pdf_graphics

The rendering core of the dart-pdf suite: a content-stream interpreter, a device interface, a font engine, and text extraction.

Pure Dart with no dart:ui or Flutter dependency, so it runs on the VM, in CLIs and servers, and on the web. The Flutter canvas device lives in pdf_editor; implement PdfDevice to render anywhere else (SVG, raster, print pipelines).

Features

• Interpreter covering the full content operator set: paths, clipping, text, images and inline images, XObjects, transparency groups, soft masks, blend modes, optional content, and type 0–4 functions.
• Fonts: Type 1, TrueType, CFF, Type 0/CID, and Type 3, embedded or with standard-14 metrics; encodings, CMaps, and glyph outlines.
• Color: ICC profiles (validated against littleCMS), Separation/DeviceN with tint transforms, Indexed, Lab, and calibrated spaces.
• Shadings and patterns: types 1–7 including the mesh families, with /Extend semantics; tiling and shading patterns.
• Text extraction: run geometry for selection and search, reading in document order, rotation-aware page geometry.
• Annotation appearance-stream rendering for view-accurate annotation and form-field display.

Conformance is pinned by two checked-in corpora: the Ghent PDF Output Suite V5.0 and 171 edge-case files from the PDF.js test corpus.

Usage

import 'package:pdf_document/pdf_document.dart';
import 'package:pdf_graphics/pdf_graphics.dart';

final doc = PdfDocument.open(bytes);

// Extract text with geometry
final text = PdfPageText.extract(doc, 0);
print(text.text);

// Render by implementing PdfDevice and feeding it to PdfInterpreter

The suite