From f1c38e25600d4b61b84a5cdba1c7c9a28cc742a7 Mon Sep 17 00:00:00 2001 From: Profpatsch Date: Sat, 2 Jan 2021 14:59:23 +0100 Subject: feat(Profpatsch): dump netencode spec & parser MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The netencode standard, a no-nonsense extension of netstrings for structured data. Includes a nix generator module and a rust parsing library. Imported from https://github.com/openlab-aux/vuizvui/tree/e409df3861f48de44d0e37277ce007e348a7a0dc/pkgs/profpatsch/netencode Original license GPLv3, but I’m the sole author, so I transfer it to whatever license depot uses. Change-Id: I4f6fa97120a0fd861eeef35085a3dd642ab7c407 Reviewed-on: https://cl.tvl.fyi/c/depot/+/2319 Tested-by: BuildkiteCI Reviewed-by: Profpatsch --- users/Profpatsch/netencode/default.nix | 48 +++ users/Profpatsch/netencode/netencode.nix | 49 +++ users/Profpatsch/netencode/netencode.rs | 553 +++++++++++++++++++++++++++++++ users/Profpatsch/netencode/spec.md | 111 +++++++ 4 files changed, 761 insertions(+) create mode 100644 users/Profpatsch/netencode/default.nix create mode 100644 users/Profpatsch/netencode/netencode.nix create mode 100644 users/Profpatsch/netencode/netencode.rs create mode 100644 users/Profpatsch/netencode/spec.md (limited to 'users') diff --git a/users/Profpatsch/netencode/default.nix b/users/Profpatsch/netencode/default.nix new file mode 100644 index 000000000000..4a06cd700e94 --- /dev/null +++ b/users/Profpatsch/netencode/default.nix @@ -0,0 +1,48 @@ +{ pkgs, depot, ... }: + +let + imports = { + inherit (depot.users.Profpatsch) + writers; + }; + + version-check = pkgs.buildRustCrate { + pname = "version-check"; + version = "0.9.2"; + crateName = "version-check"; + sha256 = "1vwvc1mzwv8ana9jv8z933p2xzgj1533qwwl5zr8mi89azyhq21v"; + }; + memchr = pkgs.buildRustCrate { + pname = "memchr"; + version = "2.3.3"; + crateName = "memchr"; + sha256 = "1ivxvlswglk6wd46gadkbbsknr94gwryk6y21v64ja7x4icrpihw"; + }; + nom = pkgs.buildRustCrate { + pname = "nom"; + version = "5.1.1"; + crateName = "nom"; + sha256 = "1gb4r6mjwd645jqh02nhn60i7qkw8cgy3xq1r4clnmvz3cmkv1l0"; + dependencies = [ memchr ]; + buildDependencies = [ version-check ]; + features = [ "std" "alloc" ]; + }; + + netencode-rs-common = tests: imports.writers.rustSimpleLib { + name = "netencode"; + dependencies = [ nom ]; + buildTests = tests; + release = false; + verbose = true; + } (builtins.readFile ./netencode.rs); + + netencode-rs-tests = netencode-rs-common true; + + netencode-rs = netencode-rs-common false; + +in { + inherit + netencode-rs + netencode-rs-tests + ; +} diff --git a/users/Profpatsch/netencode/netencode.nix b/users/Profpatsch/netencode/netencode.nix new file mode 100644 index 000000000000..6bb2fa002796 --- /dev/null +++ b/users/Profpatsch/netencode/netencode.nix @@ -0,0 +1,49 @@ +let + + netstring = tag: suffix: s: + "${tag}${toString (builtins.stringLength s)}:${s}${suffix}"; + + unit = "u,"; + + n1 = b: if b then "n1:1," else "n1:0,"; + + n = i: n: netstring "n${toString i}" "," (toString n); + i = i: n: netstring "i${toString i}" "," (toString n); + + n3 = n 3; + n6 = n 6; + n7 = n 7; + + i3 = i 3; + i6 = i 6; + i7 = i 7; + + text = netstring "t" ","; + binary = netstring "b" ","; + + tag = key: val: netstring "<" "|" key + val; + + concatStrings = builtins.concatStringsSep ""; + + record = lokv: netstring "{" "}" + (concatStrings (map (kv: tag kv.key kv.val) lokv)); + + list = l: netstring "[" "]" (concatStrings l); + +in { + inherit + unit + n1 + n3 + n6 + n7 + i3 + i6 + i7 + text + binary + tag + record + list + ; +} diff --git a/users/Profpatsch/netencode/netencode.rs b/users/Profpatsch/netencode/netencode.rs new file mode 100644 index 000000000000..57b0a094931c --- /dev/null +++ b/users/Profpatsch/netencode/netencode.rs @@ -0,0 +1,553 @@ +extern crate nom; + +use std::collections::HashMap; +use std::io::Write; + +#[derive(Debug, PartialEq, Eq, Clone)] +pub enum T { + // Unit + Unit, + // Boolean + N1(bool), + // Naturals + N3(u8), + N6(u64), + N7(u128), + // Integers + I3(i8), + I6(i64), + I7(i128), + // Text + // TODO: make into &str + Text(String), + Binary(Vec), + // Tags + // TODO: make into &str + Sum(Tag>), + // TODO: make into &str + Record(HashMap>), + List(Box>), +} + +#[derive(Debug, PartialEq, Eq, Clone)] +pub enum U<'a> { + Unit, + // Boolean + N1(bool), + // Naturals + N3(u8), + N6(u64), + N7(u128), + // Integers + I3(i8), + I6(i64), + I7(i128), + // Text + Text(&'a [u8]), + Binary(&'a [u8]), + // Tags + Sum(Tag<&'a str, Box>>), + Record(Vec<(&'a str, Box>)>), + List(&'a [u8]), +} + +#[derive(Debug, PartialEq, Eq, Clone)] +pub struct Tag { + // TODO: make into &str + pub tag: S, + pub val: A +} + +impl Tag { + fn map(self, f: F) -> Tag + where F: Fn(A) -> B { + Tag { + tag: self.tag, + val: f(self.val) + } + } +} + +fn encode_tag(w: &mut W, tag: &str, val: U) -> std::io::Result<()> { + write!(w, "<{}:{}|", tag.len(), tag)?; + encode(w, val)?; + Ok(()) +} + +pub fn encode(w: &mut W, u: U) -> std::io::Result<()> { + match u { + U::Unit => write!(w, "u,"), + U::N1(b) => if b { write!(w, "n1:1,") } else { write!(w, "n1:0,") }, + U::N3(n) => write!(w, "n3:{},", n), + U::N6(n) => write!(w, "n6:{},", n), + U::N7(n) => write!(w, "n7:{},", n), + U::I3(i) => write!(w, "i3:{},", i), + U::I6(i) => write!(w, "i6:{},", i), + U::I7(i) => write!(w, "i7:{},", i), + U::Text(s) => { + write!(w, "t{}:", s.len()); + w.write(&s); + write!(w, ",") + } + U::Binary(s) => { + write!(w, "b{}:", s.len()); + w.write(&s); + write!(w, ",") + }, + U::Sum(Tag{tag, val}) => encode_tag(w, tag, *val), + U::Record(m) => { + let mut c = std::io::Cursor::new(vec![]); + for (k, v) in m { + encode_tag(&mut c, k, *v)?; + } + write!(w, "{{{}:", c.get_ref().len())?; + w.write(c.get_ref())?; + write!(w, "}}") + }, + U::List(l) => { + write!(w, "[{}:", l.len())?; + w.write(l)?; + write!(w, "]") + } + } +} + +pub fn text(s: String) -> T { + T::Text(s) +} + +pub mod parse { + use super::{T, Tag, U}; + + use std::str::FromStr; + use std::ops::Neg; + use std::collections::HashMap; + + use nom::{IResult}; + use nom::bytes::complete::{tag, take}; + use nom::branch::{alt}; + use nom::character::complete::{digit1, char}; + use nom::sequence::{tuple}; + use nom::combinator::{map, map_res, flat_map, map_parser, opt}; + use nom::error::{context, ErrorKind, ParseError}; + + fn unit_t(s: &[u8]) -> IResult<&[u8], ()> { + let (s, _) = context("unit", tag("u,"))(s)?; + Ok((s, ())) + } + + fn usize_t(s: &[u8]) -> IResult<&[u8], usize> { + context( + "usize", + map_res( + map_res(digit1, |n| std::str::from_utf8(n)), + |s| s.parse::()) + )(s) + } + + fn sized(begin: char, end: char) -> impl Fn(&[u8]) -> IResult<&[u8], &[u8]> { + move |s: &[u8]| { + let (s, (_, len, _)) = tuple(( + char(begin), + usize_t, + char(':') + ))(s)?; + let (s, (res, _)) = tuple(( + take(len), + char(end) + ))(s)?; + Ok((s, res)) + } + } + + + fn uint_t<'a, I: FromStr + 'a>(t: &'static str) -> impl Fn(&'a [u8]) -> IResult<&'a [u8], I> { + move |s: &'a [u8]| { + let (s, (_, _, int, _)) = tuple(( + tag(t.as_bytes()), + char(':'), + map_res( + map_res(digit1, |n: &[u8]| std::str::from_utf8(n)), + |s| s.parse::() + ), + char(',') + ))(s)?; + Ok((s, int)) + } + } + + fn bool_t<'a>() -> impl Fn(&'a [u8]) -> IResult<&'a [u8], bool> { + context("bool", alt(( + map(tag("n1:0,"), |_| false), + map(tag("n1:1,"), |_| true), + ))) + } + + fn int_t<'a, I: FromStr + Neg>(t: &'static str) -> impl Fn(&'a [u8]) -> IResult<&[u8], I> { + context( + t, + move |s: &'a [u8]| { + let (s, (_, _, neg, int, _)) = tuple(( + tag(t.as_bytes()), + char(':'), + opt(char('-')), + map_res( + map_res(digit1, |n: &[u8]| std::str::from_utf8(n)), + |s| s.parse::() + ), + char(',') + ))(s)?; + let res = match neg { + Some(_) => -int, + None => int, + }; + Ok((s, res)) + } + ) + } + + fn tag_t(s: &[u8]) -> IResult<&[u8], Tag>> { + // recurses into the main parser + map(tag_g(t_t), + |Tag {tag, val}| + Tag { + tag: tag.to_string(), + val: Box::new(val) + })(s) + } + + fn tag_g<'a, P, O>(inner: P) -> impl Fn(&'a [u8]) -> IResult<&'a [u8], Tag<&'a str, O>> + where + P: Fn(&'a [u8]) -> IResult<&'a [u8], O> + { + move |s: &[u8]| { + let (s, tag) = sized('<', '|')(s)?; + let (s, val) = inner(s)?; + Ok((s, Tag { + tag: std::str::from_utf8(tag) + .map_err(|_| nom::Err::Failure((s, ErrorKind::Char)))?, + val + })) + + } + } + + /// parse text scalar (`t5:hello,`) + fn text(s: &[u8]) -> IResult<&[u8], T> { + let (s, res) = text_g()(s)?; + Ok((s, T::Text( + std::str::from_utf8(res) + .map_err(|_| nom::Err::Failure((s, ErrorKind::Char))) + .map(|s| s.to_string())?, + ))) + } + + fn text_g() -> impl Fn(&[u8]) -> IResult<&[u8], &[u8]> { + sized('t', ',') + } + + fn binary<'a>() -> impl Fn(&'a [u8]) -> IResult<&'a [u8], T> { + map(binary_g(), |b| T::Binary(b.to_owned())) + } + + fn binary_g() -> impl Fn(&[u8]) -> IResult<&[u8], &[u8]> { + sized('b', ',') + } + + fn list_t(s: &[u8]) -> IResult<&[u8], Vec> { + map_parser(list_g(), nom::multi::many0(t_t))(s) + } + + fn list_g() -> impl Fn(&[u8]) -> IResult<&[u8], &[u8]> { + sized('[', ']') + } + + fn skip() -> impl Fn(&[u8]) -> IResult<&[u8], ()> { + move |s: &[u8]| { + let (s, ()) = alt(( + // TODO: only use the sized parsers here + map(text, |_| ()), + map(unit_t, |_| ()), + map(list_g(), |_| ()), + map(t_t, |_| ()), + // TODO: add rest of parsers + ))(s)?; + Ok((s, ())) + } + } + + fn list_take<'a>(n: usize) -> impl Fn(&'a [u8]) -> IResult<&'a [u8], Vec>> { + map_parser(list_g(), nom::multi::many_m_n(n, n, u_u)) + } + + fn record_t<'a>(s: &'a [u8]) -> IResult<&'a [u8], HashMap>> { + let (s, r) = record_g(t_t)(s)?; + Ok((s, + r.into_iter() + // ignore duplicated tag names that appear later + // by reverting the vector now + .rev() + .map(|(k, v)| (k.to_string(), v)) + .collect::>())) + } + + fn record_g<'a, P, O>(inner: P) -> impl Fn(&'a [u8]) -> IResult<&'a [u8], Vec<(&'a str, Box)>> + where + O: Clone, + P: Fn(&'a [u8]) -> IResult<&'a [u8], O> + { + map_parser( + sized('{', '}'), + nom::multi::fold_many1( + tag_g(inner), + Vec::new(), + |mut acc: Vec<_>, Tag { tag, mut val }| { + acc.push((tag, Box::new(val))); + acc + } + ) + ) + } + + pub