Merge pull request #802 from facebook/variants

fix #788 : fix /programs binary variants
This commit is contained in:
Yann Collet
2017-08-19 11:57:50 -07:00
committed by GitHub
6 changed files with 290 additions and 253 deletions
+1
View File
@@ -76,6 +76,7 @@ zlibwrapper:
.PHONY: test shortest .PHONY: test shortest
test shortest: test shortest:
$(MAKE) -C $(PRGDIR) allVariants
$(MAKE) -C $(TESTDIR) $@ $(MAKE) -C $(TESTDIR) $@
.PHONY: examples .PHONY: examples
+3
View File
@@ -3,6 +3,9 @@ zstd
zstd32 zstd32
zstd-compress zstd-compress
zstd-decompress zstd-decompress
zstd-frugal
zstd-small
zstd-nolegacy
# Object files # Object files
*.o *.o
+26 -12
View File
@@ -42,7 +42,7 @@ CPPFLAGS+= -I$(ZSTDDIR) -I$(ZSTDDIR)/common -I$(ZSTDDIR)/compress \
-DZSTD_NEWAPI \ -DZSTD_NEWAPI \
-DXXH_NAMESPACE=ZSTD_ # because xxhash.o already compiled with this macro from library -DXXH_NAMESPACE=ZSTD_ # because xxhash.o already compiled with this macro from library
CFLAGS ?= -O3 CFLAGS ?= -O3
DEBUGFLAGS = -Wall -Wextra -Wcast-qual -Wcast-align -Wshadow \ DEBUGFLAGS= -Wall -Wextra -Wcast-qual -Wcast-align -Wshadow \
-Wstrict-aliasing=1 -Wswitch-enum -Wdeclaration-after-statement \ -Wstrict-aliasing=1 -Wswitch-enum -Wdeclaration-after-statement \
-Wstrict-prototypes -Wundef -Wpointer-arith -Wformat-security \ -Wstrict-prototypes -Wundef -Wpointer-arith -Wformat-security \
-Wvla -Wformat=2 -Winit-self -Wfloat-equal -Wwrite-strings \ -Wvla -Wformat=2 -Winit-self -Wfloat-equal -Wwrite-strings \
@@ -132,12 +132,15 @@ else
LZ4_MSG := $(NO_LZ4_MSG) LZ4_MSG := $(NO_LZ4_MSG)
endif endif
.PHONY: default all clean clean_decomp_o install uninstall generate_res .PHONY: default
default: zstd-release default: zstd-release
.PHONY: all
all: zstd all: zstd
.PHONY: allVariants
allVariants: zstd zstd-compress zstd-decompress zstd-small zstd-nolegacy
$(ZSTDDECOMP_O): CFLAGS += $(ALIGN_LOOP) $(ZSTDDECOMP_O): CFLAGS += $(ALIGN_LOOP)
zstd zstd4 : CPPFLAGS += $(THREAD_CPP) $(ZLIBCPP) $(LZMACPP) zstd zstd4 : CPPFLAGS += $(THREAD_CPP) $(ZLIBCPP) $(LZMACPP)
@@ -154,8 +157,9 @@ zstd zstd4 : $(ZSTDLIB_FILES) zstdcli.o fileio.o bench.o datagen.o dibio.o
ifneq (,$(filter Windows%,$(OS))) ifneq (,$(filter Windows%,$(OS)))
windres/generate_res.bat windres/generate_res.bat
endif endif
$(CC) $(FLAGS) $^ $(RES_FILE) -o zstd$(EXT) $(LDFLAGS) $(CC) $(FLAGS) $^ $(RES_FILE) -o $@$(EXT) $(LDFLAGS)
.PHONY: zstd-release
zstd-release: DEBUGFLAGS := zstd-release: DEBUGFLAGS :=
zstd-release: zstd zstd-release: zstd
@@ -166,8 +170,8 @@ ifneq (,$(filter Windows%,$(OS)))
endif endif
$(CC) -m32 $(FLAGS) $^ $(RES32_FILE) -o $@$(EXT) $(CC) -m32 $(FLAGS) $^ $(RES32_FILE) -o $@$(EXT)
zstd-nolegacy : clean_decomp_o zstd-nolegacy : $(ZSTD_FILES) $(ZDICT_FILES) zstdcli.o fileio.c bench.o datagen.o dibio.o
$(MAKE) zstd ZSTD_LEGACY_SUPPORT=0 $(CC) $(FLAGS) $^ -o $@$(EXT) $(LDFLAGS)
zstd-nomt : THREAD_CPP := zstd-nomt : THREAD_CPP :=
zstd-nomt : THREAD_LD := zstd-nomt : THREAD_LD :=
@@ -198,9 +202,9 @@ zstd-pgo : clean zstd
$(MAKE) zstd MOREFLAGS=-fprofile-use $(MAKE) zstd MOREFLAGS=-fprofile-use
# minimal target, with only zstd compression and decompression. no bench. no legacy. # minimal target, with only zstd compression and decompression. no bench. no legacy.
