Work in progress changes for packJPG encoding and decoding.
Enhance custom LibArchive filter functionlity.
This commit is contained in:
parent
75dfa6a6fb
commit
e90c52e516
9 changed files with 318 additions and 97 deletions
|
@ -43,88 +43,183 @@
|
||||||
#include "pc_archive.h"
|
#include "pc_archive.h"
|
||||||
|
|
||||||
#define PACKJPG_DEF_BUFSIZ (512 * 1024)
|
#define PACKJPG_DEF_BUFSIZ (512 * 1024)
|
||||||
#define JPG_SIZE_LIMIT (100 * 1024 * 1024)
|
#define JPG_SIZE_LIMIT (50 * 1024 * 1024)
|
||||||
|
|
||||||
struct packjpg_filter_data {
|
struct packjpg_filter_data {
|
||||||
uchar_t *buff;
|
uchar_t *buff, *in_buff;
|
||||||
size_t bufflen;
|
size_t bufflen, in_bufflen;
|
||||||
};
|
};
|
||||||
|
|
||||||
extern size_t packjpg_filter_process(uchar_t *in_buf, size_t len, uchar_t **out_buf);
|
extern size_t packjpg_filter_process(uchar_t *in_buf, size_t len, uchar_t **out_buf);
|
||||||
|
|
||||||
int packjpg_filter(struct filter_info *fi, void *filter_private);
|
int64_t packjpg_filter(struct filter_info *fi, void *filter_private);
|
||||||
|
|
||||||
void
|
void
|
||||||
add_filters_by_ext()
|
add_filters_by_type(struct type_data *typetab)
|
||||||
{
|
{
|
||||||
struct packjpg_filter_data *pjdat;
|
struct packjpg_filter_data *pjdat;
|
||||||
|
int slot;
|
||||||
|
|
||||||
pjdat = (struct packjpg_filter_data *)malloc(sizeof (struct packjpg_filter_data));
|
pjdat = (struct packjpg_filter_data *)malloc(sizeof (struct packjpg_filter_data));
|
||||||
pjdat->buff = (uchar_t *)malloc(PACKJPG_DEF_BUFSIZ);
|
pjdat->buff = (uchar_t *)malloc(PACKJPG_DEF_BUFSIZ);
|
||||||
pjdat->bufflen = PACKJPG_DEF_BUFSIZ;
|
pjdat->bufflen = PACKJPG_DEF_BUFSIZ;
|
||||||
if (insert_filter_data(packjpg_filter, pjdat, "pjg") != 0) {
|
pjdat->in_buff = NULL;
|
||||||
free(pjdat->buff);
|
pjdat->in_bufflen = 0;
|
||||||
free(pjdat);
|
|
||||||
log_msg(LOG_WARN, 0, "Failed to add filter module for packJPG.");
|
slot = TYPE_JPEG >> 3;
|
||||||
}
|
typetab[slot].filter_private = pjdat;
|
||||||
|
typetab[slot].filter_func = packjpg_filter;
|
||||||
|
typetab[slot].filter_name = "packJPG";
|
||||||
|
slot = TYPE_PACKJPG >> 3;
|
||||||
|
typetab[slot].filter_private = pjdat;
|
||||||
|
typetab[slot].filter_func = packjpg_filter;
|
||||||
|
typetab[slot].filter_name = "packJPG";
|
||||||
}
|
}
|
||||||
|
|
||||||
/* a short reminder about input/output stream types
|
static ssize_t
|
||||||
for the pjglib_init_streams() function
|
copy_archive_data(struct archive *ar, uchar_t *out_buf)
|
||||||
|
{
|
||||||
|
int64_t offset;
|
||||||
|
const void *buff;
|
||||||
|
size_t size, tot;
|
||||||
|
int r;
|
||||||
|
|
||||||
if input is file
|
tot = 0;
|
||||||
----------------
|
for (;;) {
|
||||||
in_scr -> name of input file
|
r = archive_read_data_block(ar, &buff, &size, &offset);
|
||||||
in_type -> 0
|
if (r == ARCHIVE_EOF)
|
||||||
in_size -> ignore
|
break;
|
||||||
|
if (r != ARCHIVE_OK)
|
||||||
|
return (0);
|
||||||
|
memcpy(out_buf + offset, buff, size);
|
||||||
|
tot += size;
|
||||||
|
}
|
||||||
|
return (tot);
|
||||||
|
}
|
||||||
|
|
||||||
if input is memory
|
static ssize_t
|
||||||
------------------
|
write_archive_data(struct archive *aw, uchar_t *out_buf, size_t len, int block_size)
|
||||||
in_scr -> array containg data
|
{
|
||||||
in_type -> 1
|
int64_t offset;
|
||||||
in_size -> size of data array
|
uchar_t *buff;
|
||||||
|
int r;
|
||||||
|
size_t tot;
|
||||||
|
|
||||||
if input is *FILE (f.e. stdin)
|
buff = out_buf;
|
||||||
------------------------------
|
offset = 0;
|
||||||
in_src -> stream pointer
|
tot = len;
|
||||||
in_type -> 2
|
while (len > 0) {
|
||||||
in_size -> ignore
|
if (len < block_size)
|
||||||
|
block_size = len;
|
||||||
|
r = (int)archive_write_data_block(aw, buff, block_size, offset);
|
||||||
|
if (r < ARCHIVE_WARN)
|
||||||
|
r = ARCHIVE_WARN;
|
||||||
|
if (r != ARCHIVE_OK) {
|
||||||
|
return (r);
|
||||||
|
}
|
||||||
|
offset += block_size;
|
||||||
|
len -= block_size;
|
||||||
|
}
|
||||||
|
return (tot);
|
||||||
|
}
|
||||||
|
|
||||||
vice versa for output streams! */
|
/*
|
||||||
|
* Helper routine to bridge to packJPG C++ lib, without changing packJPG itself.
