Read AFP
Write AFP
AFP Viewer
185076 - streamline suitcase travel
AFPx Packaging
 Whitepaper: AFP Facelift


“Advanced” and “functional” — that captures the essence of what we have for Advanced Functional Printing (AFP).

IBM AFP is the most significant format in mass printing [1], with the advantage of well defined and powerful data format that enables high-speed printing at over 3000 pages per minute. First created in the 1980s, over the years AFP technology had long become the de facto printing solution for corporations requiring large number of batch mailings, such as telecoms, banks, insurance companies, and even governments [2]. That remains the case today and in the near future.

Information about AFP technology can be found on the websites of AFP Consortium [3]. Today the AFP market mainly involves AFP preprocessing, conversion to or from other formats, storage and distribution of the archive. Major players in this field includes HP, Ricoh, Compart, etc. Most popular processing of AFP data stream includes extraction and assembling for reprinting, conversion to PDF for online viewing, insertion of promotional materials, and more.

MODCA Mixed Object:Document Content Architecture – The structure of AFP data stream in general and the various types of page-level data and commands.
PTOCA Presentation Text Object Content Architecture – Command sets and syntax used within a text object to generate text on hosting media.
GOCA Graphics Object Content Architecture – Command sets and syntax used within a graphics object to generate vector graphics and in situ images.
IOCA Image Object Content Architecture – Command sets and syntax used within an imaging object to create an image, a tiled pattern, or a composite.
FOCA Font Object Content Architecture – Command sets and syntax used within an native fonts including code pages, raster or outline fonts, composite fonts.
BCOCA Bar Code Object Content Architecture – Commands used to specify a barcode on page, to be rendered by the hardware.

AFP technology is specified in a number of standards covering different parts: the general page description, font architecture, image commands, text drawing commands, barcode specifications, etc. [4].  Each is a large, specialized area requiring practical experiences. Not only each different data type has its own syntax, command sets, and data structures, but the handling of data often involves additional effort such as various compression schemes. The architecture framework is called MODCA (Mixed Object Document Content Architecture). AFP technology also wraps a number of related technologies without reinventing the wheels but with special treatments nonetheless – for example, Type 1 fonts and Jpeg images are embedded within certain parts of the AFP data flow, but the way they are represented and used are distinctive. Over the course of the evolution, AFP also incorporated multiple schemes for similar tasks, for example the handling of double-byte CJK fonts — there had been as many as seven vastly different ways of doing it.

As such the entry point for dealing with AFP data stream is fairly high. The AFP data often exceeds tens of thousands of pages and range from hundreds of megabytes to multiple gigabytes, thus performance is a critical issue — a solution must be able to process the highly variable data at the fastest speed using as least resources as possible.

At Cyphia Architect, our product suite achieved this goal with leading performance and reliability. Unlike most document conversion solutions, there is no separation between a parser and a generator; the object model of AFP supports reading, modifying, creating, and writing at the same time, on physical and logical levels. For example, to transform between AFP and PDF, common approaches would typically create one parser for AFP, another for PDF, and a generator for AFP, another for PDF. That is not the case here — our solution uses only one unified AFP class and one unified PDF class, each supporting reading or writing data from either source, and being intelligent enough to import data as-is when appropriate and create new data on-the-fly when necessary.

Our AFP to PDF conversion runs at a speed of up to 10,000 pages per minute. AFP is read in a sequential way, and PDF is written in a progressive way, both ensuring timely release of memory resource. As part of our DocEvents workflow framework, the AFP can also iterate itself and take over the initiative to focus on its own content manipulation.


This is the screenshot of a PDF generated from AFP source showing various types of contents on page.

[1] According to “The Best of Both Worlds: AFP or PDF?“, 2011.
[2] See the verticals of AFP industry.
[3] In addition to AFP Consortium.
[4] See the Wikipedia page about AFP.