console: timeEnd() doesn't throw an error if no matching label #5901

ghaiklor · 2016-03-25T15:03:43Z

Pull Request check-list

Does make -j8 test (UNIX) or vcbuild test nosign (Windows) pass with
this change (including linting)?
Is the commit message formatted according to [CONTRIBUTING.md][0]?
If this change fixes a bug (or a performance problem), is a regression
test (or a benchmark) included?
Is a documentation update included (if this change modifies
existing APIs, or introduces new ones)?

Affected core subsystem(s)

console

Description of change

Based on #3514 discussion, I've removed throw an error in timeEnd() method, if there is no matching label.

cjihrig · 2016-03-25T15:06:45Z

lib/console.js

-  if (!stdout || typeof stdout.write !== 'function') {
-    throw new TypeError('Console expects a writable stream instance');
-  }
+  if (!(this instanceof Console)) return new Console(stdout, stderr);


Please revert these style changes.

@cjihrig What style guides do you using? Should I revert all the style changes or only this one?

We have an ESLint setup. But, in general, you should change the bare minimum in your PRs. Otherwise, it makes tools like git blame harder to use.

ghaiklor · 2016-03-25T15:25:34Z

@cjihrig done

cjihrig · 2016-03-25T15:41:22Z

test/parallel/test-console.js

@@ -8,7 +8,8 @@ assert.ok(process.stderr.writable);
 assert.equal('number', typeof process.stdout.fd);
 assert.equal('number', typeof process.stderr.fd);

-assert.throws(function() {
+assert.doesNotThrow(function() {
+  console.time('some label');


This is unnecessary.

ghaiklor · 2016-03-25T16:15:37Z

@cjihrig got it, so we can ignore non-existing label.

cjihrig · 2016-03-25T16:20:37Z

lib/console.js

-    throw new Error('No such label: ' + label);
-  }
+  const time = this._times.get(label);
+  if (!time) return;


Can you just replace the old throw with return;. It just needs to be a one line change.

cjihrig · 2016-03-25T16:21:43Z

I think so (see #3514 (comment)). What you had before was wrong. It would print something but it had the incorrect times.

cjihrig · 2016-03-25T16:22:11Z

test/parallel/test-console.js

@@ -8,10 +8,6 @@ assert.ok(process.stderr.writable);
 assert.equal('number', typeof process.stdout.fd);
 assert.equal('number', typeof process.stderr.fd);

-assert.throws(function() {


I didn't mean to delete the whole test. Just the one line that I commented on was unnecessary.

ghaiklor · 2016-03-25T16:38:14Z

@cjihrig here is my code:

var console = require('console');

console.time('test');
console.timeEnd('no such label');
console.timeEnd('test');

Prints only for test label:

./node test.js
test: 0.652ms

The following code:

var console = require('console');
console.timeEnd('no such label');

prints nothing.

cjihrig · 2016-03-25T17:29:12Z

lib/console.js

}
+

Please remove this newly added line.

cjihrig · 2016-03-25T17:29:20Z

LGTM with one comment.

ghaiklor · 2016-03-25T19:14:38Z

@cjihrig I see now. I must change only lines where it actually needs. Updated.

cjihrig · 2016-03-25T19:21:09Z

Thanks! LGTM. CI: https://ci.nodejs.org/job/node-test-pull-request/2067/

jasnell · 2016-03-26T04:28:16Z

LGTM

thefourtheye · 2016-03-26T05:26:01Z

We are doing this so that we would be on par with browsers? It feels like what we are doing is correct. If there is a typo in the name or if the usage is wrong we explicitly throw and let the user know what is wrong.

jasnell · 2016-03-26T05:29:15Z

How about a compromise? Don't throw but emit a warning using process.emitWarning.

ghaiklor · 2016-03-26T07:02:01Z

In case of wrong label you get nothing in the console. That way you will know that you forgot to close console.time(). Looks as expected behaviour for me.

jasnell · 2016-04-18T23:18:41Z

@thefourtheye ... do you still object to this or ...?

thefourtheye · 2016-04-19T02:47:04Z

Yes, I am still not convinced that this change is better.

To think of it, emitting a warning, as you suggested, would be the right fit here. This is not a serious error, I agree, but there should be some feedback to the user.

jasnell · 2016-04-19T03:42:34Z

lib/console.js

  if (!time) {
-    throw new Error('No such label: ' + label);
+    return;


@ghaiklor ... would it be possible for you to add a ...

process.emitWarning(`No such label: ${label}`);

...before returning here?

