Struct AdaptiveThresholdConfig 

pub struct AdaptiveThresholdConfig {
    pub median_multiplier: f32,
    pub min_threshold_pt: f32,
    pub max_threshold_pt: f32,
    pub use_iqr: bool,
    pub min_samples: usize,
}

Configuration for adaptive threshold calculation.

Determines how the threshold is computed from gap statistics. All point values assume PDF points (1/72 inch).

Fields

median_multiplier: f32

Multiplier applied to median gap when computing threshold.

Default: 1.5 Range: 0.5 - 3.0 (values outside this may be unreasonable)

Higher values → more conservative (fewer gaps marked as word boundaries) Lower values → more aggressive (more gaps marked as word boundaries)

min_threshold_pt: f32

Minimum threshold in PDF points (floor).

Default: 0.05pt (very small, close to tracking/kerning) Range: 0.01 - 0.2pt

Prevents threshold from becoming too small even with small median gaps.

max_threshold_pt: f32

Maximum threshold in PDF points (ceiling).

Default: 1.0pt (about 1/72 inch, reasonable word spacing) Range: 0.5 - 2.0pt

Prevents threshold from becoming unreasonably large.

use_iqr: bool

Use IQR instead of median for robust threshold calculation.

Default: false

When true: threshold = (p75 - p25) * multiplier When false: threshold = median * multiplier

IQR-based approach is more robust to outliers but may require different multiplier values (typically 0.3 - 0.7).

min_samples: usize

Minimum number of samples required to compute meaningful statistics.

Default: 10

If fewer gaps exist, statistics cannot be reliably computed and the function returns a default threshold instead.

Implementations

impl AdaptiveThresholdConfig

pub fn new() -> Self

Create a new configuration with default values.

pub fn balanced() -> Self

Balanced configuration (default multiplier: 1.5).

Suitable for most PDF documents with standard spacing.

pub fn aggressive() -> Self

Aggressive configuration (lower multiplier: 1.2).

Marks more gaps as word boundaries. Useful when:

• Text has tight spacing
• You want to break up large blocks more aggressively
• False negatives (missed gaps) are worse than false positives

pub fn conservative() -> Self

Conservative configuration (higher multiplier: 2.0).

Marks fewer gaps as word boundaries. Useful when:

• Text has loose spacing
• You want to avoid breaking up tightly-kerned text
• False positives (extra gaps) are worse than false negatives

pub fn policy_documents() -> Self

Optimized for policy documents with tight spacing (multiplier: 1.3).

Policy documents often have:

• Narrow margins
• Tight justified alignment
• Minimal word spacing

This configuration requires larger gaps to be detected as boundaries, and sets higher minimum threshold to avoid false positives.

pub fn academic() -> Self

Optimized for academic papers with standard spacing (multiplier: 1.6).

Academic papers typically have:

• Standard margins
• Generous word spacing
• Single or double-column layouts

This configuration is slightly more conservative than balanced to handle the higher baseline spacing.

pub fn with_multiplier(multiplier: f32) -> Self

Create a configuration with custom multiplier.

Arguments

• multiplier - Multiplier for median or IQR (typically 0.5 - 3.0)

Examples

use pdf_oxide::extractors::gap_statistics::AdaptiveThresholdConfig;

let config = AdaptiveThresholdConfig::with_multiplier(1.4);

pub fn with_iqr(self, use_iqr: bool) -> Self

Enable or disable IQR-based calculation.

Arguments

• use_iqr - If true, use IQR instead of median

Examples

use pdf_oxide::extractors::gap_statistics::AdaptiveThresholdConfig;

let config = AdaptiveThresholdConfig::default()
    .with_iqr(true);

pub fn with_min_threshold(self, min_pt: f32) -> Self

Set minimum threshold floor.

pub fn with_max_threshold(self, max_pt: f32) -> Self

Set maximum threshold ceiling.

pub fn with_min_samples(self, count: usize) -> Self

Set minimum number of samples required.

Trait Implementations

impl Clone for AdaptiveThresholdConfig

fn clone(&self) -> AdaptiveThresholdConfig

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for AdaptiveThresholdConfig

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

Formats the value using the given formatter.

impl Default for AdaptiveThresholdConfig

fn default() -> Self

Returns the “default value” for a type.

impl PartialEq for AdaptiveThresholdConfig

fn eq(&self, other: &AdaptiveThresholdConfig) -> bool

Equality operator ==.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl StructuralPartialEq for AdaptiveThresholdConfig