emo/busybox
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

bunzip: small code shrink and consmetics

Browse Source 
read_bunzip                                          276     283      +7
get_bits                                             184     162     -22
get_next_block                                      1833    1810     -23
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 1/2 up/down: 7/-45)             Total: -38 bytes
This commit is contained in:
Denis Vlasenko
2007-10-10 20:53:41 +00:00
parent db60bcf1b7
commit 52a4388d95
1 changed files with 28 additions and 22 deletions
Download Patch File Download Diff File Expand all files Collapse all files

50
archival/libunarchive/decompress_bunzip2.c
View File

@@ -44,7 +44,7 @@
#define RETVAL_LAST_BLOCK (-1) #define RETVAL_LAST_BLOCK (-1)
#define RETVAL_NOT_BZIP_DATA (-2) #define RETVAL_NOT_BZIP_DATA (-2)
#define RETVAL_UNEXPECTED_INPUT_EOF (-3) #define RETVAL_UNEXPECTED_INPUT_EOF (-3)
#define RETVAL_UNEXPECTED_OUTPUT_EOF (-4) #define RETVAL_SHORT_WRITE (-4)
#define RETVAL_DATA_ERROR (-5) #define RETVAL_DATA_ERROR (-5)
#define RETVAL_OUT_OF_MEMORY (-6) #define RETVAL_OUT_OF_MEMORY (-6)
#define RETVAL_OBSOLETE_INPUT (-7) #define RETVAL_OBSOLETE_INPUT (-7)
@@ -54,22 +54,27 @@
/* This is what we know about each Huffman coding group */ /* This is what we know about each Huffman coding group */
struct group_data { struct group_data {
/* We have an extra slot at the end of limit[] for a sentinal value. */ /* We have an extra slot at the end of limit[] for a sentinel value. */
int limit[MAX_HUFCODE_BITS+1], base[MAX_HUFCODE_BITS], permute[MAX_SYMBOLS]; int limit[MAX_HUFCODE_BITS+1], base[MAX_HUFCODE_BITS], permute[MAX_SYMBOLS];
int minLen, maxLen; int minLen, maxLen;
}; };
/* Structure holding all the housekeeping data, including IO buffers and /* Structure holding all the housekeeping data, including IO buffers and
memory that persists between calls to bunzip */ * memory that persists between calls to bunzip
* Found the most used member:
* cat this_file.c | sed -e 's/"/ /g' -e "s/'/ /g" | xargs -n1 \
* | grep 'bd->' | sed 's/^.*bd->/bd->/' | sort | $PAGER
* and moved it (inbufBitCount) to offset 0.
*/
struct bunzip_data { struct bunzip_data {
/* State for interrupting output loop */
int writeCopies, writePos, writeRunCountdown, writeCount, writeCurrent;
/* I/O tracking data (file handles, buffers, positions, etc.) */ /* I/O tracking data (file handles, buffers, positions, etc.) */
unsigned inbufBitCount, inbufBits;
int in_fd, out_fd, inbufCount, inbufPos /*, outbufPos*/; int in_fd, out_fd, inbufCount, inbufPos /*, outbufPos*/;
unsigned char *inbuf /*,*outbuf*/; unsigned char *inbuf /*,*outbuf*/;
unsigned inbufBitCount, inbufBits;
/* State for interrupting output loop */
int writeCopies, writePos, writeRunCountdown, writeCount, writeCurrent;
/* The CRC values stored in the block header and calculated from the data */ /* The CRC values stored in the block header and calculated from the data */
uint32_t headerCRC, totalCRC, writeCRC; uint32_t headerCRC, totalCRC, writeCRC;
@@ -80,7 +85,7 @@ struct bunzip_data {
/* For I/O error handling */ /* For I/O error handling */
jmp_buf jmpbuf; jmp_buf jmpbuf;
/* Big things go last (register-relative addressing can be larger for big offsets */ /* Big things go last (register-relative addressing can be larger for big offsets) */
uint32_t crc32Table[256]; uint32_t crc32Table[256];
unsigned char selectors[32768]; /* nSelectors=15 bits */ unsigned char selectors[32768]; /* nSelectors=15 bits */
struct group_data groups[MAX_GROUPS]; /* Huffman coding tables */ struct group_data groups[MAX_GROUPS]; /* Huffman coding tables */
@@ -91,7 +96,7 @@ struct bunzip_data {
/* Return the next nnn bits of input. All reads from the compressed input /* Return the next nnn bits of input. All reads from the compressed input
are done through this function. All reads are big endian */ are done through this function. All reads are big endian */
static unsigned get_bits(bunzip_data *bd, char bits_wanted) static unsigned get_bits(bunzip_data *bd, int bits_wanted)
