Function decode

nodedb-codec/src/fsst.rs:196–280

Decode FSST-compressed data back to strings.

(data: &[u8])

/// Decode FSST-compressed data back to strings.
pub fn decode(data: &[u8]) -> Result<Vec<Vec<u8>>, CodecError> {
    if data.len() < 2 {
        return Err(CodecError::Truncated {
            expected: 2,
            actual: data.len(),
        });
    }

    // Read symbol table.
    let sym_count = u16::from_le_bytes([data[0], data[1]]) as usize;
    let mut pos = 2;
    let mut symbols: Vec<Vec<u8>> = Vec::with_capacity(sym_count);

    for _ in 0..sym_count {
        if pos >= data.len() {
            return Err(CodecError::Truncated {
                expected: pos + 1,
                actual: data.len(),
            });
        }
        let len = data[pos] as usize;
        pos += 1;
        if pos + len > data.len() {
            return Err(CodecError::Truncated {
                expected: pos + len,
                actual: data.len(),
            });
        }
        symbols.push(data[pos..pos + len].to_vec());
        pos += len;
    }

    // Read header.
    if pos + 8 > data.len() {
        return Err(CodecError::Truncated {
            expected: pos + 8,
            actual: data.len(),
        });
    }
    let _total_encoded =
        u32::from_le_bytes([data[pos], data[pos + 1], data[pos + 2], data[pos + 3]]) as usize;
    pos += 4;
    let string_count =
        u32::from_le_bytes([data[pos], data[pos + 1], data[pos + 2], data[pos + 3]]) as usize;
    pos += 4;

    // Read offsets.
    let offsets_size = string_count * 4;
    if pos + offsets_size > data.len() {
        return Err(CodecError::Truncated {
            expected: pos + offsets_size,
            actual: data.len(),
        });
    }
    let mut offsets = Vec::with_capacity(string_count);
    for i in 0..string_count {
        let off_pos = pos + i * 4;
        offsets.push(u32::from_le_bytes([