dxpdf 0.1.0

A fast DOCX-to-PDF converter powered by Skia, written in Rust

dxpdf

A fast, lightweight DOCX-to-PDF converter written in Rust, powered by Skia.

Convert Microsoft Word .docx files to high-fidelity PDF documents — from the command line or as a Rust library. No Microsoft Office, LibreOffice, or cloud API required.

Built by nerdy.pro.

Why dxpdf?

• Fast — converts a 24-page document in ~115ms on Apple Silicon
• Accurate — Flutter-inspired measure→layout→paint pipeline with pixel-level fidelity
• Standalone — no external dependencies beyond Skia; no Office installation needed
• Cross-platform — runs on macOS, Linux, and Windows
• Dual-use — works as a CLI tool, Rust library (use dxpdf;), or Python package (import dxpdf)

Quick Start

Install

cargo install dxpdf

Convert a file

dxpdf input.docx                  # outputs input.pdf
dxpdf input.docx -o output.pdf    # specify output path

Use as a library

let docx_bytes = std::fs::read("document.docx")?;
let pdf_bytes = dxpdf::convert(&docx_bytes)?;
std::fs::write("output.pdf", &pdf_bytes)?;

You can also inspect or transform the parsed document model:

use dxpdf::{parse, model};

let document = parse::parse(&std::fs::read("document.docx")?)?;

for block in &document.blocks {
    match block {
        model::Block::Paragraph(p) => { /* ... */ }
        model::Block::Table(t) => { /* ... */ }
    }
}

let pdf_bytes = dxpdf::convert_document(&document)?;

Use from Python

pip install dxpdf

import dxpdf

# Bytes in, bytes out
pdf_bytes = dxpdf.convert(open("input.docx", "rb").read())

# File to file
dxpdf.convert_file("input.docx", "output.pdf")

What's Supported

dxpdf handles the most common DOCX features used in real-world business documents:

Category	Features
Text	Bold, italic, underline, font size/family/color, character spacing, superscript/subscript, run shading
Paragraphs	Alignment (left/center/right), spacing (before/after/line with auto/exact/atLeast), indentation, tab stops, paragraph borders, paragraph shading
Tables	Column widths, cell margins (3-level cascade), merged cells (gridSpan + vMerge with height distribution), row heights, borders, cell shading, nested tables
Images	Inline (PNG/JPEG/BMP/WebP), floating/anchored with alignment and percentage-based positioning
Styles	Paragraph styles, character styles (including built-in Hyperlink), basedOn inheritance, document defaults, theme fonts
Headers/Footers	Text, images, page numbers (PAGE/NUMPAGES field codes)
Lists	Bullets, decimal, lower/upper letter, lower/upper roman with counter tracking
Hyperlinks	Clickable PDF link annotations with URL resolution from relationships
Sections	Multiple page sizes/margins, section breaks, portrait/landscape
Layout	Automatic pagination, word wrapping at spaces and hyphens, line spacing modes, floating image text flow

Building from Source

Prerequisites

• Rust toolchain (1.70+)
• clang (required by skia-safe for building Skia bindings)
• Linux only: libfontconfig1-dev and libfreetype-dev (e.g., sudo apt-get install -y libfontconfig1-dev libfreetype-dev)

Build

cargo build --release

The release binary will be at target/release/dxpdf.

Architecture

dxpdf follows a measure→layout→paint pipeline inspired by Flutter's rendering model:

DOCX (ZIP) → Parse → Document Model (ADT) → Measure → Layout → Paint → Skia PDF

Each layout element (paragraphs, table cells, headers/footers) goes through three phases:

1. Measure — collect fragments, fit lines, produce draw commands with relative coordinates
2. Layout — assign absolute positions, handle page breaks, distribute heights (e.g., vMerge spans)
3. Paint — emit draw commands at final positions (shading → content → borders)

Modules

Module	Description
model	Algebraic data types representing the document tree (Document, Block, Inline, etc.)
parse	DOCX ZIP extraction, event-driven XML parser, style/numbering resolution
render/layout	Measure→layout→paint pipeline: fragment (shared line fitting), paragraph, table (three-pass), header_footer
render/painter	Skia canvas operations for PDF output
render/fonts	Font resolution: tries requested font first, falls back to metric-compatible substitutes
units	OOXML unit conversions (twips, EMUs, half-points) — spec-defined constants only

Running Tests

cargo test

The test suite includes 104 unit tests and 9 integration tests covering layout, tables, lists, floats, headers/footers, hyperlinks, superscript/subscript, field codes, and end-to-end conversion.

