mirror of
https://github.com/exoticorn/upkr.git
synced 2026-01-20 19:46:42 +01:00
Compare commits
1 Commits
fuzz
...
parity-con
| Author | SHA1 | Date | |
|---|---|---|---|
| 7d40bb8123 |
1
asm_unpackers/.gitignore
vendored
1
asm_unpackers/.gitignore
vendored
@@ -1 +0,0 @@
|
|||||||
/build/
|
|
||||||
@@ -1,30 +0,0 @@
|
|||||||
build/unpack_riscv64: ../c_unpacker/main.c ../c_unpacker/unpack.c
|
|
||||||
mkdir -p build
|
|
||||||
riscv64-linux-gnu-gcc -g -static -o $@ $^
|
|
||||||
|
|
||||||
test_riscv64: build/unpack_riscv64
|
|
||||||
qemu-riscv64 $< test_data.upk /tmp/out.bin
|
|
||||||
cmp test_data.bin /tmp/out.bin
|
|
||||||
|
|
||||||
build/unpack_armv6m: ../c_unpacker/main.c unpack_armv6m.S
|
|
||||||
mkdir -p build
|
|
||||||
arm-linux-gnueabihf-gcc -g -static -o $@ $^
|
|
||||||
|
|
||||||
test_armv6m: build/unpack_armv6m
|
|
||||||
qemu-arm $< test_data.upk /tmp/out.bin
|
|
||||||
cmp test_data.bin /tmp/out.bin
|
|
||||||
|
|
||||||
build/unpack_armv6m.bin: unpack_armv6m.S
|
|
||||||
arm-none-eabi-gcc -march=armv6-m -c -o build/unpack_armv6m.o $?
|
|
||||||
arm-none-eabi-objcopy -O binary --only-section=.text build/unpack_armv6m.o $@
|
|
||||||
|
|
||||||
build/unpack_c: ../c_unpacker/main.c ../c_unpacker/unpack.c
|
|
||||||
mkdir -p build
|
|
||||||
gcc -g -o $@ $^
|
|
||||||
|
|
||||||
test_c: build/unpack_c
|
|
||||||
$< test_data.upk /tmp/out.bin
|
|
||||||
cmp test_data.bin /tmp/out.bin
|
|
||||||
|
|
||||||
sizes: build/unpack_armv6m.bin
|
|
||||||
ls -l build/*.bin
|
|
||||||
@@ -1,99 +0,0 @@
|
|||||||
typedef unsigned char u8;
|
|
||||||
typedef unsigned short u16;
|
|
||||||
typedef unsigned long u32;
|
|
||||||
|
|
||||||
u8* upkr_data_ptr;
|
|
||||||
u8 upkr_probs[1 + 255 + 1 + 2*32 + 2*32];
|
|
||||||
#ifdef UPKR_BITSTREAM
|
|
||||||
u16 upkr_state;
|
|
||||||
u8 upkr_current_byte;
|
|
||||||
int upkr_bits_left;
|
|
||||||
#else
|
|
||||||
u32 upkr_state;
|
|
||||||
#endif
|
|
||||||
|
|
||||||
int upkr_decode_bit(int context_index) {
|
|
||||||
#ifdef UPKR_BITSTREAM
|
|
||||||
while(upkr_state < 32768) {
|
|
||||||
if(upkr_bits_left == 0) {
|
|
||||||
upkr_current_byte = *upkr_data_ptr++;
|
|
||||||
upkr_bits_left = 8;
|
|
||||||
}
|
|
||||||
upkr_state = (upkr_state << 1) + (upkr_current_byte & 1);
|
|
||||||
upkr_current_byte >>= 1;
|
|
||||||
--upkr_bits_left;
|
|
||||||
}
|
|
||||||
#else
|
|
||||||
while(upkr_state < 4096) {
|
|
||||||
upkr_state = (upkr_state << 8) | *upkr_data_ptr++;
|
|
||||||
}
|
|
||||||
#endif
|
|
||||||
|
|
||||||
int prob = upkr_probs[context_index];
|
|
||||||
int bit = (upkr_state & 255) < prob ? 1 : 0;
|
|
||||||
|
|
||||||
int tmp = prob;
|
|
||||||
if(!bit) {
|
|
||||||
tmp = 256 - tmp;
|
|
||||||
}
|
|
||||||
upkr_state = tmp * (upkr_state >> 8) + (upkr_state & 255);
|
|
||||||
tmp += (256 - tmp + 8) >> 4;
|
|
||||||
if(!bit) {
|
|
||||||
upkr_state -= prob;
|
|
||||||
tmp = 256 - tmp;
|
|
||||||
}
|
|
||||||
upkr_probs[context_index] = tmp;
|
|
||||||
|
|
||||||
return bit;
|
|
||||||
}
|
|
||||||
|
|
||||||
int upkr_decode_length(int context_index) {
|
|
||||||
int length = 0;
|
|
||||||
int bit_pos = 0;
|
|
||||||
while(upkr_decode_bit(context_index)) {
|
|
||||||
length |= upkr_decode_bit(context_index + 1) << bit_pos++;
|
|
||||||
context_index += 2;
|
|
||||||
}
|
|
||||||
return length | (1 << bit_pos);
|
|
||||||
}
|
|
||||||
|
|
||||||
void* upkr_unpack(void* destination, void* compressed_data) {
|
|
||||||