|
||||||
int
|
*/
|
||||||
|
ssize_t
|
||||||
packjpg_filter(struct filter_info *fi, void *filter_private)
|
packjpg_filter(struct filter_info *fi, void *filter_private)
|
||||||
{
|
{
|
||||||
struct packjpg_filter_data *pjdat = (struct packjpg_filter_data *)filter_private;
|
struct packjpg_filter_data *pjdat = (struct packjpg_filter_data *)filter_private;
|
||||||
uchar_t *mapbuf, *out;
|
uchar_t *mapbuf, *out;
|
||||||
size_t len;
|
size_t len, in_size = 0;
|
||||||
|
|
||||||
len = archive_entry_size(fi->entry);
|
len = archive_entry_size(fi->entry);
|
||||||
if (len > JPG_SIZE_LIMIT) // Bork on massive JPEGs
|
if (len > JPG_SIZE_LIMIT) // Bork on massive JPEGs
|
||||||
return (-1);
|
return (FILTER_RETURN_SKIP);
|
||||||
|
|
||||||
|
if (fi->compressing) {
|
||||||
mapbuf = mmap(NULL, len, PROT_READ, MAP_SHARED, fi->fd, 0);
|
mapbuf = mmap(NULL, len, PROT_READ, MAP_SHARED, fi->fd, 0);
|
||||||
if (mapbuf == NULL)
|
if (mapbuf == NULL) {
|
||||||
return (-1);
|
log_msg(LOG_ERR, 1, "Mmap failed in packJPG filter.");
|
||||||
|
return (FILTER_RETURN_ERROR);
|
||||||
|
}
|
||||||
|
|
||||||
|
/*
|
||||||
|
* We are trying to compress and this is not a jpeg. Skip.
|
||||||
|
*/
|
||||||
|
if (mapbuf[0] != 0xFF && mapbuf[1] != 0xD8) {
|
||||||
|
munmap(mapbuf, len);
|
||||||
|
return (FILTER_RETURN_SKIP);
|
||||||
|
}
|
||||||
|
|
||||||
|
} else {
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Allocate input buffer and read archive data stream for the entry
|
||||||
|
* into this buffer.
|
||||||
|
*/
|
||||||
|
if (pjdat->in_bufflen < len) {
|
||||||
|
if (pjdat->in_buff) free(pjdat->in_buff);
|
||||||
|
pjdat->in_bufflen = len;
|
||||||
|
pjdat->in_buff = malloc(pjdat->in_bufflen);
|
||||||
|
if (pjdat->in_buff == NULL) {
|
||||||
|
log_msg(LOG_ERR, 1, "Out of memory.");
|
||||||
|
return (FILTER_RETURN_ERROR);
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
in_size = copy_archive_data(fi->source_arc, pjdat->in_buff);
|
||||||
|
if (in_size != len) {
|
||||||
|
log_msg(LOG_ERR, 0, "Failed to read archive data.");
|
||||||
|
return (FILTER_RETURN_ERROR);
|
||||||
|
}
|
||||||
|
in_size = U64_P(pjdat->in_buff);
|
||||||
|
mapbuf = pjdat->in_buff + 8;
|
||||||
|
|
||||||
|
/*
|
||||||
|
* We are trying to decompress and this is not a packJPG file.
|
||||||
|
* Write the raw data and skip.
|
||||||
|
*/
|
||||||
|
if (mapbuf[0] != 'J' && mapbuf[1] != 'S') {
|
||||||
|
return (write_archive_data(fi->target_arc, mapbuf, in_size,
|
||||||
|
fi->block_size));
|
||||||
|
}
|
||||||
|
}
|
||||||
if (pjdat->bufflen < len) {
|
if (pjdat->bufflen < len) {
|
||||||
free(pjdat->buff);
|
free(pjdat->buff);
|
||||||
pjdat->bufflen = len;
|
pjdat->bufflen = len; // Include size for compressed len
|
||||||
pjdat->buff = malloc(pjdat->bufflen);
|
pjdat->buff = malloc(pjdat->bufflen);
|
||||||
if (pjdat->buff == NULL) {
|
if (pjdat->buff == NULL) {
|
||||||
log_msg(LOG_ERR, 1, "Out of memory.");
|
log_msg(LOG_ERR, 1, "Out of memory.");
|
||||||
munmap(mapbuf, len);
|
munmap(mapbuf, len);
|
||||||
return (-1);
|
return (FILTER_RETURN_ERROR);
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
/*
|
/*
|
||||||
* Helper routine to bridge to packJPG C++ lib, without changing packJPG itself.
|
* Compression case.
|
||||||
*/
|
*/
|
||||||
|
if (fi->compressing) {
|
||||||
|
ssize_t rv;
|
||||||
|
|
||||||
out = pjdat->buff;
|
out = pjdat->buff;
|
||||||
if ((len = packjpg_filter_process(mapbuf, len, &out)) == 0) {
|
if ((len = packjpg_filter_process(mapbuf, len, &out)) == 0) {
|
||||||
return (-1);
|
return (FILTER_RETURN_SKIP);
|
||||||
}
|
}
|
||||||
|
in_size = LE64(len);
|
||||||
|
rv = archive_write_data(fi->target_arc, &in_size, 8);
|
||||||
|
if (rv != 8)
|
||||||
|
return (rv);
|
||||||
return (archive_write_data(fi->target_arc, out, len));
|
return (archive_write_data(fi->target_arc, out, len));
|
||||||
|
}
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Decompression case.
