XML Binary Infoset Encoding (XBIS) is an encoding designed to eliminate most of the padding of XML text documents being passed between programs, while being faster to generate and interpret. The focus is more on speed than on size, so if document size is the major concern standard compression algorithms can offer superior results. The current Java implementation shows 4-8X performance benefits over standard XML parsers over a range of document types and sizes and across JVMs tested.
|Tags||Communications Internet Text Processing Markup XML|
|Licenses||MIT/X BSD Revised|
Release Notes: This release changed the main focus from transporting the full XML Infoset to preserving the canonical form of documents converted to and from XBIS representations. It still supports transferring more information than is needed to preserve the canonical form, though, in order to be usable as a general serialization mechanism for document models. This release also included tests to compare XBIS processing time and document size against both text and zip/gzip representations, along with updated test results.
Release Notes: This release provides Java adapters for converting to and from SAX2 events. The input adapter takes an XBIS document representation and emulates a SAX2 parser processing the original document. The output adapter takes input from a SAX2 parser and generates XBIS output. The conversions handled by the earlier XMLS code (to and from JDOM and dom4j document models) have not yet been ported to the newer XBIS core code, but are planned for the next release. The actual encoded formats are essentially the same as for XMLS.