upkr_data_ptr = (u8*)compressed_data;
|
|
||||||
upkr_state = 0;
|
|
||||||
#ifdef UPKR_BITSTREAM
|
|
||||||
upkr_bits_left = 0;
|
|
||||||
#endif
|
|
||||||
for(int i = 0; i < sizeof(upkr_probs); ++i)
|
|
||||||
upkr_probs[i] = 128;
|
|
||||||
|
|
||||||
u8* write_ptr = (u8*)destination;
|
|
||||||
|
|
||||||
int prev_was_match = 0;
|
|
||||||
int offset = 0;
|
|
||||||
for(;;) {
|
|
||||||
if(upkr_decode_bit(0)) {
|
|
||||||
if(prev_was_match || upkr_decode_bit(256)) {
|
|
||||||
offset = upkr_decode_length(257) - 1;
|
|
||||||
if(offset == 0) {
|
|
||||||
break;
|
|
||||||
}
|
|
||||||
}
|
|
||||||
int length = upkr_decode_length(257 + 64);
|
|
||||||
while(length--) {
|
|
||||||
*write_ptr = write_ptr[-offset];
|
|
||||||
++write_ptr;
|
|
||||||
}
|
|
||||||
prev_was_match = 1;
|
|
||||||
} else {
|
|
||||||
int byte = 1;
|
|
||||||
while(byte < 256) {
|
|
||||||
int bit = upkr_decode_bit(byte);
|
|
||||||
byte = (byte << 1) + bit;
|
|
||||||
}
|
|
||||||
*write_ptr++ = byte;
|
|
||||||
prev_was_match = 0;
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
return write_ptr;
|
|
||||||
}
|
|
||||||
Binary file not shown.
@@ -1,162 +0,0 @@
|
|||||||
// armv6-m upkr unpacker by yrlf
|
|
||||||
// some optimizations by exoticorn
|
|
||||||
|
|
||||||
.syntax unified
|
|
||||||
.thumb
|
|
||||||
|
|
||||||
.section .text
|
|
||||||
|
|
||||||
#define ALIGNUP(n, align) (((n) + (align) - 1) & ~((align) - 1))
|
|
||||||
#define PROB_LEN (1 + 255 + 1 + 2*32 + 2*32)
|
|
||||||
#define FRAME_SIZE ALIGNUP(PROB_LEN, 4)
|
|
||||||
|
|
||||||
// auto upkr_unpack(uint8_t * out, uint8_t * in) -> tuple<uint8_t *, uint8_t *>
|
|
||||||
.global upkr_unpack
|
|
||||||
.type upkr_unpack, %function
|
|
||||||
// r0 .. out_ptr (returned)
|
|
||||||
// r1 .. in_ptr (returned)
|
|
||||||
// r2 .. state
|
|
||||||
// r3 .. offset
|
|
||||||
// r4 .. prev_was_literal / decode_length ret
|
|
||||||
// r5 .. subroutine arg (preserved)
|
|
||||||
// r6 .. decode_bit ret
|
|
||||||
// r7 .. probs ptr
|
|
||||||
upkr_unpack:
|
|
||||||
push { r4, r5, r6, r7, lr }
|
|
||||||
sub sp, sp, #FRAME_SIZE
|
|
||||||
|
|
||||||
mov r7, sp
|
|
||||||
movs r2, #255
|
|
||||||
adds r2, r2, #(PROB_LEN - 255)
|
|
||||||
movs r3, #128
|
|
||||||
.Lclear:
|
|
||||||
subs r2, r2, #1
|
|
||||||
strb r3, [r7, r2]
|
|
||||||
bne .Lclear
|
|
||||||
|
|
||||||
.Lloop:
|
|
||||||
movs r5, #0
|
|
||||||
bl upkr_decode_bit
|
|
||||||
beq .Ldata
|
|
||||||
.Lmatch:
|
|
||||||
// r6 = 1
|
|
||||||
lsls r5, r6, #8
|
|
||||||
cmp r4, #0
|
|
||||||
beq 1f
|
|
||||||
|
|
||||||
bl upkr_decode_bit
|
|
||||||
beq 2f
|
|
||||||
|
|
||||||
1:
|
|
||||||
bl upkr_decode_length
|
|
||||||
adds r3, r4, #1
|
|
||||||
beq .Lend
|
|
||||||
2:
|
|
||||||
|
|
||||||
adds r5, r5, #64
|
|
||||||
bl upkr_decode_length
|
|
||||||
.Lcopy_loop:
|
|
||||||
ldrb r5, [r0, r3]
|
|
||||||
.Lstore:
|
|
||||||
strb r5, [r0]
|
|
||||||
adds r0, r0, #1
|
|
||||||
adds r4, r4, #1
|
|
||||||
blt .Lcopy_loop
|
|
||||||
b .Lloop
|
|
||||||
|
|
||||||
.Ldata:
|
|
||||||
movs r5, #1
|
|
||||||
|
|
||||||
.Ldata_loop:
|
|
||||||
bl upkr_decode_bit
|
|
||||||
adcs r5, r5, r5
|
|
||||||
lsrs r4, r5, #8
|
|
||||||
beq .Ldata_loop
|
|
||||||
b .Lstore
|
|
||||||
|
|
||||||
.Lend:
|
|
||||||
add sp, sp, #FRAME_SIZE
|
|
||||||
pop { r4, r5, r6, r7, pc }
|
|
||||||
|
|
||||||
.type upkr_decode_length, %function
|
|
||||||
// r0 .. -length tmp (saved)
|
|
||||||
// r1 ..