|
||||||
|
*/
|
||||||
|
out = pjdat->buff;
|
||||||
|
if ((len = packjpg_filter_process(mapbuf, in_size, &out)) == 0) {
|
||||||
|
return (FILTER_RETURN_ERROR);
|
||||||
|
}
|
||||||
|
return (write_archive_data(fi->target_arc, out, len, fi->block_size));
|
||||||
}
|
}
|
||||||
|
|
||||||
|
|
|
@ -28,6 +28,7 @@
|
||||||
|
|
||||||
#include <sys/types.h>
|
#include <sys/types.h>
|
||||||
#include <sys/stat.h>
|
#include <sys/stat.h>
|
||||||
|
#include <inttypes.h>
|
||||||
#include <unistd.h>
|
#include <unistd.h>
|
||||||
#include <archive.h>
|
#include <archive.h>
|
||||||
#include <archive_entry.h>
|
#include <archive_entry.h>
|
||||||
|
@ -36,15 +37,26 @@
|
||||||
extern "C" {
|
extern "C" {
|
||||||
#endif
|
#endif
|
||||||
|
|
||||||
|
#define FILTER_RETURN_SKIP (1)
|
||||||
|
#define FILTER_RETURN_ERROR (-1)
|
||||||
|
|
||||||
struct filter_info {
|
struct filter_info {
|
||||||
|
struct archive *source_arc;
|
||||||
struct archive *target_arc;
|
struct archive *target_arc;
|
||||||
struct archive_entry *entry;
|
struct archive_entry *entry;
|
||||||
int fd;
|
int fd;
|
||||||
|
int compressing, block_size;
|
||||||
};
|
};
|
||||||
|
|
||||||
typedef int (*filter_func_ptr)(struct filter_info *fi, void *filter_private);
|
typedef ssize_t (*filter_func_ptr)(struct filter_info *fi, void *filter_private);
|
||||||
|
|
||||||
void add_filters_by_ext();
|
struct type_data {
|
||||||
|
void *filter_private;
|
||||||
|
filter_func_ptr filter_func;
|
||||||
|
char *filter_name;
|
||||||
|
};
|
||||||
|
|
||||||
|
void add_filters_by_type(struct type_data *typetab);
|
||||||
|
|
||||||
#ifdef __cplusplus
|
#ifdef __cplusplus
|
||||||
}
|
}
|
||||||
|
|
|
@ -60,10 +60,10 @@ pthread_mutex_t init_mutex = PTHREAD_MUTEX_INITIALIZER;
|
||||||
static struct ext_hash_entry {
|
static struct ext_hash_entry {
|
||||||
uint64_t extnum;
|
uint64_t extnum;
|
||||||
int type;
|
int type;
|
||||||
void *filter_private;
|
|
||||||
filter_func_ptr filter_func;
|
|
||||||
} *exthtab = NULL;
|
} *exthtab = NULL;
|
||||||
|
|
||||||
|
struct type_data typetab[NUM_SUB_TYPES];
|
||||||
|
|
||||||
/*
|
/*
|
||||||
AE_IFREG Regular file
|
AE_IFREG Regular file
|
||||||
AE_IFLNK Symbolic link
|
AE_IFLNK Symbolic link
|
||||||
|
@ -74,12 +74,12 @@ AE_IFDIR Directory
|
||||||
AE_IFIFO Named pipe (fifo)
|
AE_IFIFO Named pipe (fifo)
|
||||||
*/
|
*/
|
||||||
|
|
||||||
#define ARC_ENTRY_OVRHEAD 500
|
#define ARC_ENTRY_OVRHEAD 1024
|
||||||
#define ARC_SCRATCH_BUFF_SIZE (64 *1024)
|
|
||||||
#define MMAP_SIZE (1024 * 1024)
|
#define MMAP_SIZE (1024 * 1024)
|
||||||
#define SORT_BUF_SIZE (65536)
|
#define SORT_BUF_SIZE (65536)
|
||||||
#define NAMELEN 4
|
#define NAMELEN 4
|
||||||
#define TEMP_MMAP_SIZE (128 * 1024)
|
#define TEMP_MMAP_SIZE (128 * 1024)
|
||||||
|
#define AW_BLOCK_SIZE (256 * 1024)
|
||||||
|
|
||||||
typedef struct member_entry {
|
typedef struct member_entry {
|
||||||
char name[NAMELEN];
|
char name[NAMELEN];
|
||||||
|
@ -104,7 +104,7 @@ static struct arc_list_state {
|
||||||
|
|
||||||
pthread_mutex_t nftw_mutex = PTHREAD_MUTEX_INITIALIZER;
|
pthread_mutex_t nftw_mutex = PTHREAD_MUTEX_INITIALIZER;
|
||||||
|
|
||||||
static int detect_type_by_ext(char *path, int pathlen);
|
static int detect_type_by_ext(const char *path, int pathlen);
|
||||||
static int detect_type_by_data(uchar_t *buf, size_t len);
|
static int detect_type_by_data(uchar_t *buf, size_t len);
|
||||||
|
|
||||||
/*
|
/*
|
||||||
|
@ -164,6 +164,12 @@ creat_write_callback(struct archive *arc, void *ctx, const void *buf, size_t len
|
||||||
uchar_t *tbuf;
|
uchar_t *tbuf;
|
||||||
|
|
||||||
tbuf = pctx->arc_buf + pctx->arc_buf_pos;
|
tbuf = pctx->arc_buf + pctx->arc_buf_pos;
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Determine if we should return the accumulated data to the caller.
|
||||||
|
* This is done if the data type changes and at least some minimum amount
|
||||||
|
* of data has accumulated in the buffer.
|
||||||
|
*/
|
||||||
if (pctx->btype != pctx->ctype) {
|
if (pctx->btype != pctx->ctype) {
|
||||||
if (pctx->btype == TYPE_UNKNOWN || pctx->arc_buf_pos == 0) {
|
if (pctx->btype == TYPE_UNKNOWN || pctx->arc_buf_pos == 0) {
|
||||||
pctx->btype = pctx->ctype;
|
pctx->btype = pctx->ctype;
|
||||||
|
@ -718,12 +724,23 @@ setup_archiver(pc_ctx_t *pctx, struct stat *sbuf)
|
||||||
err = nftw(fn->filename, add_pathname, 1024, FTW_PHYS);
|
err = nftw(fn->filename, add_pathname, 1024, FTW_PHYS);
|
||||||
} else {
|
} else {
|
||||||
int tflag;
|
int tflag;
|
||||||
|
struct FTW ftwbuf;
|
||||||
|
char *pos;
|
||||||
|
|
||||||
if (S_ISLNK(sb.st_mode))
|
if (S_ISLNK(sb.st_mode))
|
||||||
tflag = FTW_SL;
|
tflag = FTW_SL;
|
||||||
else
|
else
|
||||||
tflag = FTW_F;
|
tflag = FTW_F;
|
||||||
add_pathname(fn->filename, &sb, tflag, NULL);
|
|
||||||
|
/*
|
||||||
|
* Find out basename to mimic FTW.