Should we include console.timeEnd somewhere in the message?

Wouldn't hurt... No such label '${label}' for console.timeEnd() perhaps? Keep in mind, also, that the warning includes a stack trace.

Ah, then the message is just fine.

ghaiklor · 2016-04-19T05:07:35Z

@jasnell @thefourtheye yeah, I will update branch in few hours.

ghaiklor · 2016-04-19T06:02:25Z

Warning doesn't include a stack trace. So I updated message with @jasnell approach.

> console.timeEnd('no label');
> (node:51305) Warning: No such label 'no label' for console.timeEnd()

jasnell · 2016-04-19T06:08:31Z

@ghaiklor. It does, it's just suppressed in the default output. If you use the --trace-warnings flag it will print the stack trace. And if you watch for warnings using process.on('warning', function(warn) { /** ... **/ });, the warn object will have a stack property that returns the stack trace.

jasnell · 2016-04-19T06:11:03Z

one additional suggestion: in the test case, you can add a check to make sure the warning is issued appropriately... something like...

process.on('warning', common.mustCall((warning) => {
  assert(/No such label 'no label' for console.timeEnd\(\)/.test(warning.message));
}));

jasnell · 2016-04-19T06:11:17Z

LGTM with a suggestion.

ghaiklor · 2016-04-19T07:15:32Z

I've added check for warning event in the test case.

thefourtheye · 2016-04-19T10:24:16Z

LGTM

jasnell · 2016-04-20T17:31:51Z

test/parallel/test-console.js

@@ -1,5 +1,5 @@
 'use strict';
-require('../common');
+var common = require('../common');
 var assert = require('assert');


nit: these requires should use const

jasnell · 2016-04-20T17:33:39Z

couple more nits after adding the warning check

cjihrig · 2016-04-20T17:42:24Z

test/parallel/test-console.js

-assert.throws(function() {
+assert.doesNotThrow(function() {
+  const message = /No such label 'no such label' for console\.timeEnd\(\)/;
+  const onWarning = common.mustCall(warning => assert(message.test(warning.message)));


Does this pass linting?

It shouldn't. We added a rule where parens are explicitly required I though

cjihrig · 2016-04-20T17:42:47Z

Yea, still LGTM once the comments are addressed.

ghaiklor · 2016-04-21T12:34:00Z

@cjihrig updated and fixed conflicts with master branch.

cjihrig · 2016-04-21T13:00:19Z

It would probably be simpler to use assert.strictEqual() instead of a regular expression in your test, but I don't think it needs to hold this us. Could you squash your changes down to a single commit?

ghaiklor · 2016-04-21T14:12:50Z

@cjihrig yes, I can, but I'm not sure if I can push them with force flag. As I understand, I need to do:

git rebase -i master

Squash all commits in one and then

git push --force origin fix/3514

cjihrig · 2016-04-21T14:29:41Z

It's fine to force push your own branch. No one else is using it.

ghaiklor · 2016-04-21T15:46:51Z

@cjihrig done

jasnell · 2016-04-21T15:53:13Z

test/parallel/test-console.js

+  const message = /No such label 'no such label' for console\.timeEnd\(\)/;
+  const onWarning = (warning) => assert(message.test(warning.message));
+  process.once('warning', common.mustCall(onWarning));
+


this could likely be rewritten and simplified to just:

process.once('warning', common.mustCall((warning) => { assert(/no such label/.test(warning.message); });

There's likely very little need to test the entire warning message. You just want to verify that the warning is being issued.

jasnell · 2016-04-21T15:54:14Z

Just one last nit while I was looking it back over :-) Looking good!

When timeEnd() provided with label that doesn't exists it emits warning in the console, so developer get know about it.

ghaiklor · 2016-04-21T16:05:39Z

@jasnell done.

BTW, why do I need squash commits by myself if it can be done when merging into master (after rebase)?

jasnell · 2016-04-21T16:10:19Z

@ghaiklor ... while it's possible for the person landing the PR to squash everything, it's often easier for those of us who do the landing to have the author of the PR squash them down (typically because it saves time). I personally don't mind squashing commits when I land things.

jasnell · 2016-04-21T16:11:16Z

New CI: https://ci.nodejs.org/job/node-test-pull-request/2359/
LGTM if CI is green

ghaiklor · 2016-04-21T17:13:43Z

@jasnell I see, thanks for explanation 😃
CI is green, my first contribution ^^

jasnell · 2016-04-21T17:41:14Z

Awesome! Thank you for the contribution and for the patience! @cjihrig ... does this LGTY?

cjihrig · 2016-04-21T17:47:10Z

(yes, LGTM)

jasnell · 2016-04-21T17:48:32Z

Awesome. Given that this is a semver-major, let's give time for one more round of last call comments from @nodejs/ctc. If there are no objections by Monday I'll get this landed.

