2014-04-29 Mathieu Malaterre * [r2866] .[CPY]: [2.1] Create the 2.1 branch 2014-04-28 Antonin Descampe * [r2864] [trunk] fixed warning in j2k.c about sign conversion 2014-04-28 Mathieu Malaterre * [r2863] [trunk] Finalize API for openjpeg 2.1 public header will only contains minimal information: major, minor and build version Prefix has been changed from OPENJPEG to OPJ for consistency SOVERSION has been removed from header (ABI vs API info) OPJ_VERSION string has been removed, since it is accessible already (PACKAGE_VERSION) Fixes issue 342 * [r2862] [trunk] Properly handle failure to compress (remove file) Fixes issue 323 * [r2861] [trunk] Missing endianess setting Fixes issue 337 * [r2860] [trunk] properly handle pkg-config on non-UNIX hosts Fixes issue 340 * [r2859] [trunk] Provide #ifdef blockers for OpenJPEG Version Fixes issue 342 2014-04-27 Antonin Descampe * [r2858] [trunk] fixed typo in NEWS 2014-04-25 Antonin Descampe * [r2857] [trunk] updated NEWS for 2.1.0 2014-04-24 Mathieu Malaterre * [r2855] [trunk] Make sure OpenJPIP compiles on WIN32 2014-04-24 Antonin Descampe * [r2852] [trunk] fixed reference to opj_stream_set_user_data_v3 in jpip sources. 2014-04-23 Antonin Descampe * [r2851] [trunk]removed '-x' option for index file as long as it's broken * [r2849] [trunk]re-formatted help display in executables * [r2848] [trunk]formatted some messages. Removed -version option to disambiguate with -v(erbose) option, added version info in help display * [r2847] [trunk]Replaced deprecated opj_stream_set_user_data function from API with its 'v3' version, and removed all other 'v3' suffixes from API. 2014-04-22 Antonin Descampe * [r2845] [trunk] fixed some warning and errors formatting and add a '-version' option to opj_compress and opj_decompress binaries 2014-04-17 Antonin Descampe * [r2838] [trunk] fixed warning related to r2837 and move new field at the end of parameters structure. * [r2837] [trunk] refactoring of rsiz, profiles, and extensions management 2014-04-03 Antonin Descampe * [r2835] [trunk] updated copyright and added copyright notice required by ISO, in each file; updated AUTHORS, NEWS 2014-04-02 Antonin Descampe * [r2832] [trunk] removed debug message in CMakeLists.txt * [r2831] [trunk] fixed a bug triggering SEGFAULT on debug mode (uninitialized structure) 2014-03-27 Mathieu Malaterre * [r2815] [trunk] Properly copy/paste code from the autotools gcc correctly fails compiling this code, I needed to move the code outside the main function to get the error. This actually match the autotools (incorrect) code. We'll see how they fix this code in future release. This make the code compatible with previous openjpeg version and remove C99 requirement. Fixes issue 316 2014-03-26 Mathieu Malaterre * [r2808] [trunk] Make sure to use long long (this is not C89). However this should be ok on most compilers nowadays Update issue 316 2014-03-26 Antonin Descampe * [r2806] [trunk] fixed warnings triggered on linux related to r2802 2014-03-26 Mathieu Malaterre * [r2804] [trunk] Run test suite on issue 316 dataset Update issue 316 2014-03-26 Antonin Descampe * [r2802] [trunk] add ability in opj_compress to input subsampled images through '-F' (aka raw option). Let the user input raw images with 444, 422, 420 (etc) subsampling. To be used in conjunction with '-mct 0' option to compress subsampled YCC images. 2014-03-26 Mathieu Malaterre * [r2801] [trunk] Add small utilities to extract JP2 files from PDF 2014-03-25 Mathieu Malaterre * [r2799] [trunk] Make sure to test value not pointer Update issue 302 2014-03-25 Antonin Descampe * [r2798] [trunk] fixed bug related to r2791 2014-03-25 Mathieu Malaterre * [r2796] [trunk] Add sentinel in case list empty. I was getting a cmake error `list sub-command REMOVE_ITEM requires list to be present` * [r2795] [trunk] Import change from 2.0 branch 2014-03-25 Antonin Descampe * [r2792] [trunk] fixed warning due induced by r2791 * [r2791] [trunk] add "-mct {0,1,2}" option to opj_compress to allow MCT to be disabled if needed. 