|
|
||||||
// r2 ..
|
|
||||||
// r3 ..
|
|
||||||
// r4 .. -length (returned)
|
|
||||||
// r5 .. context index (saved)
|
|
||||||
// r6 .. (saved)
|
|
||||||
// r7 ..
|
|
||||||
upkr_decode_length:
|
|
||||||
push { r0, r5, r6, lr }
|
|
||||||
|
|
||||||
movs r0, #0
|
|
||||||
subs r4, r0, #1
|
|
||||||
.Lbit_loop:
|
|
||||||
adds r5, r5, #1
|
|
||||||
bl upkr_decode_bit
|
|
||||||
beq 1f
|
|
||||||
|
|
||||||
adds r5, r5, #1
|
|
||||||
bl upkr_decode_bit
|
|
||||||
beq 2f
|
|
||||||
adds r0, r0, r4
|
|
||||||
2:
|
|
||||||
lsls r4, r4, #1
|
|
||||||
b .Lbit_loop
|
|
||||||
1:
|
|
||||||
adds r4, r4, r0
|
|
||||||
|
|
||||||
pop { r0, r5, r6, pc }
|
|
||||||
|
|
||||||
.type upkr_decode_bit, %function
|
|
||||||
// r0 .. tmp / prob (saved)
|
|
||||||
// r1 .. in_ptr (modified)
|
|
||||||
// r2 .. state (modified)
|
|
||||||
// r3 .. scratch (saved)
|
|
||||||
// r4 ..
|
|
||||||
// r5 .. context index (preserved)
|
|
||||||
// r6 .. bit (returned)
|
|
||||||
// r7 .. probs ptr (preserved)
|
|
||||||
upkr_fill_state:
|
|
||||||
lsls r2, r2, #8
|
|
||||||
ldrb r6, [r1]
|
|
||||||
adds r1, r1, #1
|
|
||||||
orrs r2, r2, r6
|
|
||||||
|
|
||||||
upkr_decode_bit:
|
|
||||||
lsrs r6, r2, #12
|
|
||||||
beq upkr_fill_state
|
|
||||||
|
|
||||||
push { r0, r1, r3, lr }
|
|
||||||
|
|
||||||
ldrb r0, [r7, r5]
|
|
||||||
|
|
||||||
lsrs r3, r2, #8
|
|
||||||
uxtb r1, r2
|
|
||||||
|
|
||||||
subs r6, r1, r0
|
|
||||||
blt 1f
|
|
||||||
|
|
||||||
subs r1, r2, r0
|
|
||||||
rsbs r0, r0, #0
|
|
||||||
1:
|
|
||||||
|
|
||||||
muls r3, r3, r0
|
|
||||||
adds r2, r1, r3
|
|
||||||
|
|
||||||
rsbs r3, r0, #0
|
|
||||||
uxtb r3, r3
|
|
||||||
lsrs r3, r3, #4
|
|
||||||
adcs r0, r0, r3
|
|
||||||
|
|
||||||
cmp r6, #0
|
|
||||||
blt 1f
|
|
||||||
|
|
||||||
rsbs r0, r0, #0
|
|
||||||
1:
|
|
||||||
|
|
||||||
strb r0, [r7, r5]
|
|
||||||
|
|
||||||
lsrs r6, r6, #31
|
|
||||||
pop { r0, r1, r3, pc }
|
|
||||||
5
c_unpacker/.gitignore
vendored
5
c_unpacker/.gitignore
vendored
@@ -1,5 +0,0 @@
|
|||||||
unpack
|
|
||||||
unpack_bitstream
|
|
||||||
unpack_debug
|
|
||||||
*.upk
|
|
||||||
|
|
||||||
@@ -1,10 +0,0 @@
|
|||||||
all: unpack unpack_bitstream
|
|
||||||
|
|
||||||
unpack: main.c unpack.c
|
|
||||||
cc -O2 -o unpack main.c unpack.c
|
|
||||||
|
|
||||||
unpack_bitstream: main.c unpack.c
|
|
||||||
cc -O2 -D UPKR_BITSTREAM -o unpack_bitstream main.c unpack.c
|
|
||||||
|
|
||||||
unpack_debug: main.c unpack.c
|
|
||||||
cc -g -o unpack_debug main.c unpack.c
|
|
||||||
@@ -1,33 +0,0 @@
|
|||||||
int upkr_decode_bit(int context_index) {
|
|
||||||
#ifdef UPKR_BITSTREAM
|
|
||||||
while(upkr_state < 32768) {
|
|
||||||
if(upkr_bits_left == 0) {
|
|
||||||
upkr_current_byte = *upkr_data_ptr++;
|
|
||||||
upkr_bits_left = 8;
|
|
||||||
}
|
|
||||||
upkr_state = (upkr_state << 1) + (upkr_current_byte & 1);
|
|
||||||
upkr_current_byte >>= 1;
|
|
||||||
--upkr_bits_left;
|
|
||||||
}
|
|
||||||
#else
|
|
||||||
while(upkr_state < 4096) {
|
|
||||||
upkr_state = (upkr_state << 8) | *upkr_data_ptr++;
|
|
||||||
}
|
|
||||||
#endif
|
|
||||||
|
|
||||||
int prob = upkr_probs[context_index];
|
|
||||||
int bit = (upkr_state & 255) < prob ? 1 : 0;
|
|
||||||
|
|
||||||
if(bit) {
|
|
||||||
prob = 256 - prob;
|
|
||||||
}
|
|
||||||
