Documentation
CompilePDF docs
Everything you need to run the four producers, submit a job, and query the lineage chain. Sourced live from the compile-pdf engine repo — README, docs/, and the design spec.
Getting started
Overview
CompilePDF — the only writer in the Print With Synergy stack.
Architecture
How CompilePDF splits four producers across one Python package, four containers, and one Codex read-side authority.
CLI
Command reference for compile-pdf — producer subcommands plus utility commands for version, contract, health, schema, lineage, and CJD pipelines.
Reference
Contract
Compile's published contract surface — endpoints, per-producer schema versions, and the Codex sections Compile pins against.
Deploy
How to ship CompilePDF — Dockerfile, Railway envelope, the four producer services, and the sibling sidecar pattern.
Rewrite producer
Object-tree mutations on a single PDF — OCG flips, metadata patches, color-space swaps, hygiene strips, page lifecycle ops. No content-stream surgery.
Marks producer
Register / crop / color-bar / fold / proofing marks plus external mark-template ingestion. Geometry comes from codex_pdf.geom.
Impose producer
Sheet-level step-and-repeat. Layout solved by codex_pdf.geom.tile_grid; Compile drops cells onto sheets via pikepdf. No Compile-side layout math.
Trap producer
Ink-pair spread / choke trap with three engine slots — pure_python, ghostscript, external. Color resolution comes from codex_pdf.color; geometry from codex_pdf.geom.polygon_offset.