2014-03-25 Mathieu Malaterre * [r2789] [trunk] Make sure to not use alpha pointer if it is not an RGBA input Fixes issue 310 * [r2786] [trunk] Rework previous commit r2610 This way we are able to gently deprecate the old API, and preserve ABI Update issue 306 * [r2785] [trunk] Add documentation 2014-03-24 Mathieu Malaterre * [r2773] [trunk] Make sure to exit early instead of looping on every single pixels * [r2772] [trunk] Fix typo in comment (this is not j2k_dump) 2014-03-18 Mathieu Malaterre * [r2767] [trunk] Revert r2764. It breaks ~230 tests as seen on continuous. Since no dataset is available, the issue 296 will be left open. Update issue 296 * [r2766] [trunk] User can now control source/target java version Fixes issue 303 * [r2764] [trunk] Import git commit fc884aee2b69c78500e65c3d05bf216791a9ea4a from ghostscript team Fixes issue 296 * [r2763] [trunk] Import git commit 99a6f1af177c15f4db475186b79d169c993494ef from ghostscript team Update issue 296 * [r2761] [trunk] Update big endian handling no such thing as WORDS_BIGENDIAN (prefer OPJ_BIG_ENDIAN) Update issue 302 * [r2756] [trunk] Now that dataset from issue 297 have been imported, run test suite. None of the dataset triggered any buffer overflow. Recent changes in trunk handle all those cases. Update issue 297 2014-03-17 Mathieu Malaterre * [r2754] [trunk] Explicitely reject file4/file6 scenario Update issue 286 * [r2751] [trunk] Update BSD-4 copyright into a BSD-3 copyright Fixes issue 300 2014-03-14 Mathieu Malaterre * [r2748] [trunk] Since r2747 we are now able to track which files are missing from the test_suite. Add them. * [r2747] [trunk] add mecanism to track of addition of new regression files in svn/data. It will report when a J2K files is added but no test is found in the test_suite file * [r2746] [trunk] Remove duplicate content from test_inline.c Thanks to Matthieu Darbois for report (and patch) Fixes issue 285 * [r2745] [trunk] rework code from r2463. Really there has been some code duplication from r2413 Thanks to John Rogers (oracle.com) for report * [r2744] [trunk] rework code from r2463. Really there has been some code duplication from r2413 Thanks to John Rogers (oracle.com) for report * [r2743] [trunk] Now that issue 165 dataset is in non regression repository, activate code to reject them Fixes issue 165 * [r2741] [trunk] Fix issue with & vs && Fixes issue 277 * [r2740] [trunk] Add internal implementation to dump all tiles/comp info Eg: opj_dump -f 8 -i input.j2k Update issue 3 * [r2739] [trunk] Prevent a leak when reading PPT markers Update issue 295 * [r2735] [trunk] Move INLINE definition within openjpeg.h header since application may use it * [r2733] [trunk] Apply final missing patch from sumatrapdf team: http://bugs.ghostscript.com/show_bug.cgi?id=694893 Update issue 231 * [r2732] [trunk] Rework r2731 since it breaks non-regression testing * [r2731] [trunk] Be more verbose about failure. Update issue 294 * [r2730] [trunk] Add a temporary work around for issue 293 As described in the bug report the default allocation mecanism for codeblock data is too small for those dataset (16bits). We would need a finer (more granular) mecanism to reallocated only on demand. For now this help the two failing tests to pass. Fixes issue 293 * [r2726] [trunk] Mark OpenJPEG to be ABI incompatible with previous one, move to SONAME 7 Also declare this is 2.1.0 (or very close) * [r2722] [trunk] Remove warnings about unused values introduced in r2710 Thanks to Matthieu Darbois for patch Fixes issue 290 2014-03-13 Mathieu Malaterre * [r2720] [trunk] Add some sanity checks when reading Ippm(i) segments Update issue 288 * [r2719] [trunk] Make sure to reallocate ppm data buffer when multiple Ippm(i) buffer are found This handle the case where remaining data is exactly Nppm(i) Fixes issue 287 * [r2715] [trunk] Add documentation to the checkmd5refs cmake module (used for regression testing) * [r2714] [trunk] Always report when decompression failed as return code value * [r2713] [trunk] Fix test NR-DEC-issue171.jp2-71-decode-md5 after recent code change * [r2712] [trunk] Add a double check for user input * [r2711] [trunk] Fix compilation on VS2010 (snprintf is not C89). C++11 should avoid this in the future * [r2710] [trunk] Just in case this would be helpful for other, here is the code used to track issue 80 Update issue 80 * [r2709] [trunk] Tweak documentation in t2.c. Be more verbose in case of failure. * [r2708] [trunk] Add some documention in the pi code * [r2707] [trunk] Add debug code to extract decompressed image right before PCLR handling Update issue 235 * [r2705] [trunk] Add a small helper to spli ppm into 3 pgm files Update issue 235 2014-03-12 Mathieu Malaterre * [r2703] [trunk] Make sure to use 8bits buffer when applying the ICC profile. Fixes issue 281 * [r2701] [trunk] Add debug info to be able to test integration with littlecms * [r2695] [trunk] TIFF files in test suite can have 1 or 3 components * [r2694] [trunk] Simplify code when reading in TIFF images * [r2692] [trunk] Make sure when reading POC that number of layers (layer end) is within acceptable bound. Fixes issue 80 2014-03-11 Mathieu Malaterre * [r2691] [trunk] Re-active old warning about missing SOP marker * [r2690] [trunk] Add debug info to track issue 80 Update issue 80 * [r2688] [trunk] Rework assertion to work on 32bits system * [r2687] [trunk] Fix