/* Functions for dealing with sparse files
- Copyright (C) 2003, 2004, 2005, 2006 Free Software Foundation, Inc.
+ Copyright 2003-2007, 2010, 2013-2014 Free Software Foundation, Inc.
This program is free software; you can redistribute it and/or modify it
under the terms of the GNU General Public License as published by the
- Free Software Foundation; either version 2, or (at your option) any later
+ Free Software Foundation; either version 3, or (at your option) any later
version.
This program is distributed in the hope that it will be useful, but
Public License for more details.
You should have received a copy of the GNU General Public License along
- with this program; if not, write to the Free Software Foundation, Inc.,
- 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
+ with this program. If not, see <http://www.gnu.org/licenses/>. */
#include <system.h>
#include <inttostr.h>
struct tar_stat_info *st = file->stat_info;
int fd = file->fd;
char buffer[BLOCKSIZE];
- size_t count;
+ size_t count = 0;
off_t offset = 0;
struct sp_array sp = {0, 0};
- if (!lseek_or_error (file, 0))
- return false;
-
st->archive_file_size = 0;
-
- if (!tar_sparse_scan (file, scan_begin, NULL))
- return false;
- while ((count = safe_read (fd, buffer, sizeof buffer)) != 0
- && count != SAFE_READ_ERROR)
+ if (ST_NBLOCKS (st->stat) == 0)
+ offset = st->stat.st_size;
+ else
{
- /* Analyze the block. */
- if (zero_block_p (buffer, count))
+ if (!tar_sparse_scan (file, scan_begin, NULL))
+ return false;
+
+ while ((count = blocking_read (fd, buffer, sizeof buffer)) != 0
+ && count != SAFE_READ_ERROR)
{
- if (sp.numbytes)
+ /* Analyze the block. */
+ if (zero_block_p (buffer, count))
+ {
+ if (sp.numbytes)
+ {
+ sparse_add_map (st, &sp);
+ sp.numbytes = 0;
+ if (!tar_sparse_scan (file, scan_block, NULL))
+ return false;
+ }
+ }
+ else
{
- sparse_add_map (st, &sp);
- sp.numbytes = 0;
- if (!tar_sparse_scan (file, scan_block, NULL))
+ if (sp.numbytes == 0)
+ sp.offset = offset;
+ sp.numbytes += count;
+ st->archive_file_size += count;
+ if (!tar_sparse_scan (file, scan_block, buffer))
return false;
}
- }
- else
- {
- if (sp.numbytes == 0)
- sp.offset = offset;
- sp.numbytes += count;
- st->archive_file_size += count;
- if (!tar_sparse_scan (file, scan_block, buffer))
- return false;
- }
- offset += count;
+ offset += count;
+ }
}
if (sp.numbytes == 0)
memset (blk->buffer + bytes_read, 0, BLOCKSIZE - bytes_read);
bytes_left -= bytes_read;
file->dumped_size += bytes_read;
- mv_size_left (file->stat_info->archive_file_size - file->dumped_size);
set_next_block_after (blk);
}
static bool
sparse_extract_region (struct tar_sparse_file *file, size_t i)
{
- size_t write_size;
+ off_t write_size;
if (!lseek_or_error (file, file->stat_info->sparse_map[i].offset))
return false;
return false;
}
set_next_block_after (blk);
- count = full_write (file->fd, blk->buffer, wrbytes);
+ count = blocking_write (file->fd, blk->buffer, wrbytes);
write_size -= count;
file->dumped_size += count;
mv_size_left (file->stat_info->archive_file_size - file->dumped_size);
{
size_t i;
- mv_begin (file.stat_info);
+ mv_begin_write (file.stat_info->file_name,
+ file.stat_info->stat.st_size,
+ file.stat_info->archive_file_size - file.dumped_size);
for (i = 0; rc && i < file.stat_info->sparse_map_avail; i++)
rc = tar_sparse_dump_region (&file, i);
- mv_end ();
}
}
return (tar_sparse_done (&file) && rc) ? dump_status_ok : dump_status_short;
}
-/* Returns true if the file represented by stat is a sparse one */
-bool
-sparse_file_p (struct tar_stat_info *st)
-{
- return (ST_NBLOCKS (st->stat)
- < (st->stat.st_size / ST_NBLOCKSIZE
- + (st->stat.st_size % ST_NBLOCKSIZE != 0)));
-}
-
bool
sparse_member_p (struct tar_stat_info *st)
{
file.fd = -1;
rc = tar_sparse_decode_header (&file);
- skip_file (file.stat_info->archive_file_size);
+ skip_file (file.stat_info->archive_file_size - file.dumped_size);
return (tar_sparse_done (&file) && rc) ? dump_status_ok : dump_status_short;
}
static bool
check_data_region (struct tar_sparse_file *file, size_t i)
{
- size_t size_left;
+ off_t size_left;
if (!lseek_or_error (file, file->stat_info->sparse_map[i].offset))
return false;
size_left = file->stat_info->sparse_map[i].numbytes;
mv_size_left (file->stat_info->archive_file_size - file->dumped_size);
-
+
while (size_left > 0)
{
size_t bytes_read;
file.stat_info = st;
file.fd = fd;
file.seekable = true; /* File *must* be seekable for compare to work */
-
+
rc = tar_sparse_decode_header (&file);
- mv_begin (st);
+ mv_begin_read (st);
for (i = 0; rc && i < file.stat_info->sparse_map_avail; i++)
{
rc = check_sparse_region (&file,
if (!rc)
skip_file (file.stat_info->archive_file_size - file.dumped_size);
mv_end ();
-
+
tar_sparse_done (&file);
return rc;
}
/* Old GNU Format. The sparse file information is stored in the
oldgnu_header in the following manner:
- The header is marked with type 'S'. Its `size' field contains
+ The header is marked with type 'S'. Its 'size' field contains
the cumulative size of all non-empty blocks of the file. The
- actual file size is stored in `realsize' member of oldgnu_header.
