Closed easel closed 8 years ago
Wow I must have been very tired when I wrote this charset-related part of the code.
You can test the bug with this code on the server-side:
app.get('/raw_greeting', function(req, res) {
res.write("Hello world").end(); // Does not send Content-Type header
});
and this test function:
"can omit content-type" - {
HttpRequest(s"$SERVER_URL/raw_greeting")
.send()
.map({
res => res.body ==> "Hello world"
})
}
However, you will also need to correct charsetFromContentType
method which uses a hardcoded "utf-8" as default instead of the oneByteCharset
constant. Currently, the oneByteCharset
is only used if we get a header like Content-Type: type/subtype
but if we do not get any Content-Type header at all, it will use the hardcoded "utf-8" string...
The test case actually causes a failure on the jvm as well which I will look into.
Ahh, cool. I'll see about the fixups on the JS side in a few hours after I sleep. Thanks for the tip.
All right, I cleaned up my mess and fixed the issue in #37 . Thanks for the bug report!
No test yet as I can't figure out how to get express to send an empty content-type header.