|
||||||
|
*/
|
||||||
|
pos = strrchr(fn->filename, PATHSEP_CHAR);
|
||||||
|
if (pos)
|
||||||
|
ftwbuf.base = pos - fn->filename + 1;
|
||||||
|
else
|
||||||
|
ftwbuf.base = 0;
|
||||||
|
add_pathname(fn->filename, &sb, tflag, &ftwbuf);
|
||||||
a_state.arc_size = sb.st_size;
|
a_state.arc_size = sb.st_size;
|
||||||
}
|
}
|
||||||
if (a_state.bufpos > 0) {
|
if (a_state.bufpos > 0) {
|
||||||
|
@ -766,6 +783,7 @@ setup_archiver(pc_ctx_t *pctx, struct stat *sbuf)
|
||||||
return (-1);
|
return (-1);
|
||||||
}
|
}
|
||||||
archive_write_set_format_pax_restricted(arc);
|
archive_write_set_format_pax_restricted(arc);
|
||||||
|
archive_write_set_bytes_per_block(arc, 0);
|
||||||
archive_write_open(arc, pctx, arc_open_callback,
|
archive_write_open(arc, pctx, arc_open_callback,
|
||||||
creat_write_callback, creat_close_callback);
|
creat_write_callback, creat_close_callback);
|
||||||
pctx->archive_ctx = arc;
|
pctx->archive_ctx = arc;
|
||||||
|
@ -774,7 +792,7 @@ setup_archiver(pc_ctx_t *pctx, struct stat *sbuf)
|
||||||
pctx->temp_mmap_buf = mmap(NULL, pctx->temp_mmap_len, PROT_READ,
|
pctx->temp_mmap_buf = mmap(NULL, pctx->temp_mmap_len, PROT_READ,
|
||||||
MAP_SHARED, pctx->archive_members_fd, 0);
|
MAP_SHARED, pctx->archive_members_fd, 0);
|
||||||
if (pctx->temp_mmap_buf == NULL) {
|
if (pctx->temp_mmap_buf == NULL) {
|
||||||
log_msg(LOG_WARN, 1, "Unable to mmap pathlist file, switching to slower read().");
|
log_msg(LOG_WARN, 1, "Unable to mmap pathlist file, switching to read().");
|
||||||
pctx->temp_mmap_len = 0;
|
pctx->temp_mmap_len = 0;
|
||||||
}
|
}
|
||||||
pctx->temp_mmap_pos = 0;
|
pctx->temp_mmap_pos = 0;
|
||||||
|
@ -810,6 +828,28 @@ setup_extractor(pc_ctx_t *pctx)
|
||||||
return (0);
|
return (0);
|
||||||
}
|
}
|
||||||
|
|
||||||
|
static ssize_t
|
||||||
|
process_by_filter(int fd, int typ, struct archive *target_arc,
|
||||||
|
struct archive *source_arc, struct archive_entry *entry, int cmp)
|
||||||
|
{
|
||||||
|
struct filter_info fi;
|
||||||
|
int64_t wrtn;
|
||||||
|
|
||||||
|
fi.source_arc = source_arc;
|
||||||
|
fi.target_arc = target_arc;
|
||||||
|
fi.entry = entry;
|
||||||
|
fi.fd = fd;
|
||||||
|
fi.compressing = cmp;
|
||||||
|
fi.block_size = AW_BLOCK_SIZE;
|
||||||
|
wrtn = (*(typetab[(typ >> 3)].filter_func))(&fi, typetab[(typ >> 3)].filter_private);
|
||||||
|
close(fd);
|
||||||
|
if (wrtn == FILTER_RETURN_ERROR) {
|
||||||
|
log_msg(LOG_ERR, 0, "Error invoking filter module: %s",
|
||||||
|
typetab[(typ >> 3)].filter_name);
|
||||||
|
}
|
||||||
|
return (wrtn);
|
||||||
|
}
|
||||||
|
|
||||||
/*
|
/*
|
||||||
* Routines to archive members and write the file data to the callback. Portions of
|
* Routines to archive members and write the file data to the callback. Portions of
|
||||||
* the following code is adapted from some of the Libarchive bsdtar code.
|
* the following code is adapted from some of the Libarchive bsdtar code.
|
||||||
|
@ -834,6 +874,20 @@ copy_file_data(pc_ctx_t *pctx, struct archive *arc, struct archive_entry *entry,
|
||||||
return (-1);
|
return (-1);
|
||||||
}
|
}
|
||||||
|
|
||||||
|
if (typ != TYPE_UNKNOWN) {
|
||||||
|
if (typetab[(typ >> 3)].filter_func != NULL) {
|
||||||
|
int64_t rv;
|
||||||
|
|
||||||
|
rv = process_by_filter(fd, typ, arc, NULL, entry, 1);
|
||||||
|
if (rv == FILTER_RETURN_ERROR)
|
||||||
|
return (-1);
|
||||||
|
else if (rv == FILTER_RETURN_SKIP)
|
||||||
|
lseek(fd, 0, SEEK_SET);
|
||||||
|
else
|
||||||
|
return (ARCHIVE_OK);
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
/*
|
/*
|
||||||
* Use mmap for copying file data. Not necessarily for performance, but it saves on
|
* Use mmap for copying file data. Not necessarily for performance, but it saves on
|
||||||
* resident memory use.
|
* resident memory use.
|
||||||
|
@ -847,6 +901,7 @@ copy_file_data(pc_ctx_t *pctx, struct archive *arc, struct archive_entry *entry,
|
||||||
len = bytes_to_write;
|
len = bytes_to_write;
|
||||||
else
|
else
|
||||||
len = MMAP_SIZE;
|
len = MMAP_SIZE;
|
||||||
|
do_map:
|
||||||
mapbuf = mmap(NULL, len, PROT_READ, MAP_SHARED, fd, offset);
|
mapbuf = mmap(NULL, len, PROT_READ, MAP_SHARED, fd, offset);
|
||||||
if (mapbuf == NULL) {
|
if (mapbuf == NULL) {
|
||||||
/* Mmap failed; this is bad. */
|
/* Mmap failed; this is bad. */
|
||||||
|
@ -858,8 +913,28 @@ copy_file_data(pc_ctx_t *pctx, struct archive *arc, struct archive_entry *entry,
|
||||||
src = mapbuf;
|
src = mapbuf;
|
||||||
wlen = len;
|
wlen = len;
|
||||||
|
|
||||||
if (typ == TYPE_UNKNOWN)
|
if (typ == TYPE_UNKNOWN) {
|
||||||
pctx->ctype = detect_type_by_data(src, len);
|
pctx->ctype = detect_type_by_data(src, len);
|
||||||
|
if (typ != TYPE_UNKNOWN) {
|
||||||
|
if (typetab[(typ >> 3)].filter_func != NULL) {
|
||||||
|
int64_t rv;
|
||||||
|
munmap(mapbuf, len);
|
||||||
|
|
||||||
|
rv = process_by_filter(fd, typ, arc, NULL, entry, 1);
|
||||||
|
if (rv == FILTER_RETURN_ERROR) {
|
||||||
|
return (-1);
|
||||||
|
} else if (rv == FILTER_RETURN_SKIP) {
|
||||||
|
lseek(fd, 0, SEEK_SET);
|
||||||
|
typ = TYPE_COMPRESSED;
|
||||||
|
offset = 0;
|
||||||
|
goto do_map;
|
||||||
|
} else {
|
||||||
|
return (ARCHIVE_OK);
|
||||||
|
}
|
||||||
|
}
|
||||||
|
}
|
||||||
|
}
|
||||||
|
typ = TYPE_COMPRESSED; // Need to avoid calling detect_type_by_data subsequently.
