Big Faceless PDF Library

Release Notes: "bfopdf.jar" jar is now compiled for Java 5.0 or later; customers running Java 1.4 now need to use the "bfopdf-14.jar" instead, which is a backported version of the main jar. This change allows benefiting from the improvements to Java, while still continuing to support customers using Java 1.4. Fixes to setOutputProfile, which was sometimes not failing when it should have. This could result in profiles being applied incorrectly, so documents that should have failed when being converted to PDF/A were succeeding.

Release Notes: This release verifies the API against Java 7 and fixes some Java 1.4 compatibility issues. JBIG2 classes from JPedal are now included and have been optimized. There's no longer any need to use "jbig2dec" or include the "jbig2.jar" in your classpath. There are speed improvements for text extraction and PDF creation. There are improvements to XFDF reading and writing. Rendering bitmap images with a Blend Mode is now working. PropertyChangeEvent support has been added to PDFPage and PDF classes.

Release Notes: When reading or writing a PDF using AES encryption, document restrictions (such as those on printing and text extraction) weren't maintained properly. This is now fixed. A Classloader issue when executing JavaScript in PDFs under Apache Tomcat was fixed. PDFPage.setThumbnail was added, although feature is unused in modern PDF viewers. Stretching of glyphs was improved when rendering substituted fonts with different metrics to the original. Basic hooks were added for use of elliptic curve cryptography in signatures. A reversed test on FormCheckbox.isSelected was fixed.

Release Notes: bfopdf-cmap.jar was updated. If you see ObjectInputStreamException when working with documents containing Chinese/Japanese/Korean text, then make sure you have installed the latest version of this Jar. A confusion was fixed regarding FNC1 characters in barcodes: previously, "\n" was allowed to mean FNC1 in Code128 barcodes, but "\n" is actually a valid character so this was broken some time ago. FNC1 is now specified with the U+EFC1 character (defined as BarCode.FNC1), and this may be used in DataMatrix codes as well.

Release Notes: An issue where certifying a PDF sometimes only applied a regular signature was fixed. Rules for CMYK PDF/A documents relating to link annotations were tweaked. Exhaustive barcode testing fixed bugs in PDF417, MaxiCode, Codabar, Aztec, and Datamatrix BarCode output. A regression bug where valid signatures were sometimes failing to verify was fixed.