zstd-small: CFLAGS = "-Os -s" zstd-small: CFLAGS = -Os -s
zstd-frugal zstd-small: $(ZSTD_FILES) zstdcli.c fileio.c zstd-frugal zstd-small: $(ZSTD_FILES) zstdcli.c fileio.c
$(CC) $(FLAGS) -DZSTD_NOBENCH -DZSTD_NODICT $^ -o zstd$(EXT) $(CC) $(FLAGS) -DZSTD_NOBENCH -DZSTD_NODICT $^ -o $@$(EXT)
zstd-decompress: $(ZSTDCOMMON_FILES) $(ZSTDDECOMP_FILES) zstdcli.c fileio.c zstd-decompress: $(ZSTDCOMMON_FILES) $(ZSTDDECOMP_FILES) zstdcli.c fileio.c
$(CC) $(FLAGS) -DZSTD_NOBENCH -DZSTD_NODICT -DZSTD_NOCOMPRESS $^ -o $@$(EXT) $(CC) $(FLAGS) -DZSTD_NOBENCH -DZSTD_NODICT -DZSTD_NOCOMPRESS $^ -o $@$(EXT)
@@ -208,34 +212,37 @@ zstd-decompress: $(ZSTDCOMMON_FILES) $(ZSTDDECOMP_FILES) zstdcli.c fileio.c
zstd-compress: $(ZSTDCOMMON_FILES) $(ZSTDCOMP_FILES) zstdcli.c fileio.c zstd-compress: $(ZSTDCOMMON_FILES) $(ZSTDCOMP_FILES) zstdcli.c fileio.c
$(CC) $(FLAGS) -DZSTD_NOBENCH -DZSTD_NODICT -DZSTD_NODECOMPRESS $^ -o $@$(EXT) $(CC) $(FLAGS) -DZSTD_NOBENCH -DZSTD_NODICT -DZSTD_NODECOMPRESS $^ -o $@$(EXT)
# zstd is now built with Multi-threading by default # zstd is now built with multithreading enabled y default
zstdmt: zstd zstdmt: zstd
.PHONY: generate_res
generate_res: generate_res:
windres/generate_res.bat windres/generate_res.bat
.PHONY: clean
clean: clean:
$(MAKE) -C $(ZSTDDIR) clean $(MAKE) -C $(ZSTDDIR) clean
@$(RM) $(ZSTDDIR)/decompress/*.o $(ZSTDDIR)/decompress/zstd_decompress.gcda @$(RM) $(ZSTDDIR)/decompress/*.o $(ZSTDDIR)/decompress/zstd_decompress.gcda
@$(RM) core *.o tmp* result* *.gcda dictionary *.zst \ @$(RM) core *.o tmp* result* *.gcda dictionary *.zst \
zstd$(EXT) zstd32$(EXT) zstd-compress$(EXT) zstd-decompress$(EXT) \ zstd$(EXT) zstd32$(EXT) zstd-compress$(EXT) zstd-decompress$(EXT) \
zstd-small$(EXT) zstd-frugal$(EXT) zstd-nolegacy$(EXT) \
*.gcda default.profraw have_zlib$(EXT) *.gcda default.profraw have_zlib$(EXT)
@echo Cleaning completed @echo Cleaning completed
clean_decomp_o:
@$(RM) $(ZSTDDECOMP_O)
MD2ROFF = ronn MD2ROFF = ronn
MD2ROFF_FLAGS = --roff --warnings --manual="User Commands" --organization="zstd $(ZSTD_VERSION)" MD2ROFF_FLAGS = --roff --warnings --manual="User Commands" --organization="zstd $(ZSTD_VERSION)"
zstd.1: zstd.1.md zstd.1: zstd.1.md
cat $^ | $(MD2ROFF) $(MD2ROFF_FLAGS) | sed -n '/^\.\\\".*/!p' > $@ cat $^ | $(MD2ROFF) $(MD2ROFF_FLAGS) | sed -n '/^\.\\\".*/!p' > $@
.PHONY: man
man: zstd.1 man: zstd.1
.PHONY: clean-man
clean-man: clean-man:
rm zstd.1 rm zstd.1
.PHONY: preview-man
preview-man: clean-man man preview-man: clean-man man
man ./zstd.1 man ./zstd.1
@@ -244,6 +251,10 @@ preview-man: clean-man man
#----------------------------------------------------------------------------- #-----------------------------------------------------------------------------
ifneq (,$(filter $(shell uname),Linux Darwin GNU/kFreeBSD GNU OpenBSD FreeBSD NetBSD DragonFly SunOS)) ifneq (,$(filter $(shell uname),Linux Darwin GNU/kFreeBSD GNU OpenBSD FreeBSD NetBSD DragonFly SunOS))
.PHONY: list
list:
@$(MAKE) -pRrq -f $(lastword $(MAKEFILE_LIST)) : 2>/dev/null | awk -v RS= -F: '/^# File/,/^# Finished Make data base/ {if ($$1 !~ "^[#.]") {print $$1}}' | sort | egrep -v -e '^[^[:alnum:]]' -e '^$@$$' | xargs
ifneq (,$(filter $(shell uname),SunOS)) ifneq (,$(filter $(shell uname),SunOS))
INSTALL ?= ginstall INSTALL ?= ginstall
else else
@@ -264,6 +275,7 @@ INSTALL_PROGRAM ?= $(INSTALL) -m 755
INSTALL_SCRIPT ?= $(INSTALL) -m 755 INSTALL_SCRIPT ?= $(INSTALL) -m 755
INSTALL_MAN ?= $(INSTALL) -m 644 INSTALL_MAN ?= $(INSTALL) -m 644
.PHONY: install
install: zstd install: zstd
@echo Installing binaries @echo Installing binaries
@$(INSTALL) -d -m 755 $(DESTDIR)$(BINDIR)/ $(DESTDIR)$(MANDIR)/ @$(INSTALL) -d -m 755 $(DESTDIR)$(BINDIR)/ $(DESTDIR)$(MANDIR)/
@@ -279,6 +291,7 @@ install: zstd
@ln -sf zstd.1 $(DESTDIR)$(MANDIR)/unzstd.1 @ln -sf zstd.1 $(DESTDIR)$(MANDIR)/unzstd.1
@echo zstd installation completed @echo zstd installation completed
.PHONY: uninstall
uninstall: uninstall:
@$(RM) $(DESTDIR)$(BINDIR)/zstdgrep @$(RM) $(DESTDIR)$(BINDIR)/zstdgrep
@$(RM) $(DESTDIR)$(BINDIR)/zstdless @$(RM) $(DESTDIR)$(BINDIR)/zstdless
@@ -289,4 +302,5 @@ uninstall:
@$(RM) $(DESTDIR)$(MANDIR)/unzstd.1 @$(RM) $(DESTDIR)$(MANDIR)/unzstd.1
@$(RM) $(DESTDIR)$(MANDIR)/zstd.1 @$(RM) $(DESTDIR)$(MANDIR)/zstd.1