|
||||||
|
|
||||||
/*
|
/*
|
||||||
* Write the entire mmap-ed buffer. Since we are writing to the compressor
|
* Write the entire mmap-ed buffer. Since we are writing to the compressor
|
||||||
|
@ -983,10 +1058,10 @@ archiver_thread_func(void *dat) {
|
||||||
* libarchive refuses to extract these files on Linux, no matter what I try.
|
* libarchive refuses to extract these files on Linux, no matter what I try.
|
||||||
* Bug?
|
* Bug?
|
||||||
*
|
*
|
||||||
* In this case the file basename is changed and a custom extended attribute
|
* In this case the file basename is changed and a custom flag is set to
|
||||||
* is set to indicate extraction to change it back.
|
* indicate extraction to change it back.
|
||||||
*/
|
*/
|
||||||
if (bnchars[0] == '.' && bnchars[1] == '_') {
|
if (bnchars[0] == '.' && bnchars[1] == '_' && archive_entry_filetype(entry) == AE_IFREG) {
|
||||||
char *pos = strstr(name, "._");
|
char *pos = strstr(name, "._");
|
||||||
char name[] = "@.", value[] = "m";
|
char name[] = "@.", value[] = "m";
|
||||||
if (pos) {
|
if (pos) {
|
||||||
|
@ -1001,6 +1076,8 @@ archiver_thread_func(void *dat) {
|
||||||
|
|
||||||
if (archive_entry_filetype(entry) != AE_IFREG) {
|
if (archive_entry_filetype(entry) != AE_IFREG) {
|
||||||
archive_entry_set_size(entry, 0);
|
archive_entry_set_size(entry, 0);
|
||||||
|
} else {
|
||||||
|
archive_entry_set_size(entry, archive_entry_size(entry));
|
||||||
}
|
}
|
||||||
if (pctx->verbose)
|
if (pctx->verbose)
|
||||||
log_msg(LOG_INFO, 0, "%5d/%5d %8d %s", ctr, pctx->archive_members_count,
|
log_msg(LOG_INFO, 0, "%5d/%5d %8d %s", ctr, pctx->archive_members_count,
|
||||||
|
@ -1015,6 +1092,7 @@ archiver_thread_func(void *dat) {
|
||||||
ent = spare_entry;
|
ent = spare_entry;
|
||||||
spare_entry = NULL;
|
spare_entry = NULL;
|
||||||
}
|
}
|
||||||
|
archive_write_finish_entry(arc);
|
||||||
archive_entry_clear(entry);
|
archive_entry_clear(entry);
|
||||||
ctr++;
|
ctr++;
|
||||||
}
|
}
|
||||||
|
@ -1045,13 +1123,29 @@ start_archiver(pc_ctx_t *pctx) {
|
||||||
* routines, so we have to handle here.
|
* routines, so we have to handle here.
|
||||||
*/
|
*/
|
||||||
static int
|
static int
|
||||||
copy_data_out(struct archive *ar, struct archive *aw)
|
copy_data_out(struct archive *ar, struct archive *aw, struct archive_entry *entry,
|
||||||
|
int typ)
|
||||||
{
|
{
|
||||||
int64_t offset;
|
int64_t offset;
|
||||||
const void *buff;
|
const void *buff;
|
||||||
size_t size;
|
size_t size;
|
||||||
int r;
|
int r;
|
||||||
|
|
||||||
|
if (typ != TYPE_UNKNOWN) {
|
||||||
|
if (typetab[(typ >> 3)].filter_func != NULL) {
|
||||||
|
int64_t rv;
|
||||||
|
|
||||||
|
rv = process_by_filter(-1, typ, aw, ar, entry, 0);
|
||||||
|
if (rv == FILTER_RETURN_ERROR) {
|
||||||
|
archive_set_error(ar, archive_errno(aw),
|
||||||
|
"%s", archive_error_string(aw));
|
||||||
|
return (ARCHIVE_FATAL);
|
||||||
|
} else {
|
||||||
|
return (ARCHIVE_OK);
|
||||||
|
}
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
for (;;) {
|
for (;;) {
|
||||||
r = archive_read_data_block(ar, &buff, &size, &offset);
|
r = archive_read_data_block(ar, &buff, &size, &offset);
|
||||||
if (r == ARCHIVE_EOF)
|
if (r == ARCHIVE_EOF)
|
||||||
|
@ -1071,19 +1165,20 @@ copy_data_out(struct archive *ar, struct archive *aw)
|
||||||
|
|
||||||
static int
|
static int
|
||||||
archive_extract_entry(struct archive *a, struct archive_entry *entry,
|
archive_extract_entry(struct archive *a, struct archive_entry *entry,
|
||||||
struct archive *ad)
|