remainings warnings on linux/32bits arch * [r2686] [trunk] Rework fseek ifdefs blockers, it breaks the 64bits behavior for fseeko * [r2685] [trunk] Fix compilation errors when JPWL and/or MJ2 are build 2014-03-10 Mathieu Malaterre * [r2683] [trunk] Another final round of fixes for sign conversion warnings. Fixes issue 256 * [r2682] [trunk] Another round of fixes for sign conversion warnings. Update issue 256 * [r2681] [trunk] Another round of fixes for sign conversion warnings. Update issue 256 * [r2680] [trunk] Another round of fixes for sign conversion warnings. Update issue 256 * [r2679] [trunk] Remove old warnings about deprecated API * [r2677] [trunk] Remove a set of warning about sign conversion Update issue 266 * [r2676] [trunk] Remove warning about sign conversion in color.c Update issue 256 * [r2672] [trunk] Run richter test suite, correct a regression introduced in r2668 * [r2670] [trunk] Tweak JP2 conformance test suite add documentation * [r2668] [trunk] Properly handle number of components to run kakadu test suite. Update issue 266 * [r2664] [trunk] Fix simple sign conversion warnings only visible on 32bits arch * [r2663] [trunk] Handle multi components files with PGM outputs * [r2661] [trunk] Properly handle PGM file generation when fullpath is specified in the cmd line This will break a very old behavior (since 2005), but avoid writing in un-expected places for users This also make the behavior similar with PGX file generation * [r2660] [trunk] Fix inversion in code convention * [r2659] [trunk] Fix typo in the code * [r2658] [trunk] rework actual code to handle DOS/UNIX end of lines in a transparent manner. This will help cross compilation cases * [r2657] [trunk] First pass to cleanup compare_dump_files * [r2656] [trunk] Fix compilation when TIFF lib is neither found nor compiled * [r2655] [trunk] Make sure to always initialize variable * [r2654] [trunk] Fix warnings about shadow variables * [r2653] [trunk] remove a warning when using strict prototype * [r2652] [trunk] Remove 2 warnings about set but not used variables 2014-03-07 Mathieu Malaterre * [r2651] [trunk] Upon failure, need to return proper return code * [r2650] [trunk] Add new test suite that run on kakadu conformance images * [r2644] [trunk] Cleanup commit. Rename compare family since comparePGX now support TIFF. Rename compareRAWimage for clarity since it merely compare files (not images). * [r2643] [trunk] Use C-style comments * [r2642] [trunk] Rework the code to simplify cleanup code * [r2638] [trunk] Also override the default error handler for TIFF files * [r2636] [trunk] Another round of fixes for sign conversion warnings. Update issue 256 * [r2635] [trunk] Another round of fixes for sign conversion warnings. Update issue 256 * [r2634] [trunk] Rework r2597, make sure test suite is passing now. Update issue 256 * [r2633] [trunk] Another round of fixes for sign conversion warnings. Update issue 256 * [r2632] [trunk] Another round of fixes for sign conversion warnings. Update issue 256 * [r2631] [trunk] Another round of fixes for sign conversion warnings. Update issue 256 * [r2630] [trunk] Another round of fixes for sign conversion warnings. Update issue 256 * [r2629] [trunk] Another round of fixes for sign conversion warnings. Update issue 256 * [r2628] [trunk] Redo r2623, and remove guilty section Update issue 256 * [r2627] [trunk] Revert r2623 for now * [r2626] [trunk] Another round of fixes for sign conversion warnings. Update issue 256 * [r2625] [trunk] Another round of fixes for sign conversion warnings. Update issue 256 * [r2624] [trunk] Another round of fixes for sign conversion warnings. Update issue 256 * [r2623] [trunk] Another round of fixes for sign conversion warnings. Update issue 256 * [r2622] [trunk] Another round of fixes for sign conversion warnings. Update issue 256 * [r2621] [trunk] Another round of fixes for sign conversion warnings. Update issue 256 * [r2620] [trunk] Another round of fixes for sign conversion warnings. Update issue 256 * [r2619] [trunk] As reported by clang analyzer thoses values were never read * [r2618] [trunk] Remove mem leak and warning reported by clang * [r2617] [trunk] Remove simple warnings about format * [r2616] [trunk] Remove simple warning about functions not used * [r2615] [trunk] Another round of fixes for sign conversion warnings. Update issue 256 * [r2614] [trunk] Another round of fixes for sign conversion warnings. Update issue 256 * [r2613] [trunk] Fix simple warning about sign conversion Update issue 256 * [r2612] [trunk] Another of fixes for warning about sign conversion Update issue 256 * [r2611] [trunk] Fix warning about sign conversion Update issue 256 * [r2610] [trunk] un-deprecated