ZExpr is composed of several formats:
- ZType: Self-describing codecs
- ZBase: Self-describing numerical print formats
- ZLink: Hash-linked content-identifiers
- ZAtom: Efficiently-packed typed bytes
- ZCons: Efficiently-packed canonical trees
ZType is a table of prefix codes which describe the type of the encoded bytes.
This table can be found in ztype.csv.
For example, the ZAtom:
xdead_beef_dead_beef:int
indicates that that xdead_beef_dead_beef should be interpreted as a signed
integer.
All ztypes can be optionally extended with a numeric decimal suffix which lifts the bit-length of the value into the type. Without any annotation, the type is assumed to be variable length (i.e. the length is not type-relevant).
For example
xdead_beef_dead_beef:int64
Indicates that xdead_beef_dead_beef is a 64-bit integer. This bit-number
indicated by this suffix must be byte-aligned, so that e.g. int63 is not a
valid ztype. The textual format of ZType uses bit-length rather than
byte-length to align better with the common convention of saying "64-bit" rather
than "8-byte" or "8-word." However, the binary encoding of ZType uses
byte-length, so the data_length of an int64 is encoded as the unsigned integer
8
ZBase defines a self-describing print format for bytes. This is not necessarily a numeric base, since bijective numerations (useful for canonically printing numbers as text) could be valid print formats too. (Consider the sequence: "A, B, C, ..., X, Y, Z, AA, AB, AC, ..., AX, AY, AZ, BA, BB, BC, ..." which Microsoft Excel uses to number its columns.)
| encoding | code | description |
|---|---|---|
| base2, | 'b', | binary (01010101), |
| base8, | 'o', | octal, |
| base10, | 'd', | decimal, |
| base16, | 'x', | hexadecimal, |
| base32z, | 'v', | z-base-32 |
| base58btc, | 'I', | base58 bitcoin, |
| base64url, | '~', | rfc4648 no padding, |
ZAtom:
<bytes>:<ztype>
wbafyasdfoiuwer:int
xdead_beef_dead_beef:bytes
ZCons:
(<zatom> <zatom> <zatom> ... <zatom>)
[TODO]
- Why use ZExpr instead of Multiformats?
- Why use ZExpr instead of JSON?
- Why use ZExpr instead of csexp?