content_type.nx

content_type.nx -- parse a Content-Type / Accept header value. RFC 7231 §3.1.1.1. A media-type value is: type "/" subtype *( ";" param "=" value ) Examples: text/html text/html; charset=utf-8 multipart/form-data; boundary=----WebKitFormBoundary application/vnd.nishi.page+json; profile=blog Parses this into (type_off, type_len, subtype_off, subtype_len, params_off, params_len) triplets. Parameter list is left as a raw byte range -- caller can feed it into cookie.nx or query.nx style splitters if it needs per-param access. For most routing, knowing the type + subtype is enough. The Accept header uses the same grammar with an additional "q=" parameter for quality (0..1). We surface the parsed param range so caller can extract "q" on demand. Invariants: CT1 Type + subtype trimmed of surrounding whitespace. CT2 Type + subtype lowercased-in-place is CALLER's job (we're zero-alloc; won't mutate the buffer). CT3 Missing '/' returns CT_ERR_FORMAT; media types always have a slash. CT4 "*/*", "text/*", "application/*": accepted as-is; caller inspects for '*' to know whether it's a wildcard.