A level defines a finite alphabet and a bijection between valid words over that alphabet and symbols of the next level. SUL uses two distinct kinds of levels: literal and hash.
The first three levels use an exact integer bijection. Alphabet sizes grow via tetration starting from n = 2:
| Level | e | n = 2^e + 1 |
|---|---|---|
| 1 | 0 | 2 |
| 2 | 2 | 5 |
| 3 | 7 | 0x81 |
Level 4 would have 2^136 + 1 symbols — far too large for integer storage.
See literal/README.md for encoding details, tables, and streaming API.
Beyond level 3, each symbol is a 256-bit content-addressed Id. Any pair of identifiers merges into a new identifier via a SHA2-based compress, enabling unbounded nesting at a fixed size. The hash level array grows dynamically as encoding proceeds.
See hash/README.md for the encoding algorithm and streaming API.
All levels share the same word structure. A valid word has the form [s0, s1, ..., sk, t] where:
s0 > s1 > ... > sk— a strictly decreasing prefixsk <= t— a terminating symbol no less than the last prefix element
See the SUL overview for full details.