upkr_state -= prob * ((upkr_state >> 8) + (bit ^ 1));
|
|
||||||
prob -= (prob + 8) >> 4;
|
|
||||||
if(bit) {
|
|
||||||
prob = -prob;
|
|
||||||
}
|
|
||||||
upkr_probs[context_index] = prob;
|
|
||||||
|
|
||||||
return bit;
|
|
||||||
}
|
|
||||||
|
|
||||||
@@ -1,26 +0,0 @@
|
|||||||
#include <stdio.h>
|
|
||||||
#include <stdlib.h>
|
|
||||||
|
|
||||||
void* upkr_unpack(void* destination, void* compressed_data);
|
|
||||||
|
|
||||||
int main(int argn, char** argv) {
|
|
||||||
void* input_buffer = malloc(1024*1024);
|
|
||||||
void* output_buffer = malloc(1024*1024);
|
|
||||||
|
|
||||||
FILE* in_file = fopen(argv[1], "rb");
|
|
||||||
int in_size = fread(input_buffer, 1, 1024*1024, in_file);
|
|
||||||
fclose(in_file);
|
|
||||||
|
|
||||||
printf("Compressed size: %d\n", in_size);
|
|
||||||
|
|
||||||
void* end_ptr = upkr_unpack(output_buffer, input_buffer);
|
|
||||||
int out_size = (char*)end_ptr - (char*)output_buffer;
|
|
||||||
|
|
||||||
printf("Uncompressed size: %d\n", out_size);
|
|
||||||
|
|
||||||
FILE* out_file = fopen(argv[2], "wb");
|
|
||||||
fwrite(output_buffer, 1, out_size, out_file);
|
|
||||||
fclose(out_file);
|
|
||||||
|
|
||||||
return 0;
|
|
||||||
}
|
|
||||||
@@ -1,4 +0,0 @@
|
|||||||
a very simple unpacker in c, as a reference for people wanting to implement their own unpacker.
|
|
||||||
absolutely not production ready, it makes no effort to ensure the output buffer can actually
|
|
||||||
hold the uncompressed data.
|
|
||||||
!!! Never run on untrusted input !!!
|
|
||||||
@@ -1,96 +0,0 @@
|
|||||||
typedef unsigned char u8;
|
|
||||||
typedef unsigned short u16;
|
|
||||||
typedef unsigned long u32;
|
|
||||||
|
|
||||||
u8* upkr_data_ptr;
|
|
||||||
u8 upkr_probs[1 + 255 + 1 + 2*32 + 2*32];
|
|
||||||
#ifdef UPKR_BITSTREAM
|
|
||||||
u16 upkr_state;
|
|
||||||
u8 upkr_current_byte;
|
|
||||||
int upkr_bits_left;
|
|
||||||
#else
|
|
||||||
u32 upkr_state;
|
|
||||||
#endif
|
|
||||||
|
|
||||||
int upkr_decode_bit(int context_index) {
|
|
||||||
#ifdef UPKR_BITSTREAM
|
|
||||||
while(upkr_state < 32768) {
|
|
||||||
if(upkr_bits_left == 0) {
|
|
||||||
upkr_current_byte = *upkr_data_ptr++;
|
|
||||||
upkr_bits_left = 8;
|
|
||||||
}
|
|
||||||
upkr_state = (upkr_state << 1) + (upkr_current_byte & 1);
|
|
||||||
upkr_current_byte >>= 1;
|
|
||||||
--upkr_bits_left;
|
|
||||||
}
|
|
||||||
#else
|
|
||||||
while(upkr_state < 4096) {
|
|
||||||
upkr_state = (upkr_state << 8) | *upkr_data_ptr++;
|
|
||||||
}
|
|
||||||
#endif
|
|
||||||
|
|
||||||
int prob = upkr_probs[context_index];
|
|
||||||
int bit = (upkr_state & 255) < prob ? 1 : 0;
|
|
||||||
|
|
||||||
if(bit) {
|
|
||||||
upkr_state = prob * (upkr_state >> 8) + (upkr_state & 255);
|
|
||||||
prob += (256 - prob + 8) >> 4;
|
|
||||||
} else {
|
|
||||||
upkr_state = (256 - prob) * (upkr_state >> 8) + (upkr_state & 255) - prob;
|
|
||||||
prob -= (prob + 8) >> 4;
|
|
||||||
}
|
|
||||||
upkr_probs[context_index] = prob;
|
|
||||||
|
|
||||||
return bit;
|
|
||||||
}
|
|
||||||
|
|
||||||
int upkr_decode_length(int context_index) {
|
|
||||||
int length = 0;
|
|
||||||
int bit_pos = 0;
|
|
||||||
while(upkr_decode_bit(context_index)) {
|
|
||||||
length |= upkr_decode_bit(context_index + 1) << bit_pos++;
|
