Next Generation erlguten -- a PDF generation application in Erlang
Erlang Other
Switch branches/tags
Nothing to show
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.


Using Erlang in typography applications


ErlGuten. is a system for hiqh quality typesetting, ErlGuten is free software. ErlGuten aims to produce typographic quality PDF directly from XML or from a program.

The aim of ErlGuten is to produce high quality PDF from a layout language or from a program. The ErlGuten distribution includes a programming API, so that Erlang programs can produce PDF - and a typesetting system for typesetting documents written in XML.

The name ErlGuten is chosen because the program is written in Erlang - the Guten part is a reference to Gutenberg the father of printing.

ErlGuten is a system for high quality typesetting, so we take a great deal of care when formatting text, a large number of optimizations are performed which improve the quality of the printed text. Many of these optimizations are usually only found in expensive professional type-setting programs. We believe that WYSIWYG programs have destroyed the fine art of typesetting - ErlGuten is a modest attempt to improve the situation.

We have chosen XML as the input language for it's wide appeal. XML provides only a thin abstraction layer over the typesetting system

  • so the adventurous can use the programming interface to ErlGuten - to directly produce typographic quality PDF in real-time. We expect this facility to be useful for the dynamic generation of documents from web-servers.

    In ErlGuten we take the view that the highest level of abstraction is the layout of a document - we are very concerned that the user can specify the exact position of text on the printed page. At the next level of abstraction we are concerned with the typefaces that are used to format different regions of the document.

    ErlGuten is designed for the production of large and complex documents with complex layout requirements, like newspapers or books. In ErlGuten layout, content, and document management are considered separate issues. Layout is template based - Content is assumed to be stored as a large number of documents in a file system or data base, document management is considered as a mapping operation which takes documents in the content data base and maps them onto templates to produce hight quality output.

    This is normal text, set 30 picas wide in 12/14 Times Roman. Many different typefaces can be used within a paragraph. Emphasized text is set in Times-Italic. Hyphenation uses the TeX hyphenation algorithm. Any of the 35 built-in PDF typefaces with the same point size can be mixed with a paragraph. The term {person,"Joe"} is an Erlang term which has been typeset in 12 point courier. The paragraph justification algorithm does proper kerning so, for example, the word AWAY is correctly kerned! - line breaks within a paragraph are selected by a sophisticated global optimization technique.

A Mininal PDF Construction

The following code is the minimum you need to create a one page PDF with calls to the pdf module.

PDF = eg_pdf:new(),

To create a pdf into which to put content


   To set the page size of the PDF. Other choices are letter, legal, lots of A formats and B formats.


   To set the page you are working on

eg_pdf:set_font(PDF, "Victorias-Secret", 40),

  To set the font to use until told different

pdf_lib:moveAndShow(PDF, 50, 700, "Hello Joe from Gutenburg"),

   To move to position (50,700) and place the "Hello Joe from Gutenberg" content

Serialised = eg_pdf:export(PDF),

   To create all the content of the pdf as a string in the term Serialised


   To output the pdf content into a finished PDF file


   To delete the PDF object and the PDF process.

The tests


When you run comcastBill:test(), you get a two page PDF which reproduces my Comcast bill for internet services.


When you run eg_test1:test(), it produces a four page PDF (eg_test1.pdf) using the pdf module function calls. It illustrates most features of the pdf module.


When you run eg_test2:test(), it produces a pretty ugly three page PDF (eg_test2.pdf) that also tests a number of features of the pdf module.


When you run eg_test3:test(), it produces one page of PDF. It includes several tests of justification of text. These tests include columns of text put into boxes located on the page. In the background is a grid showing the location of the items on the page. This uses the pdf module directly , but then also uses the xml parsing modules to processing xml strings into the PDF content.


When you run eg_test4:test(), it produces a two page PDF. The 1st page is an A4 layout grid with a few lines of text placed on it. The second page shows 14 examples of the same text in different fonts.


When you run eg_test5:test(), it produces a page of PDF. This page shows the planning grid and on it are 8 blocks of text. The content of each block is the same, but the blocks are different sizes, different justification schemes and arranged in angular rotation. If you run eg_test5:test(X) where is X is a number from 1 to 7 you get the tuple returned by break_richText when it processes its test paragraph with different kinds of justification and line lengths. If you run eg_test:test(8), you get the tuple returned by eg_line_break:make_partitions. This takes a sentence and shows you the places where you can hyphenate the last word in the sentence.


When you run eg_test6:test(), it produces a page of PDF. This is a very simple page produced by calls to the pdf module.


When you run eg_test8_test(), it produces the same output as eg_test1, but it assembles a pdf document directly by building it from pdf "objects" rather than going through the pdf.erl API (and its pdf process). This test mostly producess the same output as eg_test1.erl._


When you run eg_test8:test(), it produces a page of PDF. This tests the eg_table module by putting size tables on the page.


When you run eg_test12:test(), it produces a 4 page PDF that looks much the same as the results of eg_test1.


When you run eg:tmo_test:file(), it produces a 9 page PDF from the file process.xml. The contents are some Erlang software development standards. This code produces the PDF based on the content of the XML file. Within this test program are functions to produce tables in a PDF. It produces the revision history table on the second page and it is used to format the section headings between the lines. This is a sophisticated example of using the pdf modules to format a document.


When you run tmo_doc:file(), you get an output file that at first glance looks the same as the results from eg_tmo_test. The internals are different in many places.


When you run kd_test1:test(), it produces a 1 page PDF. It is a commercial bill in Swedish for a recording, I think. It does a good job of including a graphic to show the vendor's logo. It looks interesting.