Request charset defaults to UTF-8, Response charset defaults to ISO_8859_1

[edatkinvey]

Both the request and response objects should default to the same charsets. This causes very hard to troubleshoot bugs with repeated requests.

from com.google.api.client.http.HttpResponse

public Charset getContentCharset() {
    return mediaType == null || mediaType.getCharsetParameter() == null
        ? Charsets.ISO_8859_1 : mediaType.getCharsetParameter();
}

from com.google.api.client.http.AbstractHttpContent

protected final Charset getCharset() {
    return mediaType == null || mediaType.getCharsetParameter() == null
        ? Charsets.UTF_8 : mediaType.getCharsetParameter();
}

Ideally both will use Charsets.UTF_8, as Charsets.ISO_8859_1 only supports a subset.

[elharo]

Per the HTTP 1.1 spec, if the server does not supply a charset in the HTTP headers, then ISO-8859-1 is assumed. Thus on the request, I think we should return Charsets.8859_1 instead of Charsets.UTF_8.

When sending, we can specify anything we like as long as we include the right header, and it's certainly better to include than not.