struct archive *ad, int typ)
|
||||||
{
|
{
|
||||||
int r, r2;
|
int r, r2;
|
||||||
|
|
||||||
r = archive_write_header(ad, entry);
|
r = archive_write_header(ad, entry);
|
||||||
if (r < ARCHIVE_WARN)
|
if (r < ARCHIVE_WARN)
|
||||||
r = ARCHIVE_WARN;
|
r = ARCHIVE_WARN;
|
||||||
if (r != ARCHIVE_OK)
|
if (r != ARCHIVE_OK) {
|
||||||
/* If _write_header failed, copy the error. */
|
/* If _write_header failed, copy the error. */
|
||||||
archive_copy_error(a, ad);
|
archive_copy_error(a, ad);
|
||||||
else if (!archive_entry_size_is_set(entry) || archive_entry_size(entry) > 0)
|
} else if (!archive_entry_size_is_set(entry) || archive_entry_size(entry) > 0) {
|
||||||
/* Otherwise, pour data into the entry. */
|
/* Otherwise, pour data into the entry. */
|
||||||
r = copy_data_out(a, ad);
|
r = copy_data_out(a, ad, entry, typ);
|
||||||
|
}
|
||||||
r2 = archive_write_finish_entry(ad);
|
r2 = archive_write_finish_entry(ad);
|
||||||
if (r2 < ARCHIVE_WARN)
|
if (r2 < ARCHIVE_WARN)
|
||||||
r2 = ARCHIVE_WARN;
|
r2 = ARCHIVE_WARN;
|
||||||
|
@ -1153,6 +1248,7 @@ extractor_thread_func(void *dat) {
|
||||||
while ((rv = archive_read_next_header(arc, &entry)) != ARCHIVE_EOF) {
|
while ((rv = archive_read_next_header(arc, &entry)) != ARCHIVE_EOF) {
|
||||||
const char *xt_name, *xt_value;
|
const char *xt_name, *xt_value;
|
||||||
size_t xt_size;
|
size_t xt_size;
|
||||||
|
int typ;
|
||||||
|
|
||||||
if (rv != ARCHIVE_OK)
|
if (rv != ARCHIVE_OK)
|
||||||
log_msg(LOG_WARN, 0, "%s", archive_error_string(arc));
|
log_msg(LOG_WARN, 0, "%s", archive_error_string(arc));
|
||||||
|
@ -1167,6 +1263,12 @@ extractor_thread_func(void *dat) {
|
||||||
continue;
|
continue;
|
||||||
}
|
}
|
||||||
|
|
||||||
|
typ = TYPE_UNKNOWN;
|
||||||
|
if (archive_entry_filetype(entry) == AE_IFREG) {
|
||||||
|
const char *fpath = archive_entry_pathname(entry);
|
||||||
|
typ = detect_type_by_ext(fpath, strlen(fpath));
|
||||||
|
}
|
||||||
|
|
||||||
/*
|
/*
|
||||||
* Workaround for libarchive weirdness on Non MAC OS X platforms for filenames
|
* Workaround for libarchive weirdness on Non MAC OS X platforms for filenames
|
||||||
* starting with '._'. See above ...
|
* starting with '._'. See above ...
|
||||||
|
@ -1191,7 +1293,7 @@ extractor_thread_func(void *dat) {
|
||||||
}
|
}
|
||||||
#endif
|
#endif
|
||||||
|
|
||||||
rv = archive_extract_entry(arc, entry, awd);
|
rv = archive_extract_entry(arc, entry, awd, typ);
|
||||||
if (rv != ARCHIVE_OK) {
|
if (rv != ARCHIVE_OK) {
|
||||||
log_msg(LOG_WARN, 0, "%s: %s", archive_entry_pathname(entry),
|
log_msg(LOG_WARN, 0, "%s: %s", archive_entry_pathname(entry),
|
||||||
archive_error_string(arc));
|
archive_error_string(arc));
|
||||||
|
@ -1251,11 +1353,10 @@ init_archive_mod() {
|
||||||
extnum = (extnum << 1) | extlist[i].ext[j];
|
extnum = (extnum << 1) | extlist[i].ext[j];
|
||||||
exthtab[slot].extnum = extnum;
|
exthtab[slot].extnum = extnum;
|
||||||
exthtab[slot].type = extlist[i].type;
|
exthtab[slot].type = extlist[i].type;
|
||||||
exthtab[slot].filter_func = NULL;
|
|
||||||
exthtab[slot].filter_private = NULL;
|
|
||||||
}
|
}
|
||||||
|
|
||||||
add_filters_by_ext();
|
memset(typetab, 0, sizeof (typetab));
|
||||||
|
add_filters_by_type(typetab);
|
||||||
inited = 1;
|
inited = 1;
|
||||||
} else {
|
} else {
|
||||||
rv = 1;
|
rv = 1;
|
||||||
|
@ -1271,9 +1372,9 @@ init_archive_mod() {
|
||||||
* outside the hash table range then the function returns unknown type.
|
* outside the hash table range then the function returns unknown type.
