All releases of cpdetector
Release Notes: This release fixes a crash in command line mode when an invalid declared charset (the "" charset) was found. The return code of the command line tool (CodepageProcessor) does not return 0 in case of an error anymore. A bug that broke the ability to reset input streams after detection was fixed.
Release Notes: This major bugfix release fixes two issues in commandline batch mode. The switch to skip moving undetected documents works now again. No attempt will be made to transcode undetected documents (the latter caused exceptional program flow).
Release Notes: This version is a stability release and fixes the byte order mark detection and incompatibility with OpenJDK. It also requires Java 1.5 now.
Release Notes: The release structure has been changed: cpdetetor.jar does not contain 3rd party library files anymore. Missing public functions are contained again. The proguard shrinker has been updated from version 3.8 to 4.2.
Release Notes: The proguard shrinker is now used, so the cpdetector jar is now more than ten times smaller. System.out is no longer used for logging in JChardetFacade. All packages were renamed with the prefix "info.monitorenter".
Release Notes: Severe errors like a potential infinite loop and incorrect file handling have been removed.
Release Notes: A bug in the Ant build of the source release has been fixed. Instructions for document tests with fit were added.
Release Notes: It is now possible let cpdetector guess the codepage out of the remaining possibilities when it is not possible to narrow down this set to one. This version marks the start of testing with FIT. A new best practice command line tool allows printing of the codepage name for file arguments.
Release Notes: This version is a result of beginning with quality assurance. Two severe bugs were fixed. Additional features include XML DTDs, test/build automation, and a new ASCII fallback detection implementation.
Release Notes: The framework now additionally detects XML encoding declarations in documents. This allows a speed up when handling mostly XML, as the fallback to the jchardet guessing and exclusion is more expensive.