|
||||||
context_index += 2;
|
|
||||||
}
|
|
||||||
return length | (1 << bit_pos);
|
|
||||||
}
|
|
||||||
|
|
||||||
void* upkr_unpack(void* destination, void* compressed_data) {
|
|
||||||
upkr_data_ptr = (u8*)compressed_data;
|
|
||||||
upkr_state = 0;
|
|
||||||
#ifdef UPKR_BITSTREAM
|
|
||||||
upkr_bits_left = 0;
|
|
||||||
#endif
|
|
||||||
for(int i = 0; i < sizeof(upkr_probs); ++i)
|
|
||||||
upkr_probs[i] = 128;
|
|
||||||
|
|
||||||
u8* write_ptr = (u8*)destination;
|
|
||||||
|
|
||||||
int prev_was_match = 0;
|
|
||||||
int offset = 0;
|
|
||||||
for(;;) {
|
|
||||||
if(upkr_decode_bit(0)) {
|
|
||||||
if(prev_was_match || upkr_decode_bit(256)) {
|
|
||||||
offset = upkr_decode_length(257) - 1;
|
|
||||||
if(offset == 0) {
|
|
||||||
break;
|
|
||||||
}
|
|
||||||
}
|
|
||||||
int length = upkr_decode_length(257 + 64);
|
|
||||||
while(length--) {
|
|
||||||
*write_ptr = write_ptr[-offset];
|
|
||||||
++write_ptr;
|
|
||||||
}
|
|
||||||
prev_was_match = 1;
|
|
||||||
} else {
|
|
||||||
int byte = 1;
|
|
||||||
while(byte < 256) {
|
|
||||||
int bit = upkr_decode_bit(byte);
|
|
||||||
byte = (byte << 1) + bit;
|
|
||||||
}
|
|
||||||
*write_ptr++ = byte;
|
|
||||||
prev_was_match = 0;
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
return write_ptr;
|
|
||||||
}
|
|
||||||
@@ -6,12 +6,11 @@ use crate::ProgressCallback;
|
|||||||
pub fn pack(
|
pub fn pack(
|
||||||
data: &[u8],
|
data: &[u8],
|
||||||
use_bitstream: bool,
|
use_bitstream: bool,
|
||||||
parity_contexts: usize,
|
|
||||||
mut progress_callback: Option<ProgressCallback>,
|
mut progress_callback: Option<ProgressCallback>,
|
||||||
) -> Vec<u8> {
|
) -> Vec<u8> {
|
||||||
let mut match_finder = MatchFinder::new(data);
|
let mut match_finder = MatchFinder::new(data);
|
||||||
let mut rans_coder = RansCoder::new(use_bitstream);
|
let mut rans_coder = RansCoder::new(use_bitstream);
|
||||||
let mut state = lz::CoderState::new(parity_contexts);
|
let mut state = lz::CoderState::new();
|
||||||
|
|
||||||
let mut pos = 0;
|
let mut pos = 0;
|
||||||
while pos < data.len() {
|
while pos < data.len() {
|
||||||
|
|||||||
20
src/lib.rs
20
src/lib.rs
@@ -13,27 +13,11 @@ pub fn pack(
|
|||||||
data: &[u8],
|
data: &[u8],
|
||||||
level: u8,
|
level: u8,
|
||||||
use_bitstream: bool,
|
use_bitstream: bool,
|
||||||
parity_contexts: usize,
|
|
||||||
progress_callback: Option<ProgressCallback>,
|
progress_callback: Option<ProgressCallback>,
|
||||||
) -> Vec<u8> {
|
) -> Vec<u8> {
|
||||||
if level == 0 {
|
if level == 0 {
|
||||||
greedy_packer::pack(data, use_bitstream, parity_contexts, progress_callback)
|
greedy_packer::pack(data, use_bitstream, progress_callback)
|
||||||
} else {
|
} else {
|
||||||
parsing_packer::pack(
|
parsing_packer::pack(data, level, use_bitstream, progress_callback)
|
||||||
data,
|
|
||||||
level,
|
|
||||||
use_bitstream,
|
|
||||||
parity_contexts,
|
|
||||||
progress_callback,
|
|
||||||
)
|
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
pub fn compressed_size(mut data: &[u8]) -> f32 {
|
|
||||||
let mut state = 0;
|
|
||||||
while state < 4096 {
|
|
||||||
state = (state << 8) | data[0] as u32;
|
|
||||||
data = &data[1..];
|
|
||||||
}
|
|
||||||
data.len() as f32 + (state as f32).log2() / 8.
