Struct FontMetrics 

pub struct FontMetrics {
    pub size: f64,
    pub line_height: f64,
    pub avg_char_width: f64,
    pub text_align: TextAlign,
    pub typeface: FontFamily,
    pub resolved_widths: Option<Vec<u16>>,
    pub resolved_upem: Option<u16>,
    pub resolved_ascender: Option<i16>,
    pub resolved_descender: Option<i16>,
}

Font properties for text measurement.

Fields

size: f64

Font size in points.

line_height: f64

Line height as a multiplier of font size (typically 1.2).

avg_char_width: f64

Average character width as a fraction of font size (fallback).

text_align: TextAlign

Horizontal text alignment (from XFA <para hAlign>).

typeface: FontFamily

Font family for per-character width lookup.

resolved_widths: Option<Vec<u16>>

Resolved glyph widths from the actual PDF font (units per resolved_upem).

resolved_upem: Option<u16>

Units-per-em of the resolved font (typically 1000 or 2048).

resolved_ascender: Option<i16>

Typographic ascender of the resolved font (font units).

resolved_descender: Option<i16>

Typographic descender of the resolved font (font units, typically negative).

Implementations

impl FontMetrics

pub fn new(size: f64) -> Self

pub fn line_height_pt(&self) -> f64

Uses resolved ascender/descender when available, else size * 1.2.

XFA Spec 3.3 §28.1 (p1228) — Adobe Non-conformance: Font metrics. Adobe’s AXTE text engine ignores font-supplied line gap and uses 20% of font height. Our approach matches: (asc−desc)/upem gives the em-square height (no line gap added), and the fallback is size × 1.2 (20% extra).

pub fn measure_width(&self, text: &str) -> f64

Measure the width of text in points.

When resolved_widths is present, widths are interpreted as per-1000 text-space units indexed directly by character code in the 0..255 range. Upstream code must therefore apply any PDF /FirstChar offset as padding before storing widths here, so that code 65 ('A') is read from resolved_widths[65].

Characters outside the available resolved-width table fall back to the width of ASCII space. This keeps wrapping stable for partially populated tables and avoids treating unknown characters as zero-width.

When resolved_widths is absent, measurement falls back to AFM tables for the generic serif/sans/monospace families. Non-ASCII characters in that path use the width of 'n' as a conservative average for one visible glyph.

The function iterates over Unicode scalar values rather than UTF-8 bytes, so multibyte characters contribute a single glyph width.

Trait Implementations

impl Clone for FontMetrics

fn clone(&self) -> FontMetrics

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for FontMetrics

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

Formats the value using the given formatter.

impl Default for FontMetrics

fn default() -> Self

Returns the “default value” for a type.