- From: Poul-Henning Kamp <phk@phk.freebsd.dk>
- Date: Fri, 26 May 2023 09:48:15 +0000
- To: ietf-http-wg@w3.org
IFF we are going to make UTF-8 a distinct data type in SF, which I still think we should not do, here is a way to do it I can live with: sf-utf8 A data type to transfer UTF8 encoding of UniCode sequences. Parsing If the first cahracter is not "!" fail Remove the "!" Call the Byte Sequence parser Serializing Emit "!" Call the Byte Sequence serializer. Security Considerations UTF8 and UniCode sequences are vectors for a large and diverse class of smuggling and confusion attacks. Implementations should document clearly where responsibility for validation lies: Caller or callee. The name clearly says both what this is and isn't. The serialization does not pose a risk if rendered unscrutinized. We reuse parser and serializer code to the max. Poul-Henning -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 phk@FreeBSD.ORG | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence.
Received on Friday, 26 May 2023 09:48:26 UTC