|
|
||||||
}
|
|
||||||
|
|||||||
58
src/lz.rs
58
src/lz.rs
@@ -9,49 +9,44 @@ pub enum Op {
|
|||||||
|
|
||||||
impl Op {
|
impl Op {
|
||||||
pub fn encode(&self, coder: &mut dyn EntropyCoder, state: &mut CoderState) {
|
pub fn encode(&self, coder: &mut dyn EntropyCoder, state: &mut CoderState) {
|
||||||
let literal_base = state.pos % state.parity_contexts * 256;
|
let base_context = 256 * (state.pos & 3);
|
||||||
match self {
|
match self {
|
||||||
&Op::Literal(lit) => {
|
&Op::Literal(lit) => {
|
||||||
encode_bit(coder, state, literal_base, false);
|
encode_bit(coder, state, base_context, false);
|
||||||
let mut context_index = 1;
|
let mut context_index = 1;
|
||||||
for i in (0..8).rev() {
|
for i in (0..8).rev() {
|
||||||
let bit = (lit >> i) & 1 != 0;
|
let bit = (lit >> i) & 1 != 0;
|
||||||
encode_bit(coder, state, literal_base + context_index, bit);
|
encode_bit(coder, state, base_context + context_index, bit);
|
||||||
context_index = (context_index << 1) | bit as usize;
|
context_index = (context_index << 1) | bit as usize;
|
||||||
}
|
}
|
||||||
state.prev_was_match = false;
|
|
||||||
state.pos += 1;
|
state.pos += 1;
|
||||||
|
state.prev_was_match = false;
|
||||||
}
|
}
|
||||||
&Op::Match { offset, len } => {
|
&Op::Match { offset, len } => {
|
||||||
encode_bit(coder, state, literal_base, true);
|
encode_bit(coder, state, base_context, true);
|
||||||
if !state.prev_was_match {
|
if !state.prev_was_match {
|
||||||
encode_bit(
|
encode_bit(coder, state, 1024, offset != state.last_offset);
|
||||||
coder,
|
|
||||||
state,
|
|
||||||
256 * state.parity_contexts,
|
|
||||||
offset != state.last_offset,
|
|
||||||
);
|
|
||||||
} else {
|
} else {
|
||||||
assert!(offset != state.last_offset);
|
assert!(offset != state.last_offset);
|
||||||
}
|
}
|
||||||
if offset != state.last_offset {
|
if offset != state.last_offset {
|
||||||
encode_length(coder, state, 256 * state.parity_contexts + 1, offset + 1);
|
encode_length(coder, state, 1025, offset + 1);
|
||||||
state.last_offset = offset;
|
state.last_offset = offset;
|
||||||
}
|
}
|
||||||
encode_length(coder, state, 256 * state.parity_contexts + 65, len);
|
encode_length(coder, state, 1025 + 64, len);
|
||||||
state.prev_was_match = true;
|
|
||||||
state.pos += len as usize;
|
state.pos += len as usize;
|
||||||
|
state.prev_was_match = true;
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
pub fn encode_eof(coder: &mut dyn EntropyCoder, state: &mut CoderState) {
|
pub fn encode_eof(coder: &mut dyn EntropyCoder, state: &mut CoderState) {
|
||||||
encode_bit(coder, state, state.pos % state.parity_contexts * 256, true);
|
encode_bit(coder, state, 256 * (state.pos & 3), true);
|
||||||
if !state.prev_was_match {
|
if !state.prev_was_match {
|
||||||
encode_bit(coder, state, 256 * state.parity_contexts, true);
|
encode_bit(coder, state, 1024, true);
|
||||||
}
|
}
|
||||||
encode_length(coder, state, 256 * state.parity_contexts + 1, 1);
|
encode_length(coder, state, 1025, 1);
|
||||||
}
|
}
|
||||||
|
|
||||||
fn encode_bit(
|
fn encode_bit(
|
||||||
@@ -84,20 +79,18 @@ fn encode_length(
|
|||||||
#[derive(Clone)]
|
#[derive(Clone)]
|
||||||
pub struct CoderState {
|
pub struct CoderState {
|
||||||
contexts: ContextState,
|
contexts: ContextState,
|
||||||
parity_contexts: usize,
|
|
||||||
last_offset: u32,
|
last_offset: u32,
|
||||||
prev_was_match: bool,
|
|
||||||
pos: usize,
|
pos: usize,
|
||||||
|
prev_was_match: bool,
|
||||||
}
|
}
|
||||||
|
|
||||||
impl CoderState {
|
impl CoderState {
|
||||||
pub fn new(parity_contexts: usize) -> CoderState {
|
pub fn new() -> CoderState {
|
||||||
CoderState {
|
CoderState {
|
||||||
contexts: ContextState::new((1 + 255) * parity_contexts + 1 + 64 + 64),
|
contexts: ContextState::new((1 + 255) * 4 + 1 + 64 + 64),
|
||||||
last_offset: 0,
|
last_offset: 0,
|
||||||
parity_contexts,
|
|
||||||
prev_was_match: false,
|
|
||||||
pos: 0,
|
pos: 0,
|
||||||
|
prev_was_match: false,
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
@@ -106,9 +99,9 @@ impl CoderState {
|
|||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
pub fn unpack(packed_data: &[u8], use_bitstream: bool, parity_contexts: usize) -> Vec<u8> {
|
pub fn unpack(packed_data: &[u8], use_bitstream: bool) -> Vec<u8> {
|
||||||
let mut decoder = RansDecoder::new(packed_data, use_bitstream);
|