fn u_u(s: &[u8]) -> IResult<&[u8], U> { + alt(( + map(text_g(), U::Text), + map(binary_g(), U::Binary), + map(unit_t, |()| U::Unit), + map(tag_g(u_u), |t| U::Sum(t.map(Box::new))), + map(list_g(), U::List), + map(record_g(u_u), U::Record), + + map(bool_t(), |u| U::N1(u)), + map(uint_t("n3"), |u| U::N3(u)), + map(uint_t("n6"), |u| U::N6(u)), + map(uint_t("n7"), |u| U::N7(u)), + map(int_t("i3"), |u| U::I3(u)), + map(int_t("i6"), |u| U::I6(u)), + map(int_t("i7"), |u| U::I7(u)), + + // less common + map(uint_t("n2"), |u| U::N3(u)), + map(uint_t("n4"), |u| U::N6(u)), + map(uint_t("n5"), |u| U::N6(u)), + map(int_t("i1"), |u| U::I3(u)), + map(int_t("i2"), |u| U::I3(u)), + map(int_t("i4"), |u| U::I6(u)), + map(int_t("i5"), |u| U::I6(u)), + // TODO: 8, 9 not supported + ))(s) + } + + fn t_t(s: &[u8]) -> IResult<&[u8], T> { + alt(( + text, + binary(), + map(unit_t, |_| T::Unit), + map(tag_t, |t| T::Sum(t)), + map(list_t, |l| T::List(Box::new(l))), + map(record_t, |p| T::Record(p)), + + map(bool_t(), |u| T::N1(u)), + // 8, 64 and 128 bit + map(uint_t("n3"), |u| T::N3(u)), + map(uint_t("n6"), |u| T::N6(u)), + map(uint_t("n7"), |u| T::N7(u)), + map(int_t("i3"), |u| T::I3(u)), + map(int_t("i6"), |u| T::I6(u)), + map(int_t("i7"), |u| T::I7(u)), + + // less common + map(uint_t("n2"), |u| T::N3(u)), + map(uint_t("n4"), |u| T::N6(u)), + map(uint_t("n5"), |u| T::N6(u)), + map(int_t("i1"), |u| T::I3(u)), + map(int_t("i2"), |u| T::I3(u)), + map(int_t("i4"), |u| T::I6(u)), + map(int_t("i5"), |u| T::I6(u)), + // TODO: 8, 9 not supported + ))(s) + } + + #[cfg(test)] + mod tests { + use super::*; + + #[test] + fn test_parse_unit_t() { + assert_eq!( + unit_t("u,".as_bytes()), + Ok(("".as_bytes(), ())) + ); + } + + #[test] + fn test_parse_bool_t() { + assert_eq!( + bool_t()("n1:0,".as_bytes()), + Ok(("".as_bytes(), false)) + ); + assert_eq!( + bool_t()("n1:1,".as_bytes()), + Ok(("".as_bytes(), true)) + ); + } + + #[test] + fn test_parse_usize_t() { + assert_eq!( + usize_t("32foo".as_bytes()), + Ok(("foo".as_bytes(), 32)) + ); + } + + #[test] + fn test_parse_int_t() { + assert_eq!( + uint_t::("n3")("n3:42,abc".as_bytes()), + Ok(("abc".as_bytes(), 42)) + ); + assert_eq!( + uint_t::("n3")("n3:1024,abc".as_bytes()), + Err(nom::Err::Error(("1024,abc".as_bytes(), nom::error::ErrorKind::MapRes))) + ); + assert_eq!( + int_t::("i6")("i6:-23,abc".as_bytes()), + Ok(("abc".as_bytes(), -23)) + ); + assert_eq!( + int_t::("i3")("i3:0,:abc".as_bytes()), + Ok((":abc".as_bytes(), 0)) + ); + assert_eq!( + uint_t::("n7")("n7:09,".as_bytes()), + Ok(("".as_bytes(), 9)) + ); + // assert_eq!( + // length("c"), + // Err(nom::Err::Error(("c", nom::error::ErrorKind::Digit))) + // ); + // assert_eq!( + // length(":"), + // Err(nom::Err::Error((":", nom::error::ErrorKind::Digit))) + // ); + } + + #[test] + fn test_parse_text() { + assert_eq!( + text("t5:hello,".as_bytes()), + Ok(("".as_bytes(), T::Text("hello".to_owned()))) + ); + assert_eq!( + text("t4:fo,".as_bytes()), + // TODO: way better parse error messages + Err(nom::Err::Error(("fo,".as_bytes(), nom::error::ErrorKind::Eof))) + ); + assert_eq!( + text("t9:今日は,".as_bytes()), + Ok(("".as_bytes(), T::Text("今日は".to_owned()))) + ); + } + + #[test] + fn test_parse_binary() { + assert_eq!( + binary()("b5:hello,".as_bytes()), + Ok(("".as_bytes(), T::Binary(Vec::from("hello".to_owned())))) + ); + assert_eq!