MIME type parsing, comments

[annevk]

Both http://searchfox.org/mozilla-central/source/netwerk/base/nsURLHelper.cpp#836 and Chrome source code indicate that the ";(" sequence means something special in MIME types and it looks like parsing is supposed to stop there, due to some legacy MIME type comment feature.

However, when I try something like

HTTP/1.1 200
Content-Length: 55
Content-Type: text/html;test=test;(;charset=gbk

<script>document.write(document.characterSet)</script>

browsers consistently end up not ignoring the charset parameter, so I'm not sure what is going on.

@mcmanus @sleevi help appreciated.

[annevk]

Ah, ";(" just means to look for either and it only works for charset parameter values. With text/html;charset=gbk(test) I can reproduce and it seems only Chrome and Firefox implement this kind of parsing. Edge and Safari will output windows-1252 rather than GBK. So maybe it's time to remove this oddity.

With text/html;x=(;charset=gbk all browsers detect GBK.

[annevk]

For text/html(;charset=gbk Chrome, Edge, and Firefox detect GBK; Safari renders it as text. For text/html(charset=gbk Chrome and Firefox will detect windows-1252, Edge will download, and Safari renders it as text.

I think that still means we can remove this quirk and will file bugs accordingly.

[annevk]

• https://bugzilla.mozilla.org/show_bug.cgi?id=1406337
• https://bugs.chromium.org/p/chromium/issues/detail?id=772343

[annevk]

Bugs filed and tested by #42, so closing.