let mut decoder = RansDecoder::new(packed_data, use_bitstream);
|
||||||
let mut contexts = ContextState::new((1 + 255) * parity_contexts + 1 + 64 + 64);
|
let mut contexts = ContextState::new((1 + 255) * 4 + 1 + 64 + 64);
|
||||||
let mut result = vec![];
|
let mut result = vec![];
|
||||||
let mut offset = 0;
|
let mut offset = 0;
|
||||||
let mut prev_was_match = false;
|
let mut prev_was_match = false;
|
||||||
@@ -131,17 +124,15 @@ pub fn unpack(packed_data: &[u8], use_bitstream: bool, parity_contexts: usize) -
|
|||||||
}
|
}
|
||||||
|
|
||||||
loop {
|
loop {
|
||||||
let literal_base = result.len() % parity_contexts * 256;
|
let base_context = 256 * (result.len() & 3);
|
||||||
if decoder.decode_with_context(&mut contexts.context_mut(literal_base)) {
|
if decoder.decode_with_context(&mut contexts.context_mut(base_context)) {
|
||||||
if prev_was_match
|
if prev_was_match || decoder.decode_with_context(&mut contexts.context_mut(1024)) {
|
||||||
|| decoder.decode_with_context(&mut contexts.context_mut(256 * parity_contexts))
|
offset = decode_length(&mut decoder, &mut contexts, 1025) - 1;
|
||||||
{
|
|
||||||
offset = decode_length(&mut decoder, &mut contexts, 256 * parity_contexts + 1) - 1;
|
|
||||||
if offset == 0 {
|
if offset == 0 {
|
||||||
break;
|
break;
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
let length = decode_length(&mut decoder, &mut contexts, 256 * parity_contexts + 65);
|
let length = decode_length(&mut decoder, &mut contexts, 1025 + 64);
|
||||||
for _ in 0..length {
|
for _ in 0..length {
|
||||||
result.push(result[result.len() - offset]);
|
result.push(result[result.len() - offset]);
|
||||||
}
|
}
|
||||||
@@ -150,8 +141,7 @@ pub fn unpack(packed_data: &[u8], use_bitstream: bool, parity_contexts: usize) -
|
|||||||
let mut context_index = 1;
|
let mut context_index = 1;
|
||||||
let mut byte = 0;
|
let mut byte = 0;
|
||||||
for i in (0..8).rev() {
|
for i in (0..8).rev() {
|
||||||
let bit = decoder
|
let bit = decoder.decode_with_context(&mut contexts.context_mut(base_context + context_index));
|
||||||
.decode_with_context(&mut contexts.context_mut(literal_base + context_index));
|
|
||||||
context_index = (context_index << 1) | bit as usize;
|
context_index = (context_index << 1) | bit as usize;
|
||||||
byte |= (bit as u8) << i;
|
byte |= (bit as u8) << i;
|
||||||
}
|
}
|
||||||
|
|||||||
30
src/main.rs
30
src/main.rs
@@ -1,6 +1,5 @@
|
|||||||
use anyhow::{bail, Result};
|
use anyhow::{bail, Result};
|
||||||
use std::io::prelude::*;
|
use std::io::prelude::*;
|
||||||
use std::process;
|
|
||||||
use std::{fs::File, path::PathBuf};
|
use std::{fs::File, path::PathBuf};
|
||||||
|
|
||||||
fn main() -> Result<()> {
|
fn main() -> Result<()> {
|
||||||
@@ -11,14 +10,6 @@ fn main() -> Result<()> {
|
|||||||
Some("pack") => {
|
Some("pack") => {
|
||||||
let level = args.opt_value_from_str(["-l", "--level"])?.unwrap_or(2u8);
|
let level = args.opt_value_from_str(["-l", "--level"])?.unwrap_or(2u8);
|
||||||
let use_bitstream = args.contains(["-b", "--bitstream"]);
|
let use_bitstream = args.contains(["-b", "--bitstream"]);
|
||||||
let parity_contexts = args
|
|
||||||
.opt_value_from_str(["-p", "--parity"])?
|
|
||||||
.unwrap_or(1usize);
|
|
||||||
|
|
||||||
if parity_contexts != 1 && parity_contexts != 2 && parity_contexts != 4 {
|
|
||||||
eprintln!("--parity has to be 1, 2 or 4");
|
|
||||||
process::exit(1);
|
|
||||||
}
|
|
||||||
|
|
||||||
let infile = args.free_from_os_str::<PathBuf, bool>(|s| Ok(s.into()))?;
|
let infile = args.free_from_os_str::<PathBuf, bool>(|s| Ok(s.into()))?;
|
||||||
let outfile = args.free_from_os_str::<PathBuf, bool>(|s| Ok(s.into()))?;
|
let outfile = args.free_from_os_str::<PathBuf, bool>(|s| Ok(s.into()))?;
|
||||||
@@ -32,7 +23,6 @@ fn main() -> Result<()> {
|
|||||||
&data,
|
&data,
|
||||||
level,
|
level,
|
||||||
use_bitstream,
|
use_bitstream,
|
||||||
parity_contexts,
|
|
||||||
Some(&mut |pos| {
|
Some(&mut |pos| {
|
||||||
pb.set(pos as u64);
|
pb.set(pos as u64);
|
||||||
}),
|
}),
|
||||||
@@ -49,21 +39,13 @@ fn main() -> Result<()> {
|
|||||||
}
|
}
|
||||||
Some("unpack") => {
|
Some("unpack") => {
|
||||||
let use_bitstream = args.contains(["-b", "--bitstream"]);
|
let use_bitstream = args.contains(["-b", "--bitstream"]);
|
||||||
let parity_contexts = args
|
|
||||||
.opt_value_from_str(["-p", "--parity"])?