( + binary()("b4:fo,".as_bytes()), + // TODO: way better parse error messages + Err(nom::Err::Error(("fo,".as_bytes(), nom::error::ErrorKind::Eof))) + ); + assert_eq!( + binary()("b9:今日は,".as_bytes()), + Ok(("".as_bytes(), T::Binary(Vec::from("今日は".as_bytes())))) + ); + } + + #[test] + fn test_list() { + assert_eq!( + list_t("[0:]".as_bytes()), + Ok(("".as_bytes(), vec![])) + ); + assert_eq!( + list_t("[6:u,u,u,]".as_bytes()), + Ok(("".as_bytes(), vec![ + T::Unit, + T::Unit, + T::Unit, + ])) + ); + assert_eq!( + list_take(2)("[6:u,u,u,]".as_bytes()), + Ok(("".as_bytes(), vec![ + U::Unit, + U::Unit, + ])) + ); + assert_eq!( + list_t("[15:u,[7:t3:foo,]u,]".as_bytes()), + Ok(("".as_bytes(), vec![ + T::Unit, + T::List(Box::new(vec![T::Text("foo".to_owned())])), + T::Unit, + ])) + ); + } + + #[test] + fn test_record() { + assert_eq!( + record_t("{21:<1:a|u,<1:b|u,<1:c|u,}".as_bytes()), + Ok(("".as_bytes(), vec![ + ("a".to_owned(), Box::new(T::Unit)), + ("b".to_owned(), Box::new(T::Unit)), + ("c".to_owned(), Box::new(T::Unit)), + ].into_iter().collect::>>())) + ); + // duplicated keys are ignored (first is taken) + assert_eq!( + record_t("{25:<1:a|u,<1:b|u,<1:a|i1:-1,}".as_bytes()), + Ok(("".as_bytes(), vec![ + ("a".to_owned(), Box::new(T::Unit)), + ("b".to_owned(), Box::new(T::Unit)), + ].into_iter().collect::>())) + ); + } + + #[test] + fn test_parse() { + assert_eq!( + t_t("n3:255,".as_bytes()), + Ok(("".as_bytes(), T::N3(255))) + ); + assert_eq!( + t_t("t6:halloo,".as_bytes()), + Ok(("".as_bytes(), T::Text("halloo".to_owned()))) + ); + assert_eq!( + t_t("<3:foo|t6:halloo,".as_bytes()), + Ok(("".as_bytes(), T::Sum (Tag { + tag: "foo".to_owned(), + val: Box::new(T::Text("halloo".to_owned())) + }))) + ); + // { a: Unit + // , foo: List } + assert_eq!( + t_t("{52:<1:a|u,<3:foo|[33:<1:A|u,<1:A|n1:1,<1:B|[7:i3:127,]]}".as_bytes()), + Ok(("".as_bytes(), T::Record(vec![ + ("a".to_owned(), Box::new(T::Unit)), + ("foo".to_owned(), Box::new(T::List(Box::new(vec![ + T::Sum(Tag { tag: "A".to_owned(), val: Box::new(T::Unit) }), + T::Sum(Tag { tag: "A".to_owned(), val: Box::new(T::N1(true)) }), + T::Sum(Tag { tag: "B".to_owned(), val: Box::new(T::List(Box::new(vec![T::I3(127)]))) }), + ])))) + ].into_iter().collect::>>()))) + ); + } + + } +} diff --git a/users/Profpatsch/netencode/spec.md b/users/Profpatsch/netencode/spec.md new file mode 100644 index 000000000000..3058e36eaf5c --- /dev/null +++ b/users/Profpatsch/netencode/spec.md @@ -0,0 +1,111 @@ +# netencode 0.1-unreleased + +[bencode][] and [netstring][]-inspired pipe format that should be trivial go generate correctly in every context (only requires a `byte_length()` and a `printf()`), easy to parse (100 lines of code or less), mostly human-decipherable for easy debugging, and support nested record and sum types. + + +## scalars + +Scalars have the format `[type prefix][size]:[value],`. + +where size is a natural number without leading zeroes. + +### unit + +The unit (`u`) has only one value. + +* The unit is: `u,` + +### numbers + +Naturals (`n`) and Integers (`i`), with a maximum size in bits. + +Bit sizes are specified in 2^n increments, 1 to 9 (`n1`..`n9`, `i1`..