@echo zstd programs successfully uninstalled @echo zstd programs successfully uninstalled
endif endif
+16 -16
View File
@@ -3,39 +3,38 @@ Command Line Interface for Zstandard library
Command Line Interface (CLI) can be created using the `make` command without any additional parameters. Command Line Interface (CLI) can be created using the `make` command without any additional parameters.
There are however other Makefile targets that create different variations of CLI: There are however other Makefile targets that create different variations of CLI:
- `zstd` : default CLI supporting gzip-like arguments; includes dictionary builder, benchmark, and support for decompression of legacy zstd versions - `zstd` : default CLI supporting gzip-like arguments; includes dictionary builder, benchmark, and support for decompression of legacy zstd formats
- `zstd32` : Same as `zstd`, but forced to compile in 32-bits mode - `zstd_nolegacy` : Same as `zstd` but without support for legacy zstd formats
- `zstd_nolegacy` : Same as `zstd` except of support for decompression of legacy zstd versions - `zstd-small` : CLI optimized for minimal size; no dictionary builder, no benchmark, and no support for legacy zstd formats
- `zstd-small` : CLI optimized for minimal size; without dictionary builder, benchmark, and support for decompression of legacy zstd versions - `zstd-compress` : version of CLI which can only compress into zstd format
- `zstd-compress` : compressor-only version of CLI; without dictionary builder, benchmark, and support for decompression of legacy zstd versions - `zstd-decompress` : version of CLI which can only decompress zstd format
- `zstd-decompress` : decompressor-only version of CLI; without dictionary builder, benchmark, and support for decompression of legacy zstd versions
#### Compilation variables #### Compilation variables
`zstd` tries to detect and use the following features automatically : `zstd` tries to detect and use the following features automatically :
- __HAVE_THREAD__ : multithreading is automatically enabled when `pthread` is detected. - __HAVE_THREAD__ : multithreading is automatically enabled when `pthread` is detected.
It's possible to disable multithread support, by either compiling `zstd-nomt` target or using HAVE_THREAD=0 variable. It's possible to disable multithread support, by setting HAVE_THREAD=0 .
Example : make zstd HAVE_THREAD=0 Example : make zstd HAVE_THREAD=0
It's also possible to force compilation with multithread support, using HAVE_THREAD=1. It's also possible to force compilation with multithread support, using HAVE_THREAD=1.
In which case, linking stage will fail if `pthread` library cannot be found. In which case, linking stage will fail if `pthread` library cannot be found.
This might be useful to prevent silent feature disabling. This might be useful to prevent silent feature disabling.
- __HAVE_ZLIB__ : `zstd` can compress and decompress files in `.gz` format. - __HAVE_ZLIB__ : `zstd` can compress and decompress files in `.gz` format.
This is done through command `--format=gzip`. This is ordered through command `--format=gzip`.
Alternatively, symlinks named `gzip` or `gunzip` will mimic intended behavior. Alternatively, symlinks named `gzip` or `gunzip` will mimic intended behavior.
`.gz` support is automatically enabled when `zlib` library is detected at build time. `.gz` support is automatically enabled when `zlib` library is detected at build time.
It's possible to disable `.gz` support, by either compiling `zstd-nogz` target or using HAVE_ZLIB=0 variable. It's possible to disable `.gz` support, by setting HAVE_ZLIB=0.
Example : make zstd HAVE_ZLIB=0 Example : make zstd HAVE_ZLIB=0
It's also possible to force compilation with zlib support, using HAVE_ZLIB=1. It's also possible to force compilation with zlib support, using HAVE_ZLIB=1.
In which case, linking stage will fail if `zlib` library cannot be found. In which case, linking stage will fail if `zlib` library cannot be found.
This might be useful to prevent silent feature disabling. This might be useful to prevent silent feature disabling.
- __HAVE_LZMA__ : `zstd` can compress and decompress files in `.xz` and `.lzma` formats. - __HAVE_LZMA__ : `zstd` can compress and decompress files in `.xz` and `.lzma` formats.
