgTypeError); | ||
| fail('windowsHide', [], invalidArgTypeError); | ||
| fail('windowsHide', {}, invalidArgTypeError); | ||
| fail('windowsHide', common.mustNotCall(), invalidArgTypeError); |
There was a problem hiding this comment.
Is this not failing? I expected that the invalidArgTypeError call number has to be changed otherwise?
There was a problem hiding this comment.
Possible. It was failing because I forgot to update it, but then I forgot to run it again
| @@ -74,7 +74,7 @@ const re = /^The "id" argument must be of type string\. Received type \w+$/; | |||
| common.expectsError( | |||
| () => { require(''); }, | |||
| { | |||
| type: Error, | |||
| type: RangeError, | |||
There was a problem hiding this comment.
Without changing the errors that result in such changes, the PR would be a patch. I personally prefer a PR that is possible to backport instead of having the PR as semver-major because just very few cases that were wrong before.
There was a problem hiding this comment.
I can remove the semver-major changes, but I think a backport would have to be done from scratch anyway, even for v9.x, because many error codes don't exist in any release line
There was a problem hiding this comment.
Most errors were already ported in 9.x as far as I know. I personally still think it would be better to have this as a semver-patch but I am fine with it if others want to land it as is.
lib/module.js
Outdated
| throw new errors.Error('ERR_INVALID_ARG_VALUE', | ||
| 'id', id, 'must be a non-empty string'); | ||
| throw new ERR_INVALID_ARG_VALUE.RangeError('id', id, | ||
| 'must be a non-empty string'); |
There was a problem hiding this comment.
To me it seems like this is a TypeError, not a RangeError. Don't you think so?
There was a problem hiding this comment.
The type is checked just before. I don't feel strongly about it but it seemed right to throw a RangeError: the must be greater than zero
There was a problem hiding this comment.
Well, I guess it is somewhat of a edge case. I do not feel strongly about it either.
There was a problem hiding this comment.
I think usually we just go with ERR_OUT_OF_RANGE for this sort of generic range errors..probably makes more sense if it's ERR_INVALID_ARG_VALUE.Error on id, or a ERR_OUT_OF_RANGE on id.length
There was a problem hiding this comment.
I changed it to a TypeError. @joyeecheung are you OK with that?
| @@ -74,7 +74,7 @@ const re = /^The "id" argument must be of type string\. Received type \w+$/; | |||
| common.expectsError( | |||
| () => { require(''); }, | |||
| { | |||
| type: Error, | |||
| type: RangeError, | |||
There was a problem hiding this comment.
Most errors were already ported in 9.x as far as I know. I personally still think it would be better to have this as a semver-patch but I am fine with it if others want to land it as is.
|
Rebased again. Hopefully last CI: https://ci.nodejs.org/job/node-test-pull-request/13484/ |
|
As a question: shall we squash this or land each on it's own? I am fine both ways, I just thought I ask because it is a lot of commits. |
|
I'm for squashing |
|
And another CI: https://ci.nodejs.org/job/node-test-pull-request/13509/ There is no single green result but no failure seem related or reproducible so I'm going to land this. |
This is a first batch of updates that touches non-underscored modules in lib. PR-URL: #19034 Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Ruben Bridgewater <ruben@bridgewater.de> Reviewed-By: Joyee Cheung <joyeec9h3@gmail.com>
|
Squashed and landed in 1e8d110 |
This is a first batch of updates that touches non-underscored modules in lib. PR-URL: nodejs#19034 Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Ruben Bridgewater <ruben@bridgewater.de> Reviewed-By: Joyee Cheung <joyeec9h3@gmail.com>
This is a first batch of updates that touches non-underscored modules in
/lib/.Checklist
make -j4 test(UNIX), orvcbuild test(Windows) passesAffected core subsystem(s)
all subsystems :)
/cc @joyeecheung @BridgeAR @jasnell