|
||||||
*/
|
*/
|
||||||
static int
|
static int
|
||||||
detect_type_by_ext(char *path, int pathlen)
|
detect_type_by_ext(const char *path, int pathlen)
|
||||||
{
|
{
|
||||||
char *ext = NULL;
|
const char *ext = NULL;
|
||||||
ub4 slot;
|
ub4 slot;
|
||||||
int i, len;
|
int i, len;
|
||||||
uint64_t extnum;
|
uint64_t extnum;
|
||||||
|
@ -1340,16 +1441,3 @@ detect_type_by_data(uchar_t *buf, size_t len)
|
||||||
|
|
||||||
return (TYPE_UNKNOWN);
|
return (TYPE_UNKNOWN);
|
||||||
}
|
}
|
||||||
|
|
||||||
int
|
|
||||||
insert_filter_data(filter_func_ptr func, void *filter_private, const char *ext)
|
|
||||||
{
|
|
||||||
ub4 slot = phash(ext, strlen(ext));
|
|
||||||
if (slot >= PHASHNKEYS || slot < 0) {
|
|
||||||
log_msg(LOG_WARN, 0, "Cannot add filter for unknown extension: %s", ext);
|
|
||||||
return (-1);
|
|
||||||
}
|
|
||||||
exthtab[slot].filter_func = func;
|
|
||||||
exthtab[slot].filter_private = filter_private;
|
|
||||||
return (0);
|
|
||||||
}
|
|
||||||
|
|
|
@ -39,6 +39,29 @@ extern "C" {
|
||||||
|
|
||||||
typedef unsigned char uchar_t;
|
typedef unsigned char uchar_t;
|
||||||
|
|
||||||
|
/* a short reminder about input/output stream types
|
||||||
|
for the pjglib_init_streams() function
|
||||||
|
|
||||||
|
if input is file
|
||||||
|
----------------
|
||||||
|
in_scr -> name of input file
|
||||||
|
in_type -> 0
|
||||||
|
in_size -> ignore
|
||||||
|
|
||||||
|
if input is memory
|
||||||
|
------------------
|
||||||
|
in_scr -> array containg data
|
||||||
|
in_type -> 1
|
||||||
|
in_size -> size of data array
|
||||||
|
|
||||||
|
if input is *FILE (f.e. stdin)
|
||||||
|
------------------------------
|
||||||
|
in_src -> stream pointer
|
||||||
|
in_type -> 2
|
||||||
|
in_size -> ignore
|
||||||
|
|
||||||
|
vice versa for output streams! */
|
||||||
|
|
||||||
size_t
|
size_t
|
||||||
packjpg_filter_process(uchar_t *in_buf, size_t len, uchar_t **out_buf)
|
packjpg_filter_process(uchar_t *in_buf, size_t len, uchar_t **out_buf)
|
||||||
{
|
{
|
||||||
|
|
|
@ -1544,13 +1544,12 @@ plain_index:
|
||||||
o_chunksize = _chunksize;
|
o_chunksize = _chunksize;
|
||||||
|
|
||||||
/* Compress data chunk. */
|
/* Compress data chunk. */
|
||||||
if ((pctx->lzp_preprocess || pctx->enable_delta2_encode) && _chunksize > 0 &&
|
if ((pctx->lzp_preprocess || pctx->enable_delta2_encode) && _chunksize > 0) {
|
||||||
PC_SUBTYPE(pctx->btype) == TYPE_CMP_MAX) {
|
|
||||||
rv = preproc_compress(pctx, tdat->compress, tdat->uncompressed_chunk + dedupe_index_sz,
|
rv = preproc_compress(pctx, tdat->compress, tdat->uncompressed_chunk + dedupe_index_sz,
|
||||||
_chunksize, compressed_chunk + index_size_cmp, &_chunksize,
|
_chunksize, compressed_chunk + index_size_cmp, &_chunksize,
|
||||||
tdat->level, 0, pctx->btype, tdat->data, tdat->props);
|
tdat->level, 0, pctx->btype, tdat->data, tdat->props);
|
||||||
|
|
||||||
} else if (_chunksize > 0 && PC_SUBTYPE(pctx->btype) == TYPE_CMP_MAX) {
|
} else if (_chunksize > 0) {
|
||||||
DEBUG_STAT_EN(double strt, en);
|
DEBUG_STAT_EN(double strt, en);
|
||||||
|
|
||||||
DEBUG_STAT_EN(strt = get_wtime_millis());
|
DEBUG_STAT_EN(strt = get_wtime_millis());
|
||||||
|
@ -1915,8 +1914,10 @@ start_compress(pc_ctx_t *pctx, const char *filename, uint64_t chunksize, int lev
|
||||||
/*
|
/*
|
||||||
* Adjust chunk size for small files. We then get an archive with
|
* Adjust chunk size for small files. We then get an archive with
|
||||||
* a single chunk for the entire file.
|
* a single chunk for the entire file.
|
||||||
|
* This is not valid for archive mode since we cannot accurately estimate
|
||||||
|
* final archive size.
|
||||||
*/
|
*/
|
||||||
if (sbuf.st_size <= chunksize) {
|
if (sbuf.st_size <= chunksize && !(pctx->archive_mode)) {
|
||||||
chunksize = sbuf.st_size;
|
chunksize = sbuf.st_size;
|
||||||
pctx->enable_rabin_split = 0; // Do not split for whole files.
|
pctx->enable_rabin_split = 0; // Do not split for whole files.
|
||||||
pctx->nthreads = 1;
|
pctx->nthreads = 1;
|
||||||
|
|
|
@ -99,8 +99,8 @@ struct ext_entry {
|
||||||
{"xpi" , TYPE_BINARY|TYPE_EXE, 3},
|
{"xpi" , TYPE_BINARY|TYPE_EXE, 3},