|
|
||||||
.unwrap_or(1usize);
|
|
||||||
|
|
||||||
if parity_contexts != 1 && parity_contexts != 2 && parity_contexts != 4 {
|
|
||||||
eprintln!("--parity has to be 1, 2 or 4");
|
|
||||||
process::exit(1);
|
|
||||||
}
|
|
||||||
|
|
||||||
let infile = args.free_from_os_str::<PathBuf, bool>(|s| Ok(s.into()))?;
|
let infile = args.free_from_os_str::<PathBuf, bool>(|s| Ok(s.into()))?;
|
||||||
let outfile = args.free_from_os_str::<PathBuf, bool>(|s| Ok(s.into()))?;
|
let outfile = args.free_from_os_str::<PathBuf, bool>(|s| Ok(s.into()))?;
|
||||||
|
|
||||||
let mut data = vec![];
|
let mut data = vec![];
|
||||||
File::open(infile)?.read_to_end(&mut data)?;
|
File::open(infile)?.read_to_end(&mut data)?;
|
||||||
let packed_data = upkr::unpack(&data, use_bitstream, parity_contexts);
|
let packed_data = upkr::unpack(&data, use_bitstream);
|
||||||
File::create(outfile)?.write_all(&packed_data)?;
|
File::create(outfile)?.write_all(&packed_data)?;
|
||||||
}
|
}
|
||||||
Some(other) => {
|
Some(other) => {
|
||||||
@@ -76,11 +58,7 @@ fn main() -> Result<()> {
|
|||||||
|
|
||||||
fn print_help() {
|
fn print_help() {
|
||||||
eprintln!("Usage:");
|
eprintln!("Usage:");
|
||||||
eprintln!(" upkr pack [-b] [-l level(0-9)] [-p N] <infile> <outfile>");
|
eprintln!(" upkr pack [-l level(0-9)] <infile> <outfile>");
|
||||||
eprintln!(" upkr unpack [-b] [-p N] <infile> <outfile>");
|
eprintln!(" upkr unpack <infile> <outfile>");
|
||||||
eprintln!();
|
std::process::exit(1);
|
||||||
eprintln!(" -b, --bitstream bitstream mode");
|
|
||||||
eprintln!(" -l, --level N compression level 0-9");
|
|
||||||
eprintln!(" -p, --parity N use N (2/4) parity contexts");
|
|
||||||
process::exit(1);
|
|
||||||
}
|
}
|
||||||
|
|||||||
@@ -6,25 +6,14 @@ use crate::match_finder::MatchFinder;
|
|||||||
use crate::rans::{CostCounter, RansCoder};
|
use crate::rans::{CostCounter, RansCoder};
|
||||||
use crate::{lz, ProgressCallback};
|
use crate::{lz, ProgressCallback};
|
||||||
|
|
||||||
pub fn pack(
|
pub fn pack(data: &[u8], level: u8, use_bitstream: bool, progress_cb: Option<ProgressCallback>) -> Vec<u8> {
|
||||||
data: &[u8],
|
let mut parse = parse(data, Config::from_level(level), progress_cb);
|
||||||
level: u8,
|
|
||||||
use_bitstream: bool,
|
|
||||||
parity_contexts: usize,
|
|
||||||
progress_cb: Option<ProgressCallback>,
|
|
||||||
) -> Vec<u8> {
|
|
||||||
let mut parse = parse(
|
|
||||||
data,
|
|
||||||
Config::from_level(level),
|
|
||||||
parity_contexts,
|
|
||||||
progress_cb,
|
|
||||||
);
|
|
||||||
let mut ops = vec![];
|
let mut ops = vec![];
|
||||||
while let Some(link) = parse {
|
while let Some(link) = parse {
|
||||||
ops.push(link.op);
|
ops.push(link.op);
|
||||||
parse = link.prev.clone();
|
parse = link.prev.clone();
|
||||||
}
|
}
|
||||||
let mut state = lz::CoderState::new(parity_contexts);
|
let mut state = lz::CoderState::new();
|
||||||
let mut coder = RansCoder::new(use_bitstream);
|
let mut coder = RansCoder::new(use_bitstream);
|
||||||
for op in ops.into_iter().rev() {
|
for op in ops.into_iter().rev() {
|
||||||
op.encode(&mut coder, &mut state);
|
op.encode(&mut coder, &mut state);
|
||||||
@@ -49,7 +38,6 @@ type Arrivals = HashMap<usize, Vec<Arrival>>;
|
|||||||
fn parse(
|
fn parse(
|
||||||
data: &[u8],
|
data: &[u8],
|
||||||
config: Config,
|
config: Config,
|
||||||
parity_contexts: usize,
|
|
||||||
mut progress_cb: Option<ProgressCallback>,
|
mut progress_cb: Option<ProgressCallback>,
|
||||||
) -> Option<Rc<Parse>> {
|
) -> Option<Rc<Parse>> {
|
||||||
let mut match_finder = MatchFinder::new(data)
|
let mut match_finder = MatchFinder::new(data)
|
||||||
@@ -141,7 +129,7 @@ fn parse(
|
|||||||
0,
|
0,
|
||||||
Arrival {
|
Arrival {
|
||||||
parse: None,
|
parse: None,
|
||||||
state: lz::CoderState::new(parity_contexts),
|
state: lz::CoderState::new(),
|
||||||
cost: 0.0,
|
cost: 0.0,
|
||||||
},
|
},
|
||||||
max_arrivals,
|
max_arrivals,
|
||||||
|
|||||||
Reference in New Issue
Block a user