Struct RegionSignature 

pub struct RegionSignature {

    pub page_number: u32,
    pub region_label: String,
    pub x_left: f32,
    pub x_right: f32,
    pub y_top: f32,
    pub y_bottom: f32,
    pub area: f32,
    pub n_lines: u32,
    pub n_visual_lines: u32,
    pub median_line_height: f32,
    pub n_tokens: u32,
    pub italic_tokens: u32,
    pub bold_tokens: u32,
    pub normal_tokens: u32,
    pub glyph_area: f32,
    pub density_raw: f32,
    pub density: f32,
    pub n_peaks: u32,
    pub n_peaks_y: u32,
    pub y_peak_cv: f32,
}

Per-region signature. The unit downstream classifiers reason about. Computed for each leaf of the Region tree from RegionStats.per_page[p]. Interior nodes do not get signatures.

Fields

page_number: u32

region_label: String

Reading-order path from the Region tree leaf, e.g. “1”, “2-1”, “3-2-1”. Resolves to the matching leaf in RegionStats.per_page[…].root.

x_left: f32

x_right: f32

y_top: f32

y_bottom: f32

area: f32

(x_right - x_left) × (y_bottom - y_top). Floor-clamped to 1.0.

n_lines: u32

Distinct y-rows the leaf’s elements occupy (raw y_top values dedup’d by the same baseline).

n_visual_lines: u32

Lines after Y-baseline dedup (sub/superscripts collapsed).

median_line_height: f32

Median Y-extent (in pt) of input elements in the leaf. Used as the per-region reference for the visual-line tolerance.

n_tokens: u32

italic_tokens: u32

bold_tokens: u32

normal_tokens: u32

glyph_area: f32

Estimated ink-bearing area: Σ per-token chars × font_size² × glyph_width_ratio.

density_raw: f32

glyph_area / area. Body prose ≈ 0.4-0.6 typically.

density: f32

Per-document normalization: density_raw / max(density_raw across all regions in the document). Lands in [0, 1].

n_peaks: u32

X-projection sustained-peak count. ≥ 2 indicates multi-column content (table or list); 1 indicates flowing prose.

n_peaks_y: u32

Y-projection presence-based peak count. One peak per visual row in the leaf’s Y-extent.

y_peak_cv: f32

Coefficient of variation of inter-Y-peak gaps. 0.0 when < 3 peaks.

Trait Implementations

impl Clone for RegionSignature

fn clone(&self) -> RegionSignature

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for RegionSignature

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

Formats the value using the given formatter.

impl Default for RegionSignature

fn default() -> RegionSignature

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for RegionSignature

fn deserialize<__D>( __deserializer: __D, ) -> Result<RegionSignature, <__D as Deserializer<'de>>::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for RegionSignature

fn serialize<__S>( &self, __serializer: __S, ) -> Result<<__S as Serializer>::Ok, <__S as Serializer>::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.