This is done through commands `--format=xz` and `--format=lzma` respectively. This is ordered through commands `--format=xz` and `--format=lzma` respectively.
Alternatively, symlinks named `xz`, `unxz`, `lzma`, or `unlzma` will mimic intended behavior. Alternatively, symlinks named `xz`, `unxz`, `lzma`, or `unlzma` will mimic intended behavior.
`.xz` and `.lzma` support is automatically enabled when `lzma` library is detected at build time. `.xz` and `.lzma` support is automatically enabled when `lzma` library is detected at build time.
It's possible to disable `.xz` and `.lzma` support, by either compiling `zstd-noxz` target or using HAVE_LZMA=0 variable. It's possible to disable `.xz` and `.lzma` support, by setting HAVE_LZMA=0 .
Example : make zstd HAVE_LZMA=0 Example : make zstd HAVE_LZMA=0
It's also possible to force compilation with lzma support, using HAVE_LZMA=1. It's also possible to force compilation with lzma support, using HAVE_LZMA=1.
In which case, linking stage will fail if `lzma` library cannot be found. In which case, linking stage will fail if `lzma` library cannot be found.
@@ -61,7 +60,7 @@ will rely more and more on previously decoded content to compress the rest of th
Usage of the dictionary builder and created dictionaries with CLI: Usage of the dictionary builder and created dictionaries with CLI:
1. Create the dictionary : `zstd --train FullPathToTrainingSet/* -o dictionaryName` 1. Create the dictionary : `zstd --train PathToTrainingSet/* -o dictionaryName`
2. Compress with the dictionary: `zstd FILE -D dictionaryName` 2. Compress with the dictionary: `zstd FILE -D dictionaryName`
3. Decompress with the dictionary: `zstd --decompress FILE.zst -D dictionaryName` 3. Decompress with the dictionary: `zstd --decompress FILE.zst -D dictionaryName`
@@ -70,8 +69,8 @@ Usage of the dictionary builder and created dictionaries with CLI:
CLI includes in-memory compression benchmark module for zstd. CLI includes in-memory compression benchmark module for zstd.
The benchmark is conducted using given filenames. The files are read into memory and joined together. The benchmark is conducted using given filenames. The files are read into memory and joined together.
It makes benchmark more precise as it eliminates I/O overhead. It makes benchmark more precise as it eliminates I/O overhead.
Many filenames can be supplied as multiple parameters, parameters with wildcards or Multiple filenames can be supplied, as multiple parameters, with wildcards,
names of directories can be used as parameters with the `-r` option. or names of directories can be used as parameters with `-r` option.
The benchmark measures ratio, compressed size, compression and decompression speed. The benchmark measures ratio, compressed size, compression and decompression speed.
One can select compression levels starting from `-b` and ending with `-e`. One can select compression levels starting from `-b` and ending with `-e`.
@@ -101,13 +100,14 @@ Advanced arguments :
-v : verbose mode; specify multiple times to increase verbosity -v : verbose mode; specify multiple times to increase verbosity
-q : suppress warnings; specify twice to suppress errors too -q : suppress warnings; specify twice to suppress errors too
-c : force write to standard output, even if it is the console -c : force write to standard output, even if it is the console
-l : print information about zstd compressed files
--ultra : enable levels beyond 19, up to 22 (requires more memory) --ultra : enable levels beyond 19, up to 22 (requires more memory)
-T# : use # threads for compression (default:1)
-B# : select size of each job (default:0==automatic)
--no-dictID : don't write dictID into header (dictionary compression) --no-dictID : don't write dictID into header (dictionary compression)
--[no-]check : integrity check (default:enabled) --[no-]check : integrity check (default:enabled)
-r : operate recursively on directories -r : operate recursively on directories
--format=gzip : compress files to the .gz format --format=gzip : compress files to the .gz format
--format=xz : compress files to the .xz format
--format=lzma : compress files to the .lzma format
--test : test compressed file integrity --test : test compressed file integrity
--[no-]sparse : sparse mode (default:disabled) --[no-]sparse : sparse mode (default:disabled)
-M# : Set a memory usage limit for decompression -M# : Set a memory usage limit for decompression
+229 -221
View File
@@ -925,223 +925,6 @@ int FIO_compressFilename(const char* dstFileName, const char* srcFileName,
return result; return result;
} }
typedef struct {
int numActualFrames;
int numSkippableFrames;
unsigned long long decompressedSize;
int decompUnavailable;
unsigned long long compressedSize;
int usesCheck;
} fileInfo_t;
/*
* Reads information from file, stores in *info
* if successful, returns 0, returns 1 for frame analysis error, returns 2 for file not compressed with zstd
* returns 3 for cases in which file could not be opened.
