Add quickfix for csso parse error #39
Conversation
|
The test doesn't fail without the fix... It doesn't seem to test the bug... Not sure... Maybe I do something wrong.. |
|
Strange... the tests are failing as expected if I just return the content without modifying it |
|
|
Only because the atrifical Construct a test case that demonstrate the bug... |
|
See my first picture. When the content is returned without adding the |
|
I have the same problem and the PR fixed it for me. Thanks 👍 |
|
👍 |
|
I would like to reprodude it before merging... So please give me a testcase. |
|
I've found the underlying problem in my input files. They were generated by the sass preprocessor, which adds a byte-order-mark to the beginning of the resulting CSS file. csso was unable to parse this file correctly so I opened a PR there: css/csso/pull/223 |
|
👍 |
Maybe we can add this to webpack... |
We already have this: https://github.com/webpack/core/blob/master/lib/NormalModuleMixin.js#L15-L22 |
|
Closing as we are no longer using csso... |
IE9 will actually drop certain things (like media queries) from the cssText property when it is read back out. To fix, store the raw css text in a separately managed array, and use that to form the entire string that should be set on the style element's cssText property. Fixes webpack-contrib#39.
…dia-queries [webpack-contrib#39] Don't use styleSheet to store state of css text
This pull request adds a quickfix for the csso parse error.
#36
css/csso#185