opj_stream_destroy API, import patch manually In some case user still want to be able to call opj_stream_destroy. Fixes issue 227 * [r2609] [trunk] Update memory management mecanism Fixes issue 253 * [r2608] [trunk] Remove self-assignement. Fixes issue 273 * [r2605] [trunk] Fix compilation in visual studio, inline cannot be used directly. Fixes issue 272 * [r2604] [trunk] Remove warning about implicitly defined function * [r2601] [trunk] Remove TIFF warning handler. This help test suite run smoothly on windows by removing the message box. Fixes issue 276 * [r2600] [trunk] Fix mem leaks reported by cppcheck * [r2599] [trunk] Simplify code and remove bogus out of bound check 2014-03-06 Mathieu Malaterre * [r2598] [trunk] Revert r2597 for now, since it breaks a lot of tests * [r2597] [trunk] Remove warning about sign conversion. Update issue 256 * [r2595] [trunk] Remove some easy warnings * [r2594] [trunk] Remove warning about sign conversion. Update issue 256 * [r2593] [trunk] Remove warning about sign conversion. Update issue 256 * [r2592] [trunk] Remove warning about sign conversion. Update issue 256 * [r2591] [trunk] Remove warning about sign conversion. Update issue 256 * [r2590] [trunk] Remove warning about sign conversion. Update issue 256 * [r2589] [trunk] Remove some easy warnings about sign compare reported by gcc * [r2588] [trunk] Remove out of bound access in code and simplify code. this is a test tool, security is not an issue here. * [r2587] [trunk] Remove simple warning about unused variables * [r2586] [trunk] Make sure that all output files from the test suite are using a different output name Fixes issue 271 2014-03-05 Mathieu Malaterre * [r2577] [trunk] Handle cmap where direct use is specified (issue235_cmapsubbox.jp2) I doubt the old code ever work. The new code copy the old codestream bytes into the new components (instead of copying the pointer). Technically the issue235.jp2 file should be handled since I.5.3.5 specifies that: ... If the JP2 Header box does not contain a Component Mapping box, the components shall be mapped directly to channels, such that component i is mapped to channel i. ... Update issue 235 2014-03-04 Mathieu Malaterre * [r2572] [trunk] Make sure to reject images such as 1336.pdf.asan.47.376.jp2 Fixes issue 267 * [r2571] [trunk] remove more conversion warning in multi-lines macro. C allow simple tricks such as to find those: gcc -E tcd.c | sed '/^\#/d' | indent -st -i2 > clean.c Update issue 256 * [r2570] [trunk] Remove some conversion warnings reported by gcc Update issue 256 * [r2569] [trunk] Remove another warning. ftell returns long int Update issue 256 * [r2568] [trunk] Remove more conversion warnings. Update issue 256 * [r2567] [trunk] Remove warning about conversion. Update issue 256 * [r2566] [trunk] Add a TIFF compare function (PEAK/MSE) Fixes issue 100 * [r2565] [trunk] Revert portion of r2562, need to investigate issue. Update issue 269 2014-03-03 Mathieu Malaterre * [r2562] [trunk] Remove remaining conversion issues in mqc.c * [r2561] [trunk] Remove remainings conversion issues in t2.c + clock.c * [r2560] [trunk] Remove a simple warning about int -> float conversion * [r2559] [trunk] rework code to avoid a warning. this also remove a division and prefer bit-shift * [r2558] [trunk] Remove some more warnings about conversion * [r2557] [trunk] Prefer compile time constant to avoid conversion double -> int * [r2556] [trunk] Review warning about types conversion, and add cast when needed * [r2553] [trunk] Remove some simple warnings about conversion * [r2552] [trunk] fgets takes an int as parameter * [r2551] [trunk] Remove warning about cast double -> int. Instead use bit-shifting (compile time computation) * [r2550] [trunk] Remove easy to fix warnings about conversion * [r2549] [trunk] Import left over from patch submitted as issue 225 (comment #14) Update issue 225 * [r2548] [trunk] remove some simple warnings about unused params * [r2547] [trunk] remove all api with invalid FILE* parameters which could leads to issues when applications are compiled with different flags from openjpeg. Fixes issue 198 * [r2546] [trunk] Explicitely trigger issue 229, there may be an impact in code execution. The assert may get removed afterward Update issue 229 * [r2545] [trunk] Remove simple warnings * [r2544] [trunk] Remove 3635.pdf.asan.77.2930.jp2 from opj_dump tests Update issue 250 * [r2539] [trunk] Avoid conflicting declarion for fseek in openjp2 vs fseek in openjpip * [r2538] [trunk] Fix md5 values for decode-md5 tests. Update issue 250 * [r2537] [trunk] Remove deprecated functions. Update issue 268 * [r2536] [trunk] Remove warning about unused function Update issue 268 * [r2535] [trunk] Remove deprecated functions (not called anywhere in the code base). Update issue 268 * [r2533] [trunk] Remove unused variables from test 2014-03-02 Mathieu Malaterre * [r2532] [trunk] Remove some simple warnings about var being set but not used * [r2531] [trunk] Remove warning about -2147483648 (aka INT_MIN), causing this decimal constant is unsigned only in ISO C90 * [r2530] [trunk] Remove simple warnings about comments, missing case in switch statement 2014-03-01 Mathieu Malaterre * [r2529] [trunk] Fix compilation on MSVC compiler 2014-02-28 Mathieu Malaterre * [r2528] [trunk] Partially rework r2506, original patch from issue 171 was totally bogus (untested?) and did break most of the test (eg. p1_04 family) convert.c duplicate a lot of code, this patch only adresses the PGX codec section of the code. Update issue 171 Update issue 264 * [r2527] [trunk] Make sure to run *-decode-md5 only after *-decode tests. Update md5 since recent code change. * [r2526] [trunk] Remove easy uint32 <-> int32 conversions * [r2525] [trunk] Re-import chunk from r2458 that made textGBR.jp2 to fail otherwise Update issue 225 * [r2524] [trunk] Rework r2458, instead we should reject 451.pdf.SIGSEGV.ce9.3723 since tile-parts are out of order (as per kakadu behavior). Update issue 225 * [r2523] [trunk] Mark a file as being invalid for opj_dump * [r2522] [trunk] Fix JPIP test with updated md5sum * [r2521] [trunk] Really apply r2460 this time, but fix the original typo in the submitted patch Update issue 225 * [r2520] [trunk] final round of minor cleanups * [r2519] [trunk] Remove PNG generation code for now * [r2518] [trunk] Prefer strcpy since it copies trailing 0 directly * [r2517] [trunk] Simplify code to centralize code cleanup * [r2516] [trunk] Some code cleanup add sentinels * [r2515] [trunk] Remove a simple warning, and use static to detect unused functions. Simplify function declarations. * [r2513] [trunk] run test suite on dataset from issue 171 Fixes issue 171 * [r2511] [trunk] Remove some simple warnings in opj_dump * [r2510] [trunk] Merge final patch from issue 171 with code simplification. also handles signed case properly. Fixes issue 171 * [r2508] [trunk] Fix find_path behavior within cross-compilation execution 2014-02-27 Mathieu Malaterre * [r2507] [trunk] Import patch from issue 218. No dataset to check, so blindly applied it. Fixes issue 218 * [r2506] [trunk] Import patch from issue 171. Only the imagetoraw part was not applied. Update issue 171 * [r2505] [trunk] Try to run test suite on invalid dataset from sumatrapdf team Update issue 225 * [r2503] [trunk] run test suite to illustrate patch from r2458 * [r2501] [trunk] Make sure that opj_decompress only output a single component Update issue 110 * [r2496] [trunk] add issue 46 to the test suite Fixes issue 46 * [r2495] [trunk] improve command line parser syntax checking * [r2493] [trunk] add a sentinel in debug code * [r2492] [trunk] simplify code using for loop 2014-02-26 Mathieu Malaterre * [r2491] [trunk] Import test case from issue 208. This appears to be fixed now. Fixes issue 208 * [r2489] [trunk] Run test suite on dataset from issue 211. Test is now passing nicely. Fixes issue 211 * [r2487] [trunk] run test suite on dataset from issue 135. Add fake md5sum to make sure tests is failing. Update issue 135 * [r2485] [trunk] Update missing md5sums * [r2484] [trunk] kakadu does not support 451.pdf.SIGSEGV.ce9.372 we should not pretend to be able to decompress it Update issue 225 * [r2483] [trunk] Run test suite on issue 141. Add a fake dataset to report that test is failing. Update issue 141 * [r2481] [trunk] Add more details in case test fails * [r2480] [trunk] run test suite on input dataset from issue 134 Fixes issue 134 * [r2478] [trunk] fixes test NR-DEC-file409752.jp2-40-decode, use proper naming convention (need to keep file extension) * [r2477] [trunk] add test suite for issue 142 Fixes issue 142 * [r2476] [trunk] When no output PGX file is found, report an error * [r2474] [trunk] Add missing include dir for wx app * [r2473] [trunk] Remove some simple warnings about cast, and unused functions * [r2471] [trunk] Remove a simple warning about a cast * [r2470] [trunk] revert r2460 since it breaks ETS-C1P0-p0_03.j2k-decode. Update issue 225 * [r2469] [trunk] Revert r2453 since it breaks conformance test ETS-C1P1-p1_06.j2k-decode * [r2468] [trunk] Import patch from sumatrapdf team. This handle testcase 1336.pdf.asan.47.376 Update issue 225 * [r2467] [trunk] Import patch from sumatrapdf team. This handle testcase 1888.pdf.asan.35.988 Update issue 225 * [r2466] [trunk] Import patch from sumatrapdf team. This handle testcase 1851.pdf.SIGSEGV.ce9.948 Update issue 225 * [r2465] [trunk] Import patch from sumatrapdf team. Start using new color space when found. Update issue 225 * [r2464] [trunk] Import patch from sumatrapdf team. This handle some Part-2 compliant file, by being more tolerant in COLR box. No testcase for now. Fixes issue 247 * [r2463] [trunk] Import patch from sumatrapdf team. This handle some testcase with no input dataset, but changes looks ok. Update issue 225 * [r2462] [trunk] Import patch from sumatrapdf team. This handle testcase 2977.pdf.asan.67.2198 Update issue 225 * [r2461] [trunk] Import (ugly) patch from sumatrapdf team. This feels like a hack rather than a solution. Fixes issue 226 * [r2460] [trunk] Import patch from sumatrapdf team. This handle testcase 3635.pdf.asan.77.2930 Update issue 225 * [r2459] [trunk] Import patch from sumatrapdf team. Original commit is 5b0c9985e3359aca9b3fcfd94424166aa61a141a. Update issue 225 * [r2458] [trunk] Import patch from sumatrapdf team. This handle testcase 451.pdf.SIGSEGV.ce9.3723 Update issue 225 * [r2457] [trunk] Import patch from sumatrapdf team. This handle testcase 2.pdf.SIGFPE.706.1112 Update issue 225 * [r2456] [trunk] Import patch from sumatrapdf team. This handle testcase 1610.pdf.SIGSEGV.59c.681 Update issue 225 * [r2455] [trunk] Import patch from sumatrapdf team. This handle testcase 2539.pdf.SIGFPE.706.1712 Update issue 225 * [r2454] [trunk] Import patch from sumatrapdf team. This handle testcase 4035.pdf.SIGSEGV.d8b.3375 Update issue 225 * [r2453] [trunk] Import patch from sumatrapdf team. Add detection for testcase 1802.pdf.SIGSEGV.36e.894 Update issue 225 * [r2452] [trunk] Import patch from sumatrapdf team. Add a new function to check color box. Update issue 225 * [r2451] [trunk] Import patch from sumatrapdf team. This patch adds new colorspace handling. Update issue 225 * [r2450] [trunk] Import patch from sumatrapdf team. This adds alpha handling. This patch changes ABI. Update issue 225 * [r2449] [trunk] Update file extension function to handle cases where filename containes multiples dots. Update issue 250 * [r2447] [trunk] Clearly indicate that mem-b2ace68c-1381.jp2 should not be decompressed. Update issue 250 * [r2446] [trunk] Fix TestJPIP1 on case sensitive system * [r2445] [trunk] regex would not be used to remove java deprecation warning from dashboard * [r2444] [trunk] Remove simple warnings about unused functions * [r2443] [trunk] Properly store and use value returned by fread. Also invert nmemb and size in fread call. Fixes issue 262 * [r2442] [trunk] Fix invalid write access in JPIP code. Fixes issue 261 2014-02-25 Mathieu Malaterre * [r2436] [trunk] Run decoder on input dataset from issue 254. Add fake md5sum to make sure the test is failing. Update issue 254 * [r2433] [trunk] deprecationg warning issued by java compiler should be ignored * [r2432] [trunk] Create a new static *_impl function to avoid a warning triggered by the deprecation mecanism Fixes issue 257 * [r2430] [trunk] New decoded files are passing. Adding reference files. Update issue 225 * [r2428] [trunk] For now assume any provided stream should not get decompressed. Update issue 225 * [r2427] [trunk] Run decoder on issue 229 test cases Update issue 229 * [r2419] [trunk] Revert r2415 for now since it introduce large regression * [r2418] [trunk] Import commit 87b08a096bb8ad61f9dbe4811e208d9c9d7fe63b from ghostpdl don't define lrintf for MSVC 2013 (fix compilation breakage) Update issue 225 * [r2417] [trunk] Import commit 4cee6ceab21025079f439bb152fb9d8ae8c5c832 from ghostpdl Bug 694906: fix potential heap overflow in opj_t2_read_packet_header Update issue 225 * [r2416] [trunk] import commit 83dad6a76536222a3a51146f942e733a2e90ec52 from ghostpdl fix potential NULL-pointer dereference caused by testing the wrong variable after opj_realloc (happens only in OOM situations) Update issue 225 * [r2415] [trunk] Import commit fc884aee2b69c78500e65c3d05bf216791a9ea4a from ghostpdl prevent heap overflow in opj_t2_read_packet_header Also prevent a double-free of segment data under OOM conditions. Problem found in a test file, 1802.pdf.SIGSEGV.36e.894 supplied by Mateusz "j00ru" Jurczyk and Gynvael Coldwind of the Google Security Team using Address Sanitizer. Many thanks! Update issue 225 * [r2414] [trunk] Import commit 8ec8321df613986e8642e2975f7182072eb4be62 from ghostpdl make opj_stream_set_user_data accept a NULL stream for consistency with opj_stream_set_* which does so. Update issue 225 * [r2413] [trunk] Import commit f4139d702559649e577a5df9cfd64b0ca6107a7a from ghostpdl Several functions accept a buffer size but never actually check whether the buffer overflows during reading/writing. This fixes all cases where a size variable has explicitly