*/
static int getFileInfo(fileInfo_t* info, const char* inFileName){
int detectError = 0;
FILE* const srcFile = FIO_openSrcFile(inFileName);
if (srcFile == NULL) {
DISPLAY("Error: could not open source file %s\n", inFileName);
return 3;
}
info->compressedSize = (unsigned long long)UTIL_getFileSize(inFileName);
/* begin analyzing frame */
for ( ; ; ) {
BYTE headerBuffer[ZSTD_FRAMEHEADERSIZE_MAX];
size_t const numBytesRead = fread(headerBuffer, 1, sizeof(headerBuffer), srcFile);
if (numBytesRead < ZSTD_frameHeaderSize_min) {
if (feof(srcFile) && numBytesRead == 0 && info->compressedSize > 0) {
break;
}
else if (feof(srcFile)) {
DISPLAY("Error: reached end of file with incomplete frame\n");
detectError = 2;
break;
}
else {
DISPLAY("Error: did not reach end of file but ran out of frames\n");
detectError = 1;
break;
}
}
{ U32 const magicNumber = MEM_readLE32(headerBuffer);
/* Zstandard frame */
if (magicNumber == ZSTD_MAGICNUMBER) {
U64 const frameContentSize = ZSTD_getFrameContentSize(headerBuffer, numBytesRead);
if (frameContentSize == ZSTD_CONTENTSIZE_ERROR || frameContentSize == ZSTD_CONTENTSIZE_UNKNOWN) {
info->decompUnavailable = 1;
} else {
info->decompressedSize += frameContentSize;
}
/* move to the end of the frame header */
{ size_t const headerSize = ZSTD_frameHeaderSize(headerBuffer, numBytesRead);
if (ZSTD_isError(headerSize)) {
DISPLAY("Error: could not determine frame header size\n");
detectError = 1;
break;
}
{ int const ret = fseek(srcFile, ((long)headerSize)-((long)numBytesRead), SEEK_CUR);
if (ret != 0) {
DISPLAY("Error: could not move to end of frame header\n");
detectError = 1;
break;
} } }
/* skip the rest of the blocks in the frame */
{ int lastBlock = 0;
do {
BYTE blockHeaderBuffer[3];
size_t const readBytes = fread(blockHeaderBuffer, 1, 3, srcFile);
if (readBytes != 3) {
DISPLAY("There was a problem reading the block header\n");
detectError = 1;
break;
}
{ U32 const blockHeader = MEM_readLE24(blockHeaderBuffer);
U32 const blockTypeID = (blockHeader >> 1) & 3;
U32 const isRLE = (blockTypeID == 1);
U32 const isWrongBlock = (blockTypeID == 3);
long const blockSize = isRLE ? 1 : (long)(blockHeader >> 3);
if (isWrongBlock) {
DISPLAY("Error: unsupported block type \n");
detectError = 1;
break;
}
lastBlock = blockHeader & 1;
{ int const ret = fseek(srcFile, blockSize, SEEK_CUR);
if (ret != 0) {
DISPLAY("Error: could not skip to end of block\n");
detectError = 1;
break;
} } }
} while (lastBlock != 1);
if (detectError) break;
}
/* check if checksum is used */
{ BYTE const frameHeaderDescriptor = headerBuffer[4];
int const contentChecksumFlag = (frameHeaderDescriptor & (1 << 2)) >> 2;
if (contentChecksumFlag) {
int const ret = fseek(srcFile, 4, SEEK_CUR);
info->usesCheck = 1;
if (ret != 0) {
DISPLAY("Error: could not skip past checksum\n");
detectError = 1;
break;
} } }
info->numActualFrames++;
}
/* Skippable frame */
else if ((magicNumber & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START) {
U32 const frameSize = MEM_readLE32(headerBuffer + 4);
long const seek = (long)(8 + frameSize - numBytesRead);
int const ret = LONG_SEEK(srcFile, seek, SEEK_CUR);
if (ret != 0) {
DISPLAY("Error: could not find end of skippable frame\n");
detectError = 1;
break;
}
info->numSkippableFrames++;
}
/* unknown content */
else {
detectError = 2;
break;
}
}
} /* end analyzing frame */
fclose(srcFile);
return detectError;
}
static void displayInfo(const char* inFileName, fileInfo_t* info, int displayLevel){
unsigned const unit = info->compressedSize < (1 MB) ? (1 KB) : (1 MB);
const char* const unitStr = info->compressedSize < (1 MB) ? "KB" : "MB";
double const compressedSizeUnit = (double)info->compressedSize / unit;
double const decompressedSizeUnit = (double)info->decompressedSize / unit;
double const ratio = (info->compressedSize == 0) ? 0 : ((double)info->decompressedSize)/info->compressedSize;
const char* const checkString = (info->usesCheck ? "XXH64" : "None");
if (displayLevel <= 2) {
if (!info->decompUnavailable) {
DISPLAYOUT("Skippable Non-Skippable Compressed Uncompressed Ratio Check Filename\n");
DISPLAYOUT("%9d %13d %7.2f %2s %9.2f %2s %5.3f %5s %s\n",
info->numSkippableFrames, info->numActualFrames,
compressedSizeUnit, unitStr, decompressedSizeUnit, unitStr,
ratio, checkString, inFileName);
} else {
DISPLAYOUT("Skippable Non-Skippable Compressed Check Filename\n");
DISPLAYOUT("%9d %13d %7.2f MB %5s %s\n",
info->numSkippableFrames, info->numActualFrames,
compressedSizeUnit, checkString, inFileName);
}
} else {
