fix: add support for aliases starting with / (options.alias)

[atinux]

What kind of change does this PR introduce?

Improvement in alias handling

Did you add tests for your changes?

Of course

Summary

Hi :)

Thank you for this great webpack loader. Actually, we are using it a lot into nuxt.js.

We have an issue with the css-loader used into the vue files, we have a global webpack alias named ~ and it's used a lot to require file this way: import '~/assets/my-file.vue'

Actually, we are using css-loader with this configuration:

{
    loader: 'css-loader',
    options: {
      minimize: true,
      importLoaders: 1,
      sourceMap: true,
      alias: {
        '/static': join(this.options.srcDir, 'static'),
        '/assets': join(this.options.srcDir, 'assets')
      }
   }
}

So our users can do this into their CSS:

.logo {
  background: url('~/static/logo.png');
}
.logo-2 {
  background: url('/logo.png');
}

Before this PR, .logo background url is transformed to /static/logo.png and considered as an url not processed (even if it's starting by an alias), .logo-2 is not processed and that's perfect.

Aftet this PR, .logo is transformed as well and considered to be a relative path (since starting by a known alias) and .logo-2 is not changed 👍

Does this PR introduce a breaking change?

No breaking change

Related issue to nuxt.js repo: nuxt/nuxt#1435

[jsf-clabot]

All committers have signed the CLA.

[alexander-akait]

Good work! Thanks!

[atinux]

I believe it could be a minor change or patch since it's only a bug fix :)

[yuchonghua]

Very good support, probably when the merger can be released?
/ping @evilebottnawi

[meiyuanbo]

probably when the merger can be released?

[alexander-akait]

@michael-ciniawsky @d3viant0ne friendly ping, need review, thanks!

[michael-ciniawsky]

startsWithAlias => isAlias[ed]

[atinux]

I pushed a change to rename it to isAlias

[michael-ciniawsky]

Relabeled to 🏷 semver: Patch for now as I don't think this would be a breaking change, but anyone might chime in and correct me if I'm wrong 😛

[alexander-akait]

@michael-ciniawsky oh, yep, it is patch, just invalid choice semver

[appinteractive]

Can someone click on merge please? 👍

[alexander-akait]

@appinteractive need minimum 3 approved votes
@michael-ciniawsky @d3viant0ne friendly ping

[HadiChen]

Can someone click on merge please? 👍

[codecov]

Codecov Report

Merging #597 into master will increase coverage by <.01%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master     #597      +/-   ##
==========================================
+ Coverage   98.91%   98.91%   +<.01%     
==========================================
  Files          10       10              
  Lines         367      369       +2     
  Branches       87       87              
==========================================
+ Hits          363      365       +2     
  Misses          4        4

Impacted Files	Coverage Δ
lib/loader.js	100% <ø> (ø)	⬆️
lib/processCss.js	98.67% <100%> (+0.01%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update e16bdeb...4367442. Read the comment docs.

[michael-ciniawsky]

@atinux Thx

[liambuchanan]

Haven't done much digging yet as this is a sub-dependency for us, but it appears merging has introduced a bug as processCss is not being instantiated with a resolve method here https://github.com/webpack-contrib/css-loader/blob/master/lib/localsLoader.js#L18