`n9`). + +* Natural `1234` that fits in 32 bits (2^5): `n5:1234,` +* Integer `-42` that fits in 8 bits (2^3): `i3:-42,` +* Integer `23` that fits in 64 bits (2^6): `i6:23,` +* Integer `-1` that fits in 512 bits (2^9): `i9:-1,` +* Natural `0` that fits in 1 bit (2^1): `n1:0,` + +An implementation can define the biggest numbers it supports, and has to throw an error for anything bigger. It has to support everything smaller, so for example if you support up to i6/n6, you have to support 1–6 as well. An implementation could support up to the current architecture’s wordsize for example. + +Floats are not supported, you can implement fixed-size decimals or ratios using integers. + +### booleans + +A boolean is represented as `n1`. + +* `n1:0,`: false +* `n1:1,`: true + +TODO: should we add `f,` and `t,`? + +### text + +Text (`t`) that *must* be encoded as UTF-8, starting with its length in bytes: + +* The string `hello world` (11 bytes): `t11:hello world,` +* The string `今日は` (9 bytes): `t9:今日は,` +* The string `:,` (2 bytes): `t2::,,` +* The empty sting `` (0 bytes): `t0:,` + +### binary + +Arbitrary binary strings (`b`) that can contain any data, starting with its length in bytes. + +* The ASCII string `hello world` as binary data (11 bytes): `b11:hello world,` +* The empty binary string (0 bytes): `b0:,` +* The bytestring with `^D` (1 byte): `b1:,` + +Since the binary strings are length-prefixd, they can contain `\0` and no escaping is required. Care has to be taken in languages with `\0`-terminated bytestrings. + +Use text (`t`) if you have utf-8 encoded data. + +## tagged values + +### tags + +A tag (`<`) gives a value a name. The tag is UTF-8 encoded, starting with its length in bytes and proceeding with the value. + +* The tag `foo` (3 bytes) tagging the text `hello` (5 bytes): `<3:foo|t5:hello,` +* The tag `` (0 bytes) tagging the 8-bit integer 0: `<0:|i3:0,` + +### records (products/records), also maps + +A record (`{`) is a concatenation of tags (`<`). It needs to be closed with `}`. +If tag names repeat the later ones should be ignored. Ordering does not matter. + +Similar to text, records start with the length of their *whole encoded content*, in bytes. This makes it possible to treat their contents as opaque bytestrings. + +* There is no empty record. (TODO: make the empty record the unit type, remove `u,`?) +* A record with one empty field, `foo`: `{9:<3:foo|u,}` +* A record with two fields, `foo` and `x`: `{21:<3:foo|u,<1:x|t3:baz,}` +* The same record: `{21:<1:x|t3:baz,<3:foo|u,}` +* The same record (later occurences of fields are ignored): `{28:<1:x|t3:baz,<3:foo|u,<1:x|u,}` + +### sums (tagged unions) + +Simply a tagged value. The tag marker `<` indicates it is a sum if it appears outside of a record. + +## lists + +A list (`[`) imposes an ordering on a sequence of values. It needs to be closed with `]`. Values in it are simply concatenated. + +Similar to records, lists start with the length of their whole encoded content. + +* The empty list: `[0:]` +* The list with one element, the string `foo`: `[7:t3:foo,]` +* The list with text `foo` followed by i3 `-42`: `[14:t3:foo,i3:-42,]` +* The list with `Some` and `None` tags: `[33:<4:Some|t3:foo,<4None|u,<4None|u,]` + +## motivation + +TODO + +## guarantees + +TODO: do I want unique representation (bijection like bencode?) This would put more restrictions on the generator, like sorting records in lexicographic order, but would make it possible to compare without decoding + + +[bencode]: https://en.wikipedia.org/wiki/Bencode +[netstring]: https://en.wikipedia.org/wiki/Netstring -- cgit 1.4.1