DISPLAYOUT("# Zstandard Frames: %d\n", info->numActualFrames);
DISPLAYOUT("# Skippable Frames: %d\n", info->numSkippableFrames);
DISPLAYOUT("Compressed Size: %.2f %2s (%llu B)\n",
compressedSizeUnit, unitStr, info->compressedSize);
if (!info->decompUnavailable) {
DISPLAYOUT("Decompressed Size: %.2f %2s (%llu B)\n",
decompressedSizeUnit, unitStr, info->decompressedSize);
DISPLAYOUT("Ratio: %.4f\n", ratio);
}
DISPLAYOUT("Check: %s\n", checkString);
DISPLAYOUT("\n");
}
}
static int FIO_listFile(const char* inFileName, int displayLevel, unsigned fileNo, unsigned numFiles){
/* initialize info to avoid warnings */
fileInfo_t info;
memset(&info, 0, sizeof(info));
DISPLAYOUT("%s (%u/%u):\n", inFileName, fileNo, numFiles);
{
int const error = getFileInfo(&info, inFileName);
if (error == 1) {
/* display error, but provide output */
DISPLAY("An error occurred with getting file info\n");
}
else if (error == 2) {
DISPLAYOUT("File %s not compressed with zstd\n", inFileName);
if (displayLevel > 2) {
DISPLAYOUT("\n");
}
return 1;
}
else if (error == 3) {
/* error occurred with opening the file */
if (displayLevel > 2) {
DISPLAYOUT("\n");
}
return 1;
}
displayInfo(inFileName, &info, displayLevel);
return error;
}
}
int FIO_listMultipleFiles(unsigned numFiles, const char** filenameTable, int displayLevel){
if (numFiles == 0) {
DISPLAYOUT("No files given\n");
return 0;
}
DISPLAYOUT("===========================================\n");
DISPLAYOUT("Printing information about compressed files\n");
DISPLAYOUT("===========================================\n");
DISPLAYOUT("Number of files listed: %u\n", numFiles);
{
int error = 0;
unsigned u;
for (u=0; u<numFiles;u++) {
error |= FIO_listFile(filenameTable[u], displayLevel, u+1, numFiles);
}
return error;
}
}
int FIO_compressMultipleFilenames(const char** inFileNamesTable, unsigned nbFiles, int FIO_compressMultipleFilenames(const char** inFileNamesTable, unsigned nbFiles,
const char* suffix, const char* suffix,
@@ -1187,10 +970,8 @@ int FIO_compressMultipleFilenames(const char** inFileNamesTable, unsigned nbFile
missed_files += FIO_compressFilename_dstFile(ress, dstFileName, inFileNamesTable[u], compressionLevel); missed_files += FIO_compressFilename_dstFile(ress, dstFileName, inFileNamesTable[u], compressionLevel);
} } } }
/* Close & Free */
FIO_freeCResources(ress); FIO_freeCResources(ress);
free(dstFileName); free(dstFileName);
return missed_files; return missed_files;
} }
@@ -1201,8 +982,8 @@ int FIO_compressMultipleFilenames(const char** inFileNamesTable, unsigned nbFile
#ifndef ZSTD_NODECOMPRESS #ifndef ZSTD_NODECOMPRESS
/* ************************************************************************** /* **************************************************************************
* Decompression * Decompression
****************************************************************************/ ***************************************************************************/
typedef struct { typedef struct {
void* srcBuffer; void* srcBuffer;
size_t srcBufferLoaded; size_t srcBufferLoaded;
@@ -1871,4 +1652,231 @@ int FIO_decompressMultipleFilenames(const char** srcNamesTable, unsigned nbFiles
return missingFiles + skippedFiles; return missingFiles + skippedFiles;
} }
/* **************************************************************************
* .zst file info (--list command)
***************************************************************************/
typedef struct {
int numActualFrames;
int numSkippableFrames;
unsigned long long decompressedSize;
int decompUnavailable;
unsigned long long compressedSize;
int usesCheck;
} fileInfo_t;
/** getFileInfo() :
* Reads information from file, stores in *info
* @return : 0 if successful
* 1 for frame analysis error
* 2 for file not compressed with zstd
* 3 for cases in which file could not be opened.
*/
static int getFileInfo(fileInfo_t* info, const char* inFileName){
int detectError = 0;
FILE* const srcFile = FIO_openSrcFile(inFileName);
if (srcFile == NULL) {
DISPLAY("Error: could not open source file %s\n", inFileName);
return 3;
}
info->compressedSize = (unsigned long long)UTIL_getFileSize(inFileName);
/* begin analyzing frame */
for ( ; ; ) {
BYTE headerBuffer[ZSTD_FRAMEHEADERSIZE_MAX];
size_t const numBytesRead = fread(headerBuffer, 1, sizeof(headerBuffer), srcFile);
if (numBytesRead < ZSTD_frameHeaderSize_min) {
if (feof(srcFile) && numBytesRead == 0 && info->compressedSize > 0) {
break;
}
else if (feof(srcFile)) {
DISPLAY("Error: reached end of file with incomplete frame\n");
