Files
mozjpeg/java
DRC df9dbff830 TurboJPEG: New param to limit virt array mem usage
This corresponds to max_memory_to_use in the jpeg_memory_mgr struct in
the libjpeg API, except that the TurboJPEG parameter is specified in
megabytes.  Because this is 2023 and computers with less than 1 MB of
memory are not a thing (at least not within the scope of libjpeg-turbo
support), it isn't useful to allow a limit less than 1 MB to be
specified.  Furthermore, because TurboJPEG parameters are signed
integers, if we allowed the memory limit to be specified in bytes, then
it would be impossible to specify a limit larger than 2 GB on 64-bit
machines.  Because max_memory_to_use is a long signed integer,
effectively we can specify a limit of up to 2 petabytes on 64-bit
machines if the TurboJPEG parameter is specified in megabytes.  (2 PB
should be enough for anybody, right?)

This commit also bumps the TurboJPEG API version to 3.0.1.  Since the
TurboJPEG API version no longer tracks the libjpeg-turbo version, it
makes sense to increment the API revision number when adding constants,
to increment the minor version number when adding functions, and to
increment the major version number for a complete overhaul.

This commit also removes the vestigial TJ_NUMPARAM macro, which was
never defined because it proved unnecessary.

Partially implements #735
2023-11-14 10:19:06 -05:00
..

TurboJPEG Java Wrapper
======================

The TurboJPEG shared library can optionally be built with a Java Native
Interface wrapper, which allows the library to be loaded and used directly from
Java applications.  The Java front end for this is defined in several classes
located under org/libjpegturbo/turbojpeg.  The source code for these Java
classes is licensed under a BSD-style license, so the files can be incorporated
directly into both open source and proprietary projects without restriction.  A
Java archive (JAR) file containing these classes is also shipped with the
"official" distribution packages of libjpeg-turbo.

TJExample.java, which should also be located in the same directory as this
README file, demonstrates how to use the TurboJPEG Java API to compress and
decompress JPEG images in memory.


Performance Pitfalls
--------------------

The TurboJPEG Java API defines several convenience methods that can allocate
image buffers or instantiate classes to hold the result of compress,
decompress, or transform operations.  However, if you use these methods, then
be mindful of the amount of new data you are creating on the heap.  It may be
necessary to manually invoke the garbage collector to prevent heap exhaustion
or to prevent performance degradation.  Background garbage collection can kill
performance, particularly in a multi-threaded environment (Java pauses all
threads when the GC runs.)

The TurboJPEG Java API always gives you the option of pre-allocating your own
source and destination buffers, which allows you to re-use those buffers for
compressing/decompressing multiple images.  If the image sequence you are
compressing or decompressing consists of images of the same size, then
pre-allocating the buffers is recommended.


Installation Directory
----------------------

The TurboJPEG Java Wrapper will look for the TurboJPEG JNI library
(libturbojpeg.so, libturbojpeg.dylib, or turbojpeg.dll) in the system library
paths or in any paths specified in LD_LIBRARY_PATH (Un*x), DYLD_LIBRARY_PATH
(Mac), or PATH (Windows.)  Failing this, on Un*x and Mac systems, the wrapper
will look for the JNI library under the library directory configured when
libjpeg-turbo was built.  If that library directory is
/opt/libjpeg-turbo/lib32, then /opt/libjpeg-turbo/lib64 is also searched, and
vice versa.

If you installed the JNI library into another directory, then you will need
to pass an argument of -Djava.library.path={path_to_JNI_library} to java, or
manipulate LD_LIBRARY_PATH, DYLD_LIBRARY_PATH, or PATH to include the directory
containing the JNI library.