been marked as unused (through a (void) cast). This was discovered while investigating an assertion caused by 7cc691f332f26802c64cdc47e17bff8b_signal_sigabrt_7ffff6d59425_2247_2509.pdf among others. Thanks to Mateusz Jurczyk and Gynvael Coldwind of the Google Security Team for providing the example files. * [r2411] [trunk] Clarify raw input in openjpeg. LSB vs MSB are now documented. Fixes issue 62 * [r2410] [trunk] Run test case for issue 228. Output has been validated. Fixes issue 228 * [r2407] [trunk] Run sumatrapdf test cases 2014-02-24 Mathieu Malaterre * [r2404] [trunk] add missing private lib. Update issue 223 * [r2402] [trunk] Make sure to always output pgx file. Handle case where multiple pgx files are generated. Fixes all tests. * [r2401] [trunk] add test demonstrating a regression in 1.5.0 and up. Update issue 205 * [r2399] [trunk] improve test suite by checking against a reference md5 * [r2398] [trunk] Fix a simple mem leak * [r2397] [trunk] Fix two simple warnings about sign mismatch * [r2395] [trunk] Fix warning about signed/unsigned mismatch * [r2394] [trunk] Make sure to handle ret value, and properly initialize output file * [r2393] [trunk] Make sure to handle ret value in test * [r2392] [trunk] Prefer the new style cmake: add_test command Fixes issues 258 * [r2388] [trunk] Fix a warning about type conversion. Use a trick where unsigned wrapping is legal 2014-02-12 Antonin Descampe * [r2375] [trunk] replaced fprintf with event_mgr in cinema-related stuff in j2k.c * [r2374] [trunk] moved logic related to cinema profiles in library (was in opj_compress.c). This enables one using the library with its own executable to easily ask for cinema profiles. * [r2373] [trunk] added a warning when tif conversion changes input image bitdepth * [r2372] [trunk]re-indenting convert.c 2014-02-07 Mathieu Malaterre * [r2371] Update kakadu cmake module 2014-01-23 Antonin Descampe * [r2369] [trunk] added some tests for Cinema2k 48fps and Cinema 4k 24 fps. Fixed a bug in comparePGXimages.c leading to huge Test.xml file uploaded to DashBoard * [r2367] [trunk] fixed another DCI compliance bug 2014-01-22 Antonin Descampe * [r2366] [trunk] uncomment tests mistakenly commented in previous commit * [r2365] [trunk] fixed several bugs in cinema mode (2K 24/48 fps, and 4K). Trunk now produces compliant DCI code-streams. 2014-01-16 Antonin Descampe * [r2364] [trunk] fixed indentation in opj_compress.c, renamed 2 internal functions, added some comments 2014-01-16 Antonin Descampe * [r2363] [trunk] fixed DCI-compliant codestream generation (-cinema2K and - 2013-10-30 Antonin Descampe * [r2352] trunk: fixed cmake config so as to be able to link opj_jpip_server with FCGI. * [r2351] trunk: fixed cmake config so as to be able to link opj_jpip_server with FCGI. 2013-10-28 Antonin Descampe * [r2350] trunk: disabling class-0 conformance tests for now (bad implentation of tests) 2013-10-15 Rex Dieter * [r2348] complete openjpeg2 pkgconfig support, bug #68 2013-09-26 Mathieu Malaterre * [r2347] [trunk] Import patch from bug #241 2013-07-11 Antonin Descampe * [r2344] test: modified class-0 conformance testing to comply with constraints 2013-03-25 Mickaël Savinaud * [r2322] [trunk] use the private version of opj_config to avoid compilation error into mj2 * [r2321] [trunk] use opj_config_private.h into the opj_inttypes.h * [r2320] [trunk] separate the opj_config file between public and private part to avoid expose unused variables into the public API. * [r2319] [trunk] update test suite because now illegalcolortransform.j2k could be decoded 2013-03-24 Mickaël Savinaud * [r2318] [trunk] enhance the support of sYCC into opj_decompress application (thanks winfried). * [r2317] [trunk] managed nicely the case where TPSot value is incorrect (thanks winfried). Update the test suite to check this case. Updates issue 202. Updates issue 206. Update issue 208 2013-03-17 Mickaël Savinaud * [r2315] [trunk] add access to opj_apps_config.h to old applications * [r2314] [trunk] manage correctly the case where meth value is not correct (thanks winfried) * [r2313] [trunk] correct issue 188 (thanks winfried) and add test about it. Fixes issue 188 * [r2308] [trunk] use OPJ_ prefix on HAVE variable also for tests * [r2307] [trunk]remove wrong code in mj2 library detected by winfried (thanks to you) * [r2306] [trunk] use everywhere the new opj_ prefix for HAVE variables and use the opj_apps_config file (thanks to winfried) 2013-03-13 Mickaël Savinaud * [r2304] [trunk] use the home made macro to ensure the existence of some include file * [r2303] [trunk] use the opj_ prefix for the HAVE_FSEEKO variable * [r2302] [trunk] correct missing variable in openjpeg config file 2013-03-03 Mickaël Savinaud * [r2301] [trunk] remove some unused variable from opj_config and rename other one with opj_ prefix * [r2300] [trunk] split into two config files config options related to the lib and to the application. 