Visual regression tests compare rendered PDFs against Word-generated references using pixel matching (see VISUAL_COMPARISON.md).

OOXML Feature Coverage

Validated against ISO 29500 (Office Open XML). 34 features fully implemented, 6 partial, 15 not implemented.

Full feature matrix (click to expand)

Text Formatting (w:rPr)

Feature	Status
Bold, italic	✅ with toggle support
Underline	✅ font-proportional stroke width
Font size, family, color	✅
Superscript/subscript	✅
Character spacing	✅
Run shading	✅
Strikethrough	❌
Highlighting	❌
Caps, smallCaps	❌
Shadow, outline, emboss, imprint	❌
Hidden text	❌

Paragraph Properties (w:pPr)

Feature	Status
Alignment (left, center, right)	✅
Alignment (justify)	⚠️ parsed, renders left-aligned
Spacing before/after, line spacing	✅ auto/exact/atLeast
Indentation (left, right, first-line, hanging)	✅
Tab stops (left)	✅
Tab stops (center, right, decimal)	⚠️ parsed, render as left
Paragraph shading	✅
Paragraph borders	✅ with adjacent border merging
Keep with next, widow/orphan control	❌

Styles

Feature	Status
Paragraph styles, character styles	✅
basedOn inheritance	✅
Document defaults, theme fonts	✅

Tables

Feature	Status
Grid columns, cell widths (dxa)	✅
Cell widths (pct, auto)	⚠️ fall back to grid
Cell margins (3-level cascade)	✅
Merged cells (gridSpan, vMerge)	✅
Row heights	✅ min / ⚠️ exact treated as min
Table borders (per-cell, per-table)	✅
Border styles (single)	✅
Border styles (double, dashed, dotted)	⚠️ render as single
Cell shading (solid)	✅
Cell shading (patterns), vertical alignment	❌
Nested tables	✅

Images

Feature	Status
Inline images	✅ PNG, JPEG, BMP, WebP
Floating images	✅ offset, align, wp14:pctPos
Wrap modes	✅ none/square/tight/through
VML images	❌

Page Layout

Feature	Status
Page size and orientation	✅
Page margins (all 6)	✅
Section breaks (nextPage)	✅
Section breaks (continuous, even, odd)	⚠️ treated as nextPage
Multi-column, page borders, doc grid	❌

Headers/Footers

Feature	Status
Default header/footer	✅
First page, even/odd, per-section	❌

Lists

Feature	Status
Bullet, decimal, letter, roman	✅
Multi-level lists	⚠️ levels parsed, nesting limited

Fields

Feature	Status
PAGE, NUMPAGES	✅
Hyperlinks	✅ clickable PDF annotations
Unknown fields	✅ cached value fallback
TOC, MERGEFIELD, DATE	❌

Other

Feature	Status
Footnotes/endnotes	❌ warned
Comments, tracked changes	❌ / ⚠️
Text boxes, shapes, SmartArt, charts	❌
RTL text, automatic hyphenation	❌

Performance

Benchmarked on Apple M3 Max with hyperfine (20 runs, 3 warmup):

Metric	Value
Mean conversion time	113 ms
Peak memory (RSS)	19 MB
Test document	3-page form with 11 tables, 2 images, 2 sections

See BENCHMARKS.md for full history.

Dependencies

Crate	Purpose
quick-xml	Event-driven XML parsing
zip	DOCX ZIP archive reading
skia-safe	PDF rendering, text measurement, link annotations
clap	CLI argument parsing
thiserror	Error types
log + env_logger	Warnings for unsupported features (RUST_LOG=warn)
pyo3 (optional)	Python bindings via maturin

Contributing

Contributions are welcome. Please open an issue before submitting large PRs.

Built by nerdy.pro.

License

MIT