https://github.com/fogti/ffstru

weird file structure routines
https://github.com/fogti/ffstru

Last synced: 26 days ago
JSON representation

weird file structure routines

• Host: GitHub
• URL: https://github.com/fogti/ffstru
• Owner: fogti
• Created: 2022-02-01T22:24:56.000Z (almost 3 years ago)
• Default Branch: main
• Last Pushed: 2022-02-01T22:31:16.000Z (almost 3 years ago)
• Last Synced: 2023-03-21T21:53:12.321Z (almost 2 years ago)
• Language: Zig
• Size: 7.81 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# ffstru

## precedents

- GHR record format

- ZTX document format

- dbtin dump format

- tgs spec format

- packlist format

- Fefe tinyldap primary format

## primary data structure

- read-optimized

- partial column-orientied, store data according to access needs

- align data at 4 bytes

- all integers are stores as little endian

- indices can omit the last 2 bits, because they are always zero

- max size: a file/document can be at most 16GiB big, and thus contain

  2^32 32bit integers

(for usage examples, look at the tests in `src/main.zig`)

```zig

magic: u32 = 0x73644679 // "yFds" little-endian

record_cnt: u32, // = sizeof(records) / (2 * sizeof(u32))

// this is an index, so it measures length in 32bit units.

strtab_len: u32,

// the string table gets padded to 4 bytes,

// filled up with zeros at the end

strtab: [strtab_len * 4]u8,

records: []Record,

indices: [.implicit]Index,

pub const Record = struct {

  number_of_attributes: u32,

  rtype: u32,

  attrs: [number_of_attributes][2]u32,

};

pub const Index = struct {

  itype: u32,

  // offset of next index

  // next = (sizeof(self) / sizeof(u32)) - 3

  next: u32,

  // : if itype == 0 :

  indexed_attribute: u32,

  // records should be sorted by lexicographic order

  records_offsets: [records_cnt][2]u32,

};

```