Makefile : can specify version number

lz4.c : corrected compression efficiency issue 97 in 64-bits chained mode (-BD) for streams > 4 GB (thanks Roman Strashkin for reporting)

Makefile : make install installs both lz4 and lz4c (Jorge Aparicio)
Makefile : removed -Wno-implicit-declaration compilation switch
lz4cli.c : include <stduni.h> for isatty() (Luca Barbato)
lz4.h : introduced LZ4_MAX_INPUT_SIZE constant (Shay Green)
lz4.h : LZ4_compressBound() : unified macro and inline definitions (Shay Green)
lz4.h : LZ4_decompressSafe_partial() : clarify comments (Shay Green)
lz4.c : LZ4_compress() verify input size condition (Shay Green)
bench.c : corrected a bug in free memory size evaluation
cmake : install into bin/ directory (Richard Yao)
cmake : check for just C compiler (Elan Ruusamae)

No commit message

lz4cli.c : Minor modifications, for easier extensibility
COPYING : added license file
LZ4_Streaming_Format.odt : modified file name to remove white space characters
exe : .exe suffix now properly added only for Windows target

lz4c still there, supports additional gzip arguments, but also keep compatibility with legacy commands
lz4 (& lz4c) display version number
Fix : Sun Studio : compatible #pragma directive (issue 81)
Fix : compatible with Objective-C (iOS) (issue 79)
Fix : minor warnings using Visual Studio x64 (issue 80)
Changed : source file lz4c.c renamed lz4cli.c

fullbench : added bench of LZ4_compress_continue(), LZ4_compress_limitedOutput_continue(), LZ4_compressHC_continue() and LZ4_compressHC_limitedOutput_continue()

Improved speed of LZ4_decompress_fast() with GCC
Improved speed of LZ4_decompress_safe() for 32-bits
Made the fast LZ4 compression compatible with low-memory systems (buffer address < 64K). Thanks Francois Gretief for report and suggestion.
Makefile : added fuzzer32
Makefile : added fullbench32
fullbench : added ability to select one specific function to benchmark
lz4.c : copy macros follow memcpy() arguments convention
Small coding style modifications, hinted by cppCheck. 

Corrected a bug into the compact XXH32() function, which affects Block Checksum results. Update is recommended if you use Block Checksum with lz4c. (Note : block checksum is disabled by default). Thanks Adrien Grand for detailed bug tracking and fixing.
Corrected a bug affected Heapmode version of LZ4_compress() (Note : heapmode is disabled by default)

lz4c : made display and arguments more compatible with gzip, for easier integration with tar (patch by Yaakov Selkowitz)
Correction : large files support on 32-bits unix (reported by Karthik Rajeswaran)
lz4c : reduce the amount of displayed information in default mode; introduce a verbose mode
lz4c : changed help message
Updated xxHash to r31
Made bench.c compatible with tcc
Corrected : a few minor warnings found by CppCheck, as suggested by Brian White
lz4.c : Pushed BIG_ENDIAN_NATIVE_BUT_INCOMPATIBLE farther in the code, since it is reported as providing little benefit
Corrected : minor 64K input condition, detected by Mat Hostetter

Added : fullbench : can select compression tests or decompression tests
Removed extern inline, for compatibility with GNU89, as reported by Maciej Adamczyk
lz4.c : made forceinline more explicit
Decompression : corrected corner case behaviors (inputSize == 0 and outputSize == 0), thanks Adrien for detailed suggestions
Makefile : Removed -march=native parameter, due to incompatibility with some GCC versions 

Decompression speed improved under GCC
Improved speed of LZ4_decompress_safe_partial()
Added new utility : fullbench 
Modified x64 detection macro, as suggested by David Karner
Improved Fuzzer tool
Updated xxHash to r30

New experimental mode : compress blocks using data from previous blocks (option -BD) (limitation : -hc mode only)
Changed deprecated function names to "static", in order to avoid duplicate definition. Thanks Maciej Adamczyk for reporting.
Changed a few command line options
Prettify help text

Cleaner C99 detection, thanks to Elad
Modified Basic Types, as recommended by Wayne Scott
Simplified compression function
Corrected : warning on missing #undef in lz4hc, thanks Arseny Kapoulkine
Modified length loops as suggested by Wouter Vermaelen
Modified pragma pack directive for older versions of GCC, as reported by Peter Costello
Updated xxHash to r29
Indent lz4hc.c with spaces 

- New function : LZ4_decompress_safe_partial(), to decompress just enough data within a compressed block, saving CPU cycles
- New source files : lz4_decoder.h, lz4_encoder.h, lz4hc_encoder.h
- Improved speed of LZ4_decompress_fast()
- Improved speed for compression of small blocks < 64KB
- Improved speed for HC compression

Updated : LZ4 Streaming Format.odt, to version 1.4
New : LZ4c now supports Stream Checksum (default) and Skippable chunks
Updated : Fuzzer, testing LZ4_compressHC_limitedOutput()

Added : LZ4c command-line utility, supporting the new streaming format
Added : xxhash library
Removed : lz4demo is now replaced by lz4.c
Removed : a few level 4 warnings (issue 64)
Updated : makefiles

Updated : cmake/pack/CMakeLists.txt, by Dmitry Cherepanov
lz4demo : CLI accept aggregated commands
lz4demo : detect overwrite output
lz4demo : new commands options (-hc, -y)

- Improved decoding speed, even more for 64-bits, and "safe" version
- Slight speed increase for LZ4-HC
- Pushed a useless parameter down the list in lz4.c

Corrected issue 60 in lz4.h, reported by Takayuki Matsuoka
Added : a cmake/pack installer, by Dmitry Cherepanov

Bench mode : option to pause at the end

Removed unnecessary assignment within LZ4_uncompress_unknownSize() (reported by Shivram)

Fuzzer : more tests cases
lz4demo : detect write errors. Thanks to Dima Tisnek
bench.c : compatibility with Solaris 64. Thanks to Thorbjørn Willoch
LZ4_compressBound() : now both in inline function and macro format. Thanks to Jacob Gorm Hansen

Corrected issue 40 : Detect early ending of compressed stream. Thanks Adrian Grand.
Corrected issue 41 : minor comment editing on lz4.h

minor macro renaming (fuzzer.c)

- Corrected issue 31 : LZ4 correctly accepts compressing data when the output buffer has exactly the required size (it was a bit over-cautious in previous version).
- Added : a fuzzer tool
Thanks to Andrew Mahone, for contribution on both points

Correct issue 36 on LZ4_uncompress_unknownOutputSize(). Thanks to Clayton Stangeland and Maciej Adamczyk for notifying.
Converted tabs to space

LZ4.c : changed inline to "static inline" to comply with limitations from GCC/LLVM combination under OS-X.

Removed : functions with explicit *ctx management (LZ4_compressCtx & LZ4_compress64kCtx). Functions are still present in the .c
Changed : LZ4_compressBound() now a macro

Makefile : default produces native binary, all produce native & 32-bits binaries