Struct CompressionConfig 

pub struct CompressionConfig {
    pub min_array_length: usize,
    pub min_string_length: usize,
    pub min_frequency_count: u32,
    pub uuid_compression_potential: f32,
    pub min_net_savings: usize,
    pub delta_threshold: f32,
    pub min_delta_potential: f32,
    pub run_length_threshold: f32,
    pub min_compression_potential: f32,
    pub min_numeric_sequence_size: usize,
}

Configuration constants for compression algorithms

Fields

min_array_length: usize

Minimum array length for pattern analysis

min_string_length: usize

Minimum string length for dictionary inclusion

min_frequency_count: u32

Minimum frequency for dictionary inclusion

uuid_compression_potential: f32

Minimum compression potential for UUID patterns

min_net_savings: usize

Minimum net wire-byte saving required to select CompressionStrategy::Dictionary (or the dictionary half of CompressionStrategy::Hybrid). The net saving is computed per candidate string as gain - cost, summed across all kept entries and reduced by a fixed metadata envelope, using the same size accounting the reported compressed_size uses. This makes dictionary selection a modelled net-positive decision, not a guarantee: see the known imprecisions below, which can make an accepted payload net-negative on adversarial input. The wire-byte report (compressed_size and everything derived from it) is unaffected and always measured, never modelled (see issue #333).

For a string of length L repeated c times with per-occurrence marker overhead m = 1 + decimal_digits(index) and per-entry dictionary-array cost L + 3 (the string, its quotes, one separator), an entry only pays off once c*(L-m) > L+3, i.e. L > (c*m + 3) / (c - 1). The smallest achievable m is 2 (index 0 is one decimal digit), so for c = 2 that means L > 7. "active" (L = 6, c = 2) fails this per-entry gate (gain 8 < cost 9) and is pruned before min_net_savings is consulted; a payload whose only repeated string is "active" yields an empty dictionary and CompressionStrategy::None. Even one kept entry rarely clears the floor alone: at c = 2 it needs L >= 27 to reach the default min_net_savings of 10 after the envelope.

Known imprecisions, both of which shift the modelled net toward being optimistic (a real payload can save fewer bytes than modelled, never more): this accounting does not model JSON string escaping inside dictionary strings (e.g. embedded quotes or backslashes), and it does not model the 1-byte-per-instance cost of escaping a payload string that legitimately starts with the sentinel byte (see substitute_dictionary_strings). Both are symmetric across ordinary payloads and shift the byte count only slightly, but a payload engineered to maximize sentinel-led strings can make a selected Dictionary strategy net-negative in the real, measured report.

delta_threshold: f32

Threshold score for delta compression

min_delta_potential: f32

Minimum delta potential for numeric compression

run_length_threshold: f32

Threshold for run-length compression

min_compression_potential: f32

Minimum compression potential for pattern selection

min_numeric_sequence_size: usize

Minimum array size for numeric sequence analysis

Implementations

impl CompressionConfig

pub fn validate(&self) -> Result<(), ConfigError>

Validate compression configuration invariants.

Errors

Returns ConfigError::InconsistentBounds when a potential/ratio field (uuid_compression_potential, min_delta_potential, min_compression_potential) is outside 0.0..=1.0, or when a threshold field (delta_threshold, run_length_threshold) is negative or non-finite.

Examples

use pjson_rs::compression::CompressionConfig;

CompressionConfig::default().validate().expect("defaults are valid");

Trait Implementations

impl Clone for CompressionConfig

fn clone(&self) -> CompressionConfig

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for CompressionConfig

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for CompressionConfig

fn default() -> Self

Returns the “default value” for a type.