2013-02-18 Mickaël Savinaud * [r2298] [trunk] add the prefix opj_ to deprecated everywhere 2013-02-17 Mickaël Savinaud * [r2297] [trunk] update documentation of new functions _v3 and mark deprecated related functions * [r2296] [trunk] move to the new API for function opj_stream_create_default_file_stream. Use now opj_stream_create_default_file_stream_v3 (WIP) 2013-02-16 Mickaël Savinaud * [r2295] [trunk] move to the new API for function opj_stream_destroy. Use now opj_stream_destroy_v3 (WIP) * [r2294] [trunk] rename deprecated macro with opj_ prefix and use it for opj_stream_destroy function * [r2293] [trunk] add functions to avoid to use FILE* into the API (thanks winfried). Update issue 120 and update issue 198 2013-02-04 Mickaël Savinaud * [r2292] [trunk] add a test related to issue 202 2013-02-03 Mickaël Savinaud * [r2290] [trunk] update convert about png writing with additional metadata (thanks winfried) * [r2287] [trunk] update test suite for issue 5 and 62 * [r2285] [trunk] backout one part of the rev 2266 2013-01-03 Mathieu Malaterre * [r2272] [trunk] JP3D: Fix issue with Z handling in packet iterator * [r2271] [trunk] JP3D: Handles errors properly * [r2270] [trunk] JP3D: remove old debug 2013-01-02 Mathieu Malaterre * [r2269] [trunk] JP3D: convert from DOS to UNIX eol * [r2268] [trunk] JP3D: remove exe bit from source files 2012-12-18 Mickaël Savinaud * [r2266] [trunk] merge from branch 1.5 a correction about large code-block data (decoding part) and add tests about issue 5 and issue 62 2012-12-11 Mathieu Malaterre * [r2263] [trunk] remove left over from v2 transition * [r2262] [trunk] JP3D: completely redo the opj_tgt_create function. Take as input the initial opj_tgt_create as found in openjpeg 1.5, and extend with the 3rd dimension. 2012-12-07 Rex Dieter * [r2259] Doxyfile: HTML_TIMESTAMP = NO helps when comparing the output of multiple runs, and distros will appreciate avoiding multilib conflicts. 2012-12-04 Mickaël Savinaud * [r2255] [trunk] correct the new name of OPJ_HAVE_STDINT_H everywhere * [r2254] [trunk] remove SSIZE which is unused in openjpeg and rename HAVE_STDINT_H to OPJ_HAVE_STDINT_H * [r2253] [trunk] add openjpeg.h in some files of openjpip to use some opj type (thanks winfried) 2012-12-03 Mickaël Savinaud * [r2252] [trunk] move typedef from opj_stdint.h to openjpeg.h and add an include of stdio.h in openjpeg.h. Fixes issue 197 * [r2251] [trunk] correct the encoding of colr box information (thanks Winfried) * [r2250] [trunk]add the pkgconfig support for openjp2 (thanks Winfried) 2012-11-28 Mathieu Malaterre * [r2249] [trunk] JP3D codec was trying to pull information from the registration parameter in COM. This is non-standard. * [r2248] [trunk] As per T.809 code block are not shifted in JP3D Table A.7 3D code-block width, height and depth exponent values xcb = value, ycb = value or zcb = value. NOTE – This redefines Rec. ITU-T T.800 | ISO/IEC 15444-1 significantly! The 3D code-block width, height and depth are limited to powers of two with the minimum size being 20 and the maximum being 210. Further, the 3D code-block size is restricted so that 4 ≤ xcb+ycb+zcb ≤ 18. * [r2247] [trunk] JP3D: this is safe to write comment using Registration: ISO-8859-15 * [r2246] [trunk] JP3D fix issue with Registration value in COM marker. Move COM marker just before COD. * [r2245] [trunk] jp3d: fix command line parsing for subsampling * [r2244] [trunk] Fix issue with ITU-T T.809 (05/2011) - ZSI is now known as NSI (FF54) - NSI now stores Ndim (=3) - CAP extended marker was not set properly * [r2243] [trunk] Fix compilation issue where size_t would be undefined 2012-11-27 Mickaël Savinaud * [r2242] [test] update the list of files which will be gracefully rejected when they are dump * [r2241] [trunk] add empty test about the issue 197. Updates issue 197 2012-11-26 Mickaël Savinaud * [r2240] [trun[trunk] check if the dx and dy steps are not incorrect according to the JPEG2000 norm (thanks Even). Test associated to this issue should nicely fail. Fixes issue 194 * [r2239] [trunk] check if the number of tiles is not incorrect regarding the JPEG2000 norm (thanks Even). Test associated to this issue should nicely fail. Fixes issue 163 * [r2238] [trunk] test about issue 192 should nicely fail * [r2237] [trunk] check the index of the component in read_qcc (thanks Even). fixes issue 192 * [r2235] [trunk] update test suite with issue 191 to 194 and update patch for 191 which should fail properly