When timeEnd() provided with label that doesn't exists it emits warning in the console, so developer get know about it. PR-URL: #5901 Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Sakthipriyan Vairamani <thechargingvolcano@gmail.com>

jasnell · 2016-04-25T15:30:01Z

Landed in 1c84579

When timeEnd() provided with label that doesn't exists it emits warning in the console, so developer get know about it. PR-URL: nodejs#5901 Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Sakthipriyan Vairamani <thechargingvolcano@gmail.com>

When timeEnd() provided with label that doesn't exists it emits warning in the console, so developer get know about it. PR-URL: #5901 Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Sakthipriyan Vairamani <thechargingvolcano@gmail.com>

cjihrig reviewed Mar 25, 2016
View reviewed changes

mscdex added the console Issues and PRs related to the console subsystem. label Mar 25, 2016

cjihrig reviewed Mar 25, 2016
View reviewed changes

lib/console.js

}

Copy link

Contributor

cjihrig Mar 25, 2016

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Please remove this newly added line.

cjihrig added the semver-major PRs that contain breaking changes and should be released in the next major version. label Mar 25, 2016

jasnell reviewed Apr 19, 2016
View reviewed changes

jasnell reviewed Apr 20, 2016
View reviewed changes

cjihrig reviewed Apr 20, 2016
View reviewed changes

jasnell reviewed Apr 21, 2016
View reviewed changes

console: timeEnd() with no label emits warning

6196d12

When timeEnd() provided with label that doesn't exists it emits warning in the console, so developer get know about it.

jasnell added this to the 6.0.0 milestone Apr 22, 2016

jasnell closed this Apr 25, 2016

ghaiklor deleted the fix/3514 branch April 25, 2016 17:37

jasnell mentioned this pull request Apr 27, 2016

console.timeEnd shouldn't throw if there is no matching label #3514

Closed

console: timeEnd() doesn't throw an error if no matching label #5901

console: timeEnd() doesn't throw an error if no matching label #5901

Conversation

ghaiklor commented Mar 25, 2016

Pull Request check-list

Affected core subsystem(s)

Description of change

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

ghaiklor commented Mar 25, 2016

Choose a reason for hiding this comment

ghaiklor commented Mar 25, 2016

Choose a reason for hiding this comment

cjihrig commented Mar 25, 2016

Choose a reason for hiding this comment

ghaiklor commented Mar 25, 2016

Choose a reason for hiding this comment

cjihrig commented Mar 25, 2016

ghaiklor commented Mar 25, 2016

cjihrig commented Mar 25, 2016

jasnell commented Mar 26, 2016

thefourtheye commented Mar 26, 2016

jasnell commented Mar 26, 2016

ghaiklor commented Mar 26, 2016

jasnell commented Apr 18, 2016

thefourtheye commented Apr 19, 2016

jasnell Apr 19, 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

ghaiklor commented Apr 19, 2016

ghaiklor commented Apr 19, 2016

jasnell commented Apr 19, 2016

jasnell commented Apr 19, 2016

jasnell commented Apr 19, 2016

ghaiklor commented Apr 19, 2016

thefourtheye commented Apr 19, 2016

Choose a reason for hiding this comment

jasnell commented Apr 20, 2016

Choose a reason for hiding this comment

Choose a reason for hiding this comment

cjihrig commented Apr 20, 2016

ghaiklor commented Apr 21, 2016

cjihrig commented Apr 21, 2016

ghaiklor commented Apr 21, 2016

cjihrig commented Apr 21, 2016

ghaiklor commented Apr 21, 2016

Choose a reason for hiding this comment

jasnell commented Apr 21, 2016

ghaiklor commented Apr 21, 2016

jasnell commented Apr 21, 2016

jasnell commented Apr 21, 2016

ghaiklor commented Apr 21, 2016

jasnell commented Apr 21, 2016

cjihrig commented Apr 21, 2016

jasnell commented Apr 21, 2016

jasnell commented Apr 25, 2016

jasnell Apr 19, 2016 •

edited

Loading