|
||||||
{"off" , TYPE_BINARY|TYPE_EXE, 3},
|
{"off" , TYPE_BINARY|TYPE_EXE, 3},
|
||||||
{"pdf" , TYPE_BINARY, 3},
|
{"pdf" , TYPE_BINARY, 3},
|
||||||
{"jpg" , TYPE_BINARY|TYPE_COMPRESSED|TYPE_CMP_MAX, 3},
|
{"jpg" , TYPE_BINARY|TYPE_COMPRESSED|TYPE_JPEG, 3},
|
||||||
{"jpeg" , TYPE_BINARY|TYPE_COMPRESSED|TYPE_CMP_MAX, 4},
|
{"jpeg" , TYPE_BINARY|TYPE_COMPRESSED|TYPE_JPEG, 4},
|
||||||
{"png" , TYPE_BINARY|TYPE_COMPRESSED|TYPE_COMPRESSED_GZ, 3},
|
{"png" , TYPE_BINARY|TYPE_COMPRESSED|TYPE_COMPRESSED_GZ, 3},
|
||||||
{"mp3" , TYPE_BINARY|TYPE_COMPRESSED, 3},
|
{"mp3" , TYPE_BINARY|TYPE_COMPRESSED, 3},
|
||||||
{"wma" , TYPE_BINARY|TYPE_COMPRESSED, 3},
|
{"wma" , TYPE_BINARY|TYPE_COMPRESSED, 3},
|
||||||
|
@ -114,8 +114,8 @@ struct ext_entry {
|
||||||
{"flac" , TYPE_BINARY|TYPE_COMPRESSED, 4},
|
{"flac" , TYPE_BINARY|TYPE_COMPRESSED, 4},
|
||||||
{"pac" , TYPE_BINARY|TYPE_COMPRESSED, 3},
|
{"pac" , TYPE_BINARY|TYPE_COMPRESSED, 3},
|
||||||
{"gif" , TYPE_BINARY|TYPE_COMPRESSED|TYPE_COMPRESSED_LZW, 3},
|
{"gif" , TYPE_BINARY|TYPE_COMPRESSED|TYPE_COMPRESSED_LZW, 3},
|
||||||
{"jp2" , TYPE_BINARY|TYPE_COMPRESSED|TYPE_CMP_MAX, 3},
|
{"jp2" , TYPE_BINARY|TYPE_COMPRESSED|TYPE_JPEG, 3},
|
||||||
{"pjg" , TYPE_BINARY|TYPE_COMPRESSED|TYPE_CMP_MAX, 3},
|
{"pjg" , TYPE_BINARY|TYPE_COMPRESSED|TYPE_PACKJPG, 3},
|
||||||
{"gz" , TYPE_BINARY|TYPE_COMPRESSED|TYPE_COMPRESSED_GZ, 2},
|
{"gz" , TYPE_BINARY|TYPE_COMPRESSED|TYPE_COMPRESSED_GZ, 2},
|
||||||
{"tgz" , TYPE_BINARY|TYPE_COMPRESSED|TYPE_COMPRESSED_GZ, 3},
|
{"tgz" , TYPE_BINARY|TYPE_COMPRESSED|TYPE_COMPRESSED_GZ, 3},
|
||||||
{"bz2" , TYPE_BINARY|TYPE_COMPRESSED|TYPE_COMPRESSED_BZ2, 3},
|
{"bz2" , TYPE_BINARY|TYPE_COMPRESSED|TYPE_COMPRESSED_BZ2, 3},
|
||||||
|
|
|
@ -89,8 +89,8 @@ com,TYPE_BINARY|TYPE_EXE
|
||||||
xpi,TYPE_BINARY|TYPE_EXE
|
xpi,TYPE_BINARY|TYPE_EXE
|
||||||
off,TYPE_BINARY|TYPE_EXE
|
off,TYPE_BINARY|TYPE_EXE
|
||||||
pdf,TYPE_BINARY
|
pdf,TYPE_BINARY
|
||||||
jpg,TYPE_BINARY|TYPE_COMPRESSED|TYPE_CMP_MAX
|
jpg,TYPE_BINARY|TYPE_COMPRESSED|TYPE_JPEG
|
||||||
jpeg,TYPE_BINARY|TYPE_COMPRESSED|TYPE_CMP_MAX
|
jpeg,TYPE_BINARY|TYPE_COMPRESSED|TYPE_JPEG
|
||||||
png,TYPE_BINARY|TYPE_COMPRESSED|TYPE_COMPRESSED_GZ
|
png,TYPE_BINARY|TYPE_COMPRESSED|TYPE_COMPRESSED_GZ
|
||||||
mp3,TYPE_BINARY|TYPE_COMPRESSED
|
mp3,TYPE_BINARY|TYPE_COMPRESSED
|
||||||
wma,TYPE_BINARY|TYPE_COMPRESSED
|
wma,TYPE_BINARY|TYPE_COMPRESSED
|
||||||
|
@ -104,8 +104,8 @@ ofr,TYPE_BINARY|TYPE_COMPRESSED
|
||||||
flac,TYPE_BINARY|TYPE_COMPRESSED
|
flac,TYPE_BINARY|TYPE_COMPRESSED
|
||||||
pac,TYPE_BINARY|TYPE_COMPRESSED
|
pac,TYPE_BINARY|TYPE_COMPRESSED
|
||||||
gif,TYPE_BINARY|TYPE_COMPRESSED|TYPE_COMPRESSED_LZW
|
gif,TYPE_BINARY|TYPE_COMPRESSED|TYPE_COMPRESSED_LZW
|
||||||
jp2,TYPE_BINARY|TYPE_COMPRESSED|TYPE_CMP_MAX
|
jp2,TYPE_BINARY|TYPE_COMPRESSED|TYPE_JPEG
|
||||||
pjg,TYPE_BINARY|TYPE_COMPRESSED|TYPE_CMP_MAX
|
pjg,TYPE_BINARY|TYPE_COMPRESSED|TYPE_PACKJPG
|
||||||
gz,TYPE_BINARY|TYPE_COMPRESSED|TYPE_COMPRESSED_GZ
|
gz,TYPE_BINARY|TYPE_COMPRESSED|TYPE_COMPRESSED_GZ
|
||||||
tgz,TYPE_BINARY|TYPE_COMPRESSED|TYPE_COMPRESSED_GZ
|
tgz,TYPE_BINARY|TYPE_COMPRESSED|TYPE_COMPRESSED_GZ
|
||||||
bz2,TYPE_BINARY|TYPE_COMPRESSED|TYPE_COMPRESSED_BZ2
|
bz2,TYPE_BINARY|TYPE_COMPRESSED|TYPE_COMPRESSED_BZ2
|
||||||
|
|
|
@ -242,8 +242,9 @@ typedef enum {
|
||||||
/*
|
/*
|
||||||
* Sub-types.
|
* Sub-types.
|
||||||
*/
|
*/
|
||||||
|
#define NUM_SUB_TYPES 20
|
||||||
TYPE_EXE = 8,
|
TYPE_EXE = 8,
|
||||||
TYPE_CMP_MAX = 16,
|
TYPE_JPEG = 16,
|
||||||
TYPE_MARKUP = 24,
|
TYPE_MARKUP = 24,
|
||||||
TYPE_COMPRESSED_GZ = 32,
|
TYPE_COMPRESSED_GZ = 32,
|
||||||
TYPE_COMPRESSED_LZW = 40,
|
TYPE_COMPRESSED_LZW = 40,
|
||||||
|
@ -260,7 +261,8 @@ typedef enum {
|
||||||
TYPE_COMPRESSED_RAR = 128,
|
TYPE_COMPRESSED_RAR = 128,
|
||||||
TYPE_COMPRESSED_LZ = 136,
|
TYPE_COMPRESSED_LZ = 136,
|
||||||
TYPE_COMPRESSED_PPMD = 144,
|
TYPE_COMPRESSED_PPMD = 144,
|
||||||
TYPE_COMPRESSED_ZPAQ = 152
|
TYPE_COMPRESSED_ZPAQ = 152,
|
||||||
|
TYPE_PACKJPG = 160
|
||||||
} data_type_t;
|
} data_type_t;
|
||||||
|
|
||||||
/*
|
/*
|
||||||
|
|
Loading…
Reference in a new issue