detectError = 2;
break;
}
else {
DISPLAY("Error: did not reach end of file but ran out of frames\n");
detectError = 1;
break;
}
}
{ U32 const magicNumber = MEM_readLE32(headerBuffer);
/* Zstandard frame */
if (magicNumber == ZSTD_MAGICNUMBER) {
U64 const frameContentSize = ZSTD_getFrameContentSize(headerBuffer, numBytesRead);
if (frameContentSize == ZSTD_CONTENTSIZE_ERROR || frameContentSize == ZSTD_CONTENTSIZE_UNKNOWN) {
info->decompUnavailable = 1;
} else {
info->decompressedSize += frameContentSize;
}
/* move to the end of the frame header */
{ size_t const headerSize = ZSTD_frameHeaderSize(headerBuffer, numBytesRead);
if (ZSTD_isError(headerSize)) {
DISPLAY("Error: could not determine frame header size\n");
detectError = 1;
break;
}
{ int const ret = fseek(srcFile, ((long)headerSize)-((long)numBytesRead), SEEK_CUR);
if (ret != 0) {
DISPLAY("Error: could not move to end of frame header\n");
detectError = 1;
break;
} } }
/* skip the rest of the blocks in the frame */
{ int lastBlock = 0;
do {
BYTE blockHeaderBuffer[3];
size_t const readBytes = fread(blockHeaderBuffer, 1, 3, srcFile);
if (readBytes != 3) {
DISPLAY("There was a problem reading the block header\n");
detectError = 1;
break;
}
{ U32 const blockHeader = MEM_readLE24(blockHeaderBuffer);
U32 const blockTypeID = (blockHeader >> 1) & 3;
U32 const isRLE = (blockTypeID == 1);
U32 const isWrongBlock = (blockTypeID == 3);
long const blockSize = isRLE ? 1 : (long)(blockHeader >> 3);
if (isWrongBlock) {
DISPLAY("Error: unsupported block type \n");
detectError = 1;
break;
}
lastBlock = blockHeader & 1;
{ int const ret = fseek(srcFile, blockSize, SEEK_CUR);
if (ret != 0) {
DISPLAY("Error: could not skip to end of block\n");
detectError = 1;
break;
} } }
} while (lastBlock != 1);
if (detectError) break;
}
/* check if checksum is used */
{ BYTE const frameHeaderDescriptor = headerBuffer[4];
int const contentChecksumFlag = (frameHeaderDescriptor & (1 << 2)) >> 2;
if (contentChecksumFlag) {
int const ret = fseek(srcFile, 4, SEEK_CUR);
info->usesCheck = 1;
if (ret != 0) {
DISPLAY("Error: could not skip past checksum\n");
detectError = 1;
break;
} } }
info->numActualFrames++;
}
/* Skippable frame */
else if ((magicNumber & 0xFFFFFFF0U) == ZSTD_MAGIC_SKIPPABLE_START) {
U32 const frameSize = MEM_readLE32(headerBuffer + 4);
long const seek = (long)(8 + frameSize - numBytesRead);
int const ret = LONG_SEEK(srcFile, seek, SEEK_CUR);
if (ret != 0) {
DISPLAY("Error: could not find end of skippable frame\n");
detectError = 1;
break;
}
info->numSkippableFrames++;
}
/* unknown content */
else {
detectError = 2;
break;
}
}
} /* end analyzing frame */
fclose(srcFile);
return detectError;
}
static void displayInfo(const char* inFileName, fileInfo_t* info, int displayLevel){
unsigned const unit = info->compressedSize < (1 MB) ? (1 KB) : (1 MB);
const char* const unitStr = info->compressedSize < (1 MB) ? "KB" : "MB";
double const compressedSizeUnit = (double)info->compressedSize / unit;
double const decompressedSizeUnit = (double)info->decompressedSize / unit;
double const ratio = (info->compressedSize == 0) ? 0 : ((double)info->decompressedSize)/info->compressedSize;
const char* const checkString = (info->usesCheck ? "XXH64" : "None");
if (displayLevel <= 2) {
if (!info->decompUnavailable) {
DISPLAYOUT("Skippable Non-Skippable Compressed Uncompressed Ratio Check Filename\n");
DISPLAYOUT("%9d %13d %7.2f %2s %9.2f %2s %5.3f %5s %s\n",
info->numSkippableFrames, info->numActualFrames,
compressedSizeUnit, unitStr, decompressedSizeUnit, unitStr,
ratio, checkString, inFileName);
} else {
DISPLAYOUT("Skippable Non-Skippable Compressed Check Filename\n");
DISPLAYOUT("%9d %13d %7.2f MB %5s %s\n",
info->numSkippableFrames, info->numActualFrames,
compressedSizeUnit, checkString, inFileName);
}
} else {
DISPLAYOUT("# Zstandard Frames: %d\n", info->numActualFrames);
DISPLAYOUT("# Skippable Frames: %d\n", info->numSkippableFrames);
DISPLAYOUT("Compressed Size: %.2f %2s (%llu B)\n",
compressedSizeUnit, unitStr, info->compressedSize);
if (!info->decompUnavailable) {
DISPLAYOUT("Decompressed Size: %.2f %2s (%llu B)\n",
decompressedSizeUnit, unitStr, info->decompressedSize);
DISPLAYOUT("Ratio: %.4f\n", ratio);
}
DISPLAYOUT("Check: %s\n", checkString);
DISPLAYOUT("\n");
}
}
static int FIO_listFile(const char* inFileName, int displayLevel, unsigned fileNo, unsigned numFiles){
/* initialize info to avoid warnings */
fileInfo_t info;
memset(&info, 0, sizeof(info));
DISPLAYOUT("%s (%u/%u):\n", inFileName, fileNo, numFiles);
