fog-human-json

Crates.iofog-human-json
lib.rsfog-human-json
version0.2.1
sourcesrc
created_at2023-07-12 02:42:27.493736
updated_at2023-08-14 01:54:57.996509
descriptionHuman-readable fog-pack data, through JSON.
homepagehttps://github.com/Cognoscan/fog-human-json
repositoryhttps://github.com/Cognoscan/fog-human-json
max_upload_size
id914243
size58,425
Richard Scott Teal (Cognoscan)

documentation

https://docs.rs/fog-human-json

README

fog-human-json: human-readable fog-pack values

This crate provides functions to go back and forth between fog-pack and JSON, making it relatively easy for users to view pretty-printed fog-pack values and edit them with existing JSON tooling. A common complaint with binary data formats like fog-pack is that reading them is painful, and lowering that pain with JSON is exactly what this crate is for.

This is not a crate for turning regular JSON into fog-pack data. It uses a number of special string prefixes to encode fog-pack types in JSON, which can interfere with arbitrary JSON-to-fog conversions.

So, what does this actually do for conversion? Well, it takes each fog-pack type and either directly converts it to a corresponding JSON type, or it specially encodes it in a string that starts with $fog-. So a 32-bit floating point value could be specifically encoded as $fog-F32: 1.23. The full list of types is:

  • Str: A regular string. This is just prepended so fog-pack strings that start with $fog- won't get caught by the parser.
  • Bin: Encodes the binary data as Base64 using the "standard" encoding (bonus symbols of +/, no padding used, padding is accepted when parsing).
  • F32Hex / F64Hex: Encodes a binary32/64 IEEE floating-point value in big-endian hex. The fog-to-json process should only do this when writing out a NaN or Infinity.
  • F32 / F64 / Int: Prints a standard JSON Number, but includes the type information. This done by telling the converter to do it specifically, by a user adding type information, or by the converter for any F32 value (as serde_json will always use F64 for floating-point).
  • Time: Encodes the time as a RFC 3339 formatted string.
  • Hash / Identity / StreamId / LockId: Encodes the corresponding primitive as a base58 string (in the Bitcoin base58 style).
  • DataLockbox / IdentityLockbox / StreamLockbox / LockLockbox: Encodes the corresponding lockbox as Base64 data, just like with the "Bin" type.

That covers conversion between fog-pack Values and JSON values, but not Documents and Entries. Those are converted into JSON objects with the following key-value pairs:

  • Documents:
    • "schema": If present, a $fog-Hash:HASH with the schema.
    • "signer": If present, a $fog-Identity:IDENTITY with the signer's Identity.
    • "compression": If not present, uses default compression. If present and null, no compression is used. If set to a number between 0-255, uses that as the compression level.
    • "data": The document content. Must be present.
  • Entries:
    • "parent": Parent document's hash.
    • "key": Entry's string key.
    • "signer": If present, holds the signer's Identity.
    • "compression": If not present, uses default compression. If present and null, no compression is used. If set to a number between 0 & 255, uses that as the compression level.
    • "data": The entry content. Must be present.

When going from JSON to a Document or Entry, if there's a "signer" specified, it will attempt to pull a matching IdentityKey from a provided Vault and use that to reproduce the signature. If it can't, then the conversion will fail.

Commit count: 6

cargo fmt