[fitsbits] Java FITS library released with full support for tiled compressed images

[Tom McGlynn (NASA/GSFC Code 660.1)]

Richard van Nieuwenhoven has just released a significant upgrade, 
v1.14.0,  of the nom.tam Java FITS library that includes complete 
support for the tiled/compressed images.  The new library is available 
from GITHub (https://github.com/nom-tam-fits/nom-tam-fits) and I have 
added it to the standard location at  the HEASARC 
(http://heasarc.gsfc.nasa.gov/docs/heasarc/fits/java/v1.0). Javadocs for 
the library can be found at
http://heasarc.gsfc.nasa.gov/docs/heasarc/fits/java/v1.0/javadoc). A 
number of other smaller enhancements and bug fixes are included. The 
release note is included below.  Given the current state where the 
compression proposals are being reviewed for inclusion in the FITS 
standard, this release is particularly timely.

Since the Java implementation is directly based upon the CFITSIO code, 
the two should be fully compatible.  This release does not support 
general compressed binary tables but that is anticipated in a future 
release.  Many thanks to Richard for this work.

     Tom McGlynn

Release notes for V1.14.0

Image compression and tiling are now fully supported by nom-tam-fits. A 
100% Java implementation of the compression libraries available in 
cfitsio was implemented. An API for easy handling of compressed images 
is now provided.

Support for binary table compression and the NULL_PIXEL_MASK features is
anticipated in the next release.

When [de]compressing all available CPU's are automatically utilized.

Internal compression allows FITS files to be created where the data are
efficiently stored, but the metadata is still easily accessible. The 
tiling of images is particularly critical for supporting efficient 
access to subsets of very large images. A user can easily access only 
the tiles that overlap the region of interest and can skip data not of 
interest. While some skipping might be possible with uncompressed FITS 
files (i.e., read only the rows overlapping the desired subset), 
internal tiles can be much more efficient when the image is
substantially larger than the subset. Most compression algorithms 
interfere with the ability to skip uninteresting data, but tiles are 
compressed independently, so users can benefit both from the compression 
and the selection of only a subset of the image.

Other changes in this edition include:

      - Added a [de]compression API supporting all compression methods 
in the proposed
        updates to the FITS standard.
      - Wrong checksum calculation corrected.
      - Some problems with data segments that are bigger than 2GB corrected.
      - Header parsing performance optimization.
      - Comment style cards with a empty key value can now be used 
multiple times.
      - Alignment of hierarch headercard values deactivated.
      - The formatting of hierarch card keys can mow be be controlled. 
Two formats are  provided.