{
int const error = getFileInfo(&info, inFileName);
if (error == 1) {
/* display error, but provide output */
DISPLAY("An error occurred with getting file info\n");
}
else if (error == 2) {
DISPLAYOUT("File %s not compressed with zstd\n", inFileName);
if (displayLevel > 2) {
DISPLAYOUT("\n");
}
return 1;
}
else if (error == 3) {
/* error occurred with opening the file */
if (displayLevel > 2) {
DISPLAYOUT("\n");
}
return 1;
}
displayInfo(inFileName, &info, displayLevel);
return error;
}
}
int FIO_listMultipleFiles(unsigned numFiles, const char** filenameTable, int displayLevel){
if (numFiles == 0) {
DISPLAYOUT("No files given\n");
return 0;
}
DISPLAYOUT("===========================================\n");
DISPLAYOUT("Printing information about compressed files\n");
DISPLAYOUT("===========================================\n");
DISPLAYOUT("Number of files listed: %u\n", numFiles);
{
int error = 0;
unsigned u;
for (u=0; u<numFiles;u++) {
error |= FIO_listFile(filenameTable[u], displayLevel, u+1, numFiles);
}
return error;
}
}
#endif /* #ifndef ZSTD_NODECOMPRESS */ #endif /* #ifndef ZSTD_NODECOMPRESS */
+15 -4
View File
@@ -634,7 +634,10 @@ int main(int argCount, const char* argv[])
filenameTable[filenameIdx++] = argument; filenameTable[filenameIdx++] = argument;
} }
if (lastCommand) { DISPLAY("error : command must be followed by argument \n"); CLEAN_RETURN(1); } /* forgotten argument */ if (lastCommand) { /* forgotten argument */
DISPLAY("error : command must be followed by argument \n");
CLEAN_RETURN(1);
}
/* Welcome message (if verbose) */ /* Welcome message (if verbose) */
DISPLAYLEVEL(3, WELCOME_MESSAGE); DISPLAYLEVEL(3, WELCOME_MESSAGE);
@@ -642,7 +645,7 @@ int main(int argCount, const char* argv[])
DISPLAYLEVEL(4, "_POSIX_C_SOURCE defined: %ldL\n", (long) _POSIX_C_SOURCE); DISPLAYLEVEL(4, "_POSIX_C_SOURCE defined: %ldL\n", (long) _POSIX_C_SOURCE);
#endif #endif
#ifdef _POSIX_VERSION #ifdef _POSIX_VERSION
DISPLAYLEVEL(4, "_POSIX_VERSION defined: %ldL\n", (long) _POSIX_VERSION); DISPLAYLEVEL(4, "_POSIX_VERSION defined: %ldL \n", (long) _POSIX_VERSION);
#endif #endif
#ifdef PLATFORM_POSIX_VERSION #ifdef PLATFORM_POSIX_VERSION
DISPLAYLEVEL(4, "PLATFORM_POSIX_VERSION defined: %ldL\n", (long) PLATFORM_POSIX_VERSION); DISPLAYLEVEL(4, "PLATFORM_POSIX_VERSION defined: %ldL\n", (long) PLATFORM_POSIX_VERSION);
@@ -651,7 +654,7 @@ int main(int argCount, const char* argv[])
if (nbThreads == 0) { if (nbThreads == 0) {
/* try to guess */ /* try to guess */
nbThreads = UTIL_countPhysicalCores(); nbThreads = UTIL_countPhysicalCores();
DISPLAYLEVEL(3, "Note: %d physical core(s) detected\n", nbThreads); DISPLAYLEVEL(3, "Note: %d physical core(s) detected \n", nbThreads);
} }
g_utilDisplayLevel = g_displayLevel; g_utilDisplayLevel = g_displayLevel;
@@ -678,10 +681,17 @@ int main(int argCount, const char* argv[])
} }
} }
#endif #endif
if (operation == zom_list) { if (operation == zom_list) {
#ifndef ZSTD_NODECOMPRESS
int const ret = FIO_listMultipleFiles(filenameIdx, filenameTable, g_displayLevel); int const ret = FIO_listMultipleFiles(filenameIdx, filenameTable, g_displayLevel);
CLEAN_RETURN(ret); CLEAN_RETURN(ret);
#else
DISPLAY("file information is not supported \n");
CLEAN_RETURN(1);
#endif
} }
/* Check if benchmark is selected */ /* Check if benchmark is selected */
if (operation==zom_bench) { if (operation==zom_bench) {
#ifndef ZSTD_NOBENCH #ifndef ZSTD_NOBENCH
@@ -691,7 +701,7 @@ int main(int argCount, const char* argv[])
BMK_setNbSeconds(bench_nbSeconds); BMK_setNbSeconds(bench_nbSeconds);
BMK_benchFiles(filenameTable, filenameIdx, dictFileName, cLevel, cLevelLast, &compressionParams, setRealTimePrio); BMK_benchFiles(filenameTable, filenameIdx, dictFileName, cLevel, cLevelLast, &compressionParams, setRealTimePrio);
#endif #endif
(void)bench_nbSeconds; (void)bench_nbSeconds; (void)blockSize; (void)setRealTimePrio;
goto _end; goto _end;
} }
@@ -762,6 +772,7 @@ int main(int argCount, const char* argv[])
else else
operationResult = FIO_compressMultipleFilenames(filenameTable, filenameIdx, outFileName ? outFileName : suffix, dictFileName, cLevel, &compressionParams); operationResult = FIO_compressMultipleFilenames(filenameTable, filenameIdx, outFileName ? outFileName : suffix, dictFileName, cLevel, &compressionParams);
#else #else
(void)suffix;
DISPLAY("Compression not supported\n"); DISPLAY("Compression not supported\n");
#endif #endif
} else { /* decompression or test */ } else { /* decompression or test */