Release iText 5.4.1

Almost two months have passed since 13th anniversary release of iText. We've spent these months on some projects for customers, on developing new business models (visiting different companies in the US) and on tying the loose ends that surfaced due to the major changes in 5.4.0. Let's take a look at the most important changes.

Release Notes

iText Core 5.4.1

The previous release was more or less the PDF/UA-release, introducing a better way to create and reuse structured PDFs aka Tagged PDF. In this release, we've added the support for extra attributes and we fixed some minor problems.

We've introduced plenty of new ways to create tabs. This functionality was inspired by a question of one of our customers who had a specific requirement that couldn't be met using ordinary "Chunk Tabs". We decided to make a project out of it, and we just went ahead rewriting all the tab functionality.

The same goes for the introduction of the concept of word spacing in the context of Chunk elements. A customer needed this feature, we provided it.

Another customer performed stress tests on the new IO package introduced in iText 5.4.0. He found several situations where iText 5.4.0 doesn't release file handles in a timely fashion. That's now fixed, but it remains important for you to close PdfReader instances once you no longer need them.

Yet another customer noticed that GIF files aren't always fully downloaded when creating an Image object using an internet URL. That problem has been solved too. You may have seen the tweets about the increasing volume of sales. More sales means more customers; more customers means more useful feedback; and more useful feedback means a better product!

Changelog

iText Core 5.4.1

• Changes made by Paulo Soares

  • DocumentFont: Undo of wrong simplification introduced in iText 5.4.0

• Changes made by Kevin Day
  • io package: Handling small and zero-sized files.
  • Reduced the memory use in the getContentBytesFromContentObject() method of ContentByteUtils.
  • Improved fix by Bruno that closes RandomAccessSources when the construction of a PdfReader object fails
• Changes made by Alexander Chingarev
  • Chunk: getContent() optimization (after the introduction of tabs)
  • Optimization assigning the UUIDs to iText elements
  • Improved fix by Kevin of fix by Bruno that closes RandomAccessSources when the construction of a PdfReader object fails
  • TrueTypeFont: Glyphs hashSet should be copied when creating SubSet
• Changes made by Michaël Demey
  • AcroFields: fixed the getAppearanceStates() for Comboboxes (with or without export values). It now returns an ordered array of the display values.
  • GifImage: fixed remote loading of images. GIF stored on a remote server weren't always loaded completely.
• Changes made by Pavel Alay
  • Tagged PDF: fix writing of attributes for list item in structure element.
  • Minor refactoring in PdfCopy, PdfStructureElement, and CompareTool
  • Support for extra attributes in PdfStructureElement
  • Performance improvements in PdfCopy and PdfSmartCopy
• Changes made by Eugene Markovskyi
  • Introduction of new Tab functionality.
  • Fixed issue with PDF/A-1B compliance, digital signature and keywords
  • Fixed incorrect sizes of PdfChunk based on scaled image.
  • Fix: The background-color position for image wasn't always correct
• Changes made by Raf Hens
  • Fixed a NullPointerException problem in PdfAWriter
  • Introduction of a hasUsageRights method that checks if a PDF has usage rights (Reader-enabling)
  • Fix: char spacing was taken into account twice in certain circumstances
  • Introduced the concept of word spacing for Chunk objects
  • Fix: the wrong bit-flag was used for Rich Text fields in PdfFormField
  • Fix: keeptogether for paragraphs was broken with the introduction of new table splitting logic
• Changes made by Bruno
  • Making sure RandomAccessSources are closed when the construction of a PdfReader object fails; improved by Kevin Day.
  • Quick and dirty fix in case the image content contains [blockcode]<whitespace>[/blockcode] EI [blockcode]<whitespace>[/blockcode] .

XML Worker 5.4.1

• Changes made by Eugene Markovskyi

  • Fix: The background-color position for image wasn't not correct

• Changes made by Pavel Alay
  • Fixed bug with separated selectors
  • Fixed height calculation for rows
  • Cleaned up default.css by removing tags, moving them to StyleAttrCSSResolver
  • Fix for XFA worker in ChunkCssApplier: add support of [blockcode]<font>[/blockcode],[blockcode]<b>[/blockcode] and [blockcode]<u>[/blockcode] elements.
  • Fix for XFA worker in ChunkCssApplier: possible null reference exception of dataPipeline in XFAFlattener.flatten()