{ {
unsigned bits = 0; unsigned bits = 0;
@@ -121,7 +126,7 @@ static unsigned get_bits(bunzip_data *bd, char bits_wanted)
/* Grab next 8 bits of input from buffer. */ /* Grab next 8 bits of input from buffer. */
bd->inbufBits = (bd->inbufBits<<8) | bd->inbuf[bd->inbufPos++]; bd->inbufBits = (bd->inbufBits << 8) | bd->inbuf[bd->inbufPos++];
bd->inbufBitCount += 8; bd->inbufBitCount += 8;
} }
@@ -205,7 +210,7 @@ static int get_next_block(bunzip_data *bd)
/* Get next value */ /* Get next value */
for (j = 0; get_bits(bd, 1); j++) for (j = 0; get_bits(bd, 1); j++)
if (j>=groupCount) return RETVAL_DATA_ERROR; if (j >= groupCount) return RETVAL_DATA_ERROR;
/* Decode MTF to get the next selector */ /* Decode MTF to get the next selector */
@@ -320,19 +325,20 @@ static int get_next_block(bunzip_data *bd)
t += temp[i]; t += temp[i];
base[i+1] = pp - t; base[i+1] = pp - t;
} }
limit[maxLen+1] = INT_MAX; /* Sentinal value for reading next sym. */ limit[maxLen+1] = INT_MAX; /* Sentinel value for reading next sym. */
limit[maxLen] = pp + temp[maxLen] - 1; limit[maxLen] = pp + temp[maxLen] - 1;
base[minLen] = 0; base[minLen] = 0;
} }
/* We've finished reading and digesting the block header. Now read this /* We've finished reading and digesting the block header. Now read this
block's Huffman coded symbols from the file and undo the Huffman coding block's Huffman coded symbols from the file and undo the Huffman coding
and run length encoding, saving the result into dbuf[dbufCount++]=uc */ and run length encoding, saving the result into dbuf[dbufCount++] = uc */
/* Initialize symbol occurrence counters and symbol Move To Front table */ /* Initialize symbol occurrence counters and symbol Move To Front table */
memset(byteCount, 0, sizeof(byteCount)); /* smaller, maybe slower? */
for (i = 0; i < 256; i++) { for (i = 0; i < 256; i++) {
byteCount[i] = 0; //byteCount[i] = 0;
mtfSymbol[i] = (unsigned char)i; mtfSymbol[i] = (unsigned char)i;
} }
@@ -547,7 +553,7 @@ int read_bunzip(bunzip_data *bd, char *outbuf, int len)
/* If the output buffer is full, snapshot state and return */ /* If the output buffer is full, snapshot state and return */
if (gotcount >= len) { if (gotcount >= len) {
bd->writePos =pos; bd->writePos = pos;
bd->writeCurrent = current; bd->writeCurrent = current;
bd->writeCopies++; bd->writeCopies++;
return len; return len;
@@ -573,8 +579,8 @@ int read_bunzip(bunzip_data *bd, char *outbuf, int len)
current = pos & 0xff; current = pos & 0xff;
pos >>= 8; pos >>= 8;
/* After 3 consecutive copies of the same byte, the 4th is a repeat /* After 3 consecutive copies of the same byte, the 4th
count. We count down from 4 instead * is a repeat count. We count down from 4 instead
* of counting up because testing for non-zero is faster */ * of counting up because testing for non-zero is faster */
if (--bd->writeRunCountdown) { if (--bd->writeRunCountdown) {
@@ -606,7 +612,7 @@ int read_bunzip(bunzip_data *bd, char *outbuf, int len)
/* If this block had a CRC error, force file level CRC error. */ /* If this block had a CRC error, force file level CRC error. */
if (bd->writeCRC != bd->headerCRC) { if (bd->writeCRC != bd->headerCRC) {
bd->totalCRC = bd->headerCRC+1; bd->totalCRC = bd->headerCRC + 1;
return RETVAL_LAST_BLOCK; return RETVAL_LAST_BLOCK;
} }
} }
@@ -713,8 +719,8 @@ unpack_bz2_stream(int src_fd, int dst_fd)
for (;;) { for (;;) {
i = read_bunzip(bd, outbuf, IOBUF_SIZE); i = read_bunzip(bd, outbuf, IOBUF_SIZE);
if (i <= 0) break; if (i <= 0) break;
if (i != safe_write(dst_fd, outbuf, i)) { if (i != full_write(dst_fd, outbuf, i)) {
i = RETVAL_UNEXPECTED_OUTPUT_EOF; i = RETVAL_SHORT_WRITE;
break; break;
} }
USE_DESKTOP(total_written += i;) USE_DESKTOP(total_written += i;)
@@ -729,8 +735,8 @@ unpack_bz2_stream(int src_fd, int dst_fd)
} else { } else {
i = RETVAL_OK; i = RETVAL_OK;
} }
} else if (i == RETVAL_UNEXPECTED_OUTPUT_EOF) { } else if (i == RETVAL_SHORT_WRITE) {
bb_error_msg("unexpected EOF"); bb_error_msg("short write");
} else { } else {
bb_error_msg("bunzip error %d", i); bb_error_msg("bunzip error %d", i);
} }