Revision control
Copy as Markdown
Other Tools
use {QueryType, QueryClass, Name, Class, Header, RData};
use rdata::opt;
/// Parsed DNS packet
#[derive(Debug)]
#[allow(missing_docs)] // should be covered by spec
pub struct Packet<'a> {
pub header: Header,
pub questions: Vec<Question<'a>>,
pub answers: Vec<ResourceRecord<'a>>,
pub nameservers: Vec<ResourceRecord<'a>>,
pub additional: Vec<ResourceRecord<'a>>,
/// Optional Pseudo-RR
/// When present it is sent as an RR in the additional section. In this RR
/// the `class` and `ttl` fields store max udp packet size and flags
/// respectively. To keep `ResourceRecord` clean we store the OPT record
/// here.
pub opt: Option<opt::Record<'a>>,
}
/// A parsed chunk of data in the Query section of the packet
#[derive(Debug)]
#[allow(missing_docs)] // should be covered by spec
pub struct Question<'a> {
pub qname: Name<'a>,
/// Whether or not we prefer unicast responses.
/// This is used in multicast DNS.
pub prefer_unicast: bool,
pub qtype: QueryType,
pub qclass: QueryClass,
}
/// A single DNS record
///
/// We aim to provide whole range of DNS records available. But as time is
/// limited we have some types of packets which are parsed and other provided
/// as unparsed slice of bytes.
#[derive(Debug)]
#[allow(missing_docs)] // should be covered by spec
pub struct ResourceRecord<'a> {
pub name: Name<'a>,
/// Whether or not the set of resource records is fully contained in the
/// packet, or whether there will be more resource records in future
/// packets. Only used for multicast DNS.
pub multicast_unique: bool,
pub cls: Class,
pub ttl: u32,
pub data: RData<'a>,
}