Feature/toggle rules in src #402
Conversation
|
This is for #70 |
|
Please include the certificate of origin sign off and let us know when it's ready for comment and review.
|
|
DCO 1.1 Signed-off-by: Donald Abrams donabrams@gmail.com |
| return getToggledRules(ast); | ||
| }; | ||
|
|
||
| describe.only('rule toggling', function () { |
There was a problem hiding this comment.
Be sure to remove this .only so that Coveralls will report an accurate coverage.
There was a problem hiding this comment.
Good catch. I usually add an eslint warning for this.
|
Ready for comment and review. I'll need to make a separate issue for the doc-disabled-rules rule described in https://github.com/donabrams/sass-lint/blob/ab7e4d8ea2b8f65b09ac97a69e635f7802e758e5/docs/rules/doc-disabled-rules.md |
|
Coveralls is unhappy that you didn't test the other branches inside of Do you know anything about the performance impact of these extra traversals and checks? Visually there doesn't seem to be any significant impact. Just curious. |
|
I'll get this into our sub branch a little later on and then work on it a little too, it's looking good though. @benthemonkey don't worry about coveralls at the moment this isn't done and it's not going straight into develop. |
|
Performance is split into two independent parts per file, detection of disable blocks and result filtering. Detection of disable blocks is done in a single traversal. Time spent will grow linearly with # of comment blocks. No worries there. The result filtering is completed after results are calculated. Filtering time grows at NUM_RESULTS * NUM_GLOBAL_DISABLES + (per Rule: NUM_RULE_RESULTS * NUM_RULE_DISABLES). This is theoretically worrisome, but not in reality: 1) Global disable is usually only done at the top of the page, not many times. 2) If a rule is being ignored a ton of times (say > 5), it's likely it should be ignored once at the top of the page. |
|
Good call @benthemonkey |
|
Thanks for the thorough explanation, @donabrams. Glad my suggestion caught some typos. |
|
Ok, i'll have a look over this tomorrow, then we should get it into the feature branch we made and throw some more tests at it and maybe some more documentation. Is there anything you're still looking to add? Once we're good with that lets get a plan together to release this. Massive thankyou @donabrams it's looking great and is going to make sass-lint even better! Fingers crossed we may even get this out as part of 1.4.0 |
|
@donabrams This is being a bit picky, but do you mind going through and updating these commits to be authored by your GitHub account? I think it is helpful to link commits to the GitHub account of the person committing them. Plus, this way you can also show off your rank on the "Contributors" page 😄 There's a nice tool to help you do this: https://github.com/davidfokkema/git-rewrite-author |
|
@benthemonkey So git-rewrite-author definitely changed way more than I hoped for. As @DanPurdy pointed out, just adding a secondary email to my account was much easier. |
|
Sounds good. I apologize that it caused trouble. When I used that tool everything just worked. This PR looks good to me. |
| var first = tokens[0], | ||
| rules = tokens.slice(1); | ||
| switch (first) { | ||
| case 'sass-lint:disable': |
There was a problem hiding this comment.
This indentation isn't flagging up? Looks odd to me but if it's ok with eslint
|
I'll carry on with this tomorrow but so far it's looking excellent. For the NPE you're seeing when unable to reliably find a parent block I reckon we should just output an error on that line coming clean that we can't do what they want and suggest using line disables or something. Would that be a fair workaround for now? I'm assuming this is all going to change a little bit within the new gonzales. Also as i found with the property count rule the As i mentioned I'll look again tomorrow and when we're happy get this onto that feature branch and throw a lot more tests at it before we commit to putting it out. Absolutely awesome work though @donabrams this is great! |
|
Hi @donabrams sorry I never got round to this again, been a crazy couple of weeks! I'll get round to this soon and workto get the sass format tests in. sorry again. |
| } | ||
| var isRuleEnabled = toggledRules.ruleEnable[ruleId] | ||
| .reduce(function (acc, toggleRange) { | ||
| return isBeforeOrSame(line, column, toggleRange[1], toggleRange[2]) |
There was a problem hiding this comment.
Should do ref counting here (ie [disable] [disable] [enable] means it's still disabled.)
|
No worries, I have things I should update on this too... Been swamped past few weeks but I'll get on this soon. Need to:
|
|
Hey guys, any update on this? :) |
|
I think once we've got the gonzales update out of the way this will be the next major priority. |
|
Alright thanks, appreciate the update |
|
+1 for this feature... thank you! |
|
+1! Thanks! |
|
+1 |
|
Please don't +1 in pull requests it doesn't make anything happen quicker. |
|
And it annoys the hell out of everyone having subscribed to the issue. :) |
|
This should be visible in all PRs/issues: https://github.com/blog/2119-add-reactions-to-pull-requests-issues-and-comments |
|
@ACOSW please please please don't use +1's read the thread before commenting we'll just delete these comments now |
|
Are there any chance to merge this important issue? |
|
@danielbayerlein unfortunately this is very out of date right now, its one of our next big hurdles to get sorted but merging this right now would probably break a lot of things. We are looking to get it moving though as we know how necessary it is. 👍 |
|
I tried the merge and got just one merge conflict on my local on the index.js file. I ran the tests and i see the cli tests are failing which i believe is a setup issue on my end. |
|
We made massive changes in 1.6 and the AST we are using had massive changes too so we need to go through and review this and throw a lot more tests at it, it will be happening soon though don't worry! |
|
Merged this into our test branch for now. Intend to work on this soon but basically updated it will alllllll of the progress we've made since this began. 1 conflict which was minimal and not really a problem. Will open a new PR into develop to see the progress here. |
|
Ongoing work in #677 Any work should be a PR into feature/disable-linters branch NOT develop. Will look to update it all soon. |
So there are a ton more tests to add, and I am definitely having trouble finding the nearest block parent for a comment line. However, I wanted to give you an idea where I was headed.