feat: add support to analyze gzipped bundles #379
Conversation
xitter
changed the title
README.md
Outdated
| @@ -68,6 +68,7 @@ new BundleAnalyzerPlugin(options?: object) | |||
| |**`statsOptions`**|`null` or `{Object}`|Default: `null`. Options for `stats.toJson()` method. For example you can exclude sources of your modules from stats file with `source: false` option. [See more options here](https://webpack.js.org/configuration/stats/). | | |||
| |**`excludeAssets`**|`{null\|pattern\|pattern[]}` where `pattern` equals to `{String\|RegExp\|function}`|Default: `null`. Patterns that will be used to match against asset names to exclude them from the report. If pattern is a string it will be converted to RegExp via `new RegExp(str)`. If pattern is a function it should have the following signature `(assetName: string) => boolean` and should return `true` to *exclude* matching asset. If multiple patterns are provided asset should match at least one of them to be excluded. | | |||
| |**`logLevel`**|One of: `info`, `warn`, `error`, `silent`|Default: `info`. Used to control how much details the plugin outputs.| | |||
| |**`decompressExtenstion`**|`null` or `{Object}`|Default: `null`. Used to show stats for a compressed file, object has compressed file's extension as key and it's value is decompression algorithm of [zlib](https://nodejs.org/api/zlib.html) to use, like: `{ gz: { algorithm: 'unzipSync' } }`| | |||
There was a problem hiding this comment.
I don't think we need this option - decompression should be transparent to the end-user. It should just work without any configuration.
There was a problem hiding this comment.
I too agree on this, but then should it be basis extension itself, i.e. detect for gz/br and apply respective decompression?
There was a problem hiding this comment.
@th0r does the above comment look good?
There was a problem hiding this comment.
I think so. We can think about customizable extensions if someone will raise an issue about it.
test/parseUtils.js
Outdated
| @@ -16,7 +21,7 @@ describe('parseBundle', function () { | |||
| .forEach(bundleName => { | |||
| it(`should parse ${_.lowerCase(bundleName)}`, function () { | |||
| const bundleFile = `${BUNDLES_DIR}/${bundleName}.js`; | |||
| const bundle = parseBundle(bundleFile); | |||
| const bundle = parseBundle(bundleFile, {}); | |||
There was a problem hiding this comment.
Options parameter should be optional.
test/parseUtils.js
Outdated
| const bundleFile = `${BUNDLES_DIR}/validBundleWithArrowFunction.js`; | ||
| const compressedBundleFile = `${bundleFile}.${extension}`; | ||
| const expectedModules = JSON.parse(fs.readFileSync(`${BUNDLES_DIR}/validBundleWithArrowFunction.modules.json`)); | ||
| if (!zlib[algorithm]) { |
There was a problem hiding this comment.
We don't need this condition - it swallows valid error and we won't notice it.
There was a problem hiding this comment.
For brotli, node version needs to be >= v11.7.0. Thus I put it to pass ci-checks which run on node 8. How do you suggest handling it?
There was a problem hiding this comment.
For brotli, node version needs to be >= v11.7.0
We need to mention it in the docs and properly handle in the code.
There was a problem hiding this comment.
How do you suggest handling it?
Check in tests, whether current version of Node supports it and expect successful decoding in one case and throwing of expected error in the other.
There was a problem hiding this comment.
We need to mention it in the docs and properly handle in the code.
Brotli compression works for version >= v11.7.0, so it is kind of an already set expectation for brotli assets. Would you prefer it including in docs anyway?
There was a problem hiding this comment.
Yes. Maybe compression is done with some external tool instead of Node.js.
Feature request: #377
This would add functionality to generate stats for
.gzand.br, i.e. compressed assets.