+ actual file size is stored in 'realsize' member of oldgnu_header.
- The map of the file is stored in a list of `struct sparse'.
+ The map of the file is stored in a list of 'struct sparse'.
Each struct contains offset to the block of data and its
size (both as octal numbers). The first file header contains
at most 4 such structs (SPARSES_IN_OLDGNU_HEADER). If the map
- contains more structs, then the field `isextended' of the main
- header is set to 1 (binary) and the `struct sparse_header'
+ contains more structs, then the field 'isextended' of the main
+ header is set to 1 (binary) and the 'struct sparse_header'
header follows, containing at most 21 following structs
- (SPARSES_IN_SPARSE_HEADER). If more structs follow, `isextended'
+ (SPARSES_IN_SPARSE_HEADER). If more structs follow, 'isextended'
field of the extended header is set and next next extension header
follows, etc... */
if (s->numbytes[0] == '\0')
return add_finish;
sp.offset = OFF_FROM_HEADER (s->offset);
- sp.numbytes = SIZE_FROM_HEADER (s->numbytes);
- if (sp.offset < 0
+ sp.numbytes = OFF_FROM_HEADER (s->numbytes);
+ if (sp.offset < 0 || sp.numbytes < 0
+ || INT_ADD_OVERFLOW (sp.offset, sp.numbytes)
|| file->stat_info->stat.st_size < sp.offset + sp.numbytes
|| file->stat_info->archive_file_size < 0)
return add_fail;
{
/* NOTE! st_size was initialized from the header
which actually contains archived size. The following fixes it */
+ off_t realsize = OFF_FROM_HEADER (current_header->oldgnu_header.realsize);
file->stat_info->archive_file_size = file->stat_info->stat.st_size;
- file->stat_info->stat.st_size =
- OFF_FROM_HEADER (current_header->oldgnu_header.realsize);
- return true;
+ file->stat_info->stat.st_size = max (0, realsize);
+ return 0 <= realsize;
}
/* Convert old GNU format sparse data to internal representation */
{
OFF_TO_CHARS (file->stat_info->sparse_map[*pindex].offset,
sp->offset);
- SIZE_TO_CHARS (file->stat_info->sparse_map[*pindex].numbytes,
- sp->numbytes);
+ OFF_TO_CHARS (file->stat_info->sparse_map[*pindex].numbytes,
+ sp->numbytes);
}
}
{
/* NOTE! st_size was initialized from the header
which actually contains archived size. The following fixes it */
+ off_t realsize = OFF_FROM_HEADER (current_header->star_in_header.realsize);
file->stat_info->archive_file_size = file->stat_info->stat.st_size;
- file->stat_info->stat.st_size =
- OFF_FROM_HEADER (current_header->star_in_header.realsize);
- return true;
+ file->stat_info->stat.st_size = max (0, realsize);
+ return 0 <= realsize;
}
/* Convert STAR format sparse data to internal representation */
set_next_block_after (h);
for (i = 0; i < SPARSES_IN_STAR_EXT_HEADER && rc == add_ok; i++)
rc = oldgnu_add_sparse (file, &h->star_ext_header.sp[i]);
+ file->dumped_size += BLOCKSIZE;
}
if (rc == add_fail)
\f
/* GNU PAX sparse file format. There are several versions:
-
* 0.0
The initial version of sparse format used by tar 1.14-1.15.1.
instances of GNU.sparse.offset/GNU.sparse.numbytes variables, whereas
POSIX requires the latest occurrence of the variable to override all
previous occurrences.
-
+
To avoid this incompatibility two following versions were introduced.
* 0.1
Used by tar 1.15.2 -- 1.15.91 (alpha releases).
-
+
The sparse file map is stored in
x header:
GNU.sparse.size Real size of the stored file
GNU.sparse.numblocks Number of blocks in the sparse map
GNU.sparse.map Map of non-null data chunks. A string consisting
- of comma-separated values "offset,size[,offset,size]..."
+ of comma-separated values "offset,size[,offset,size]..."
The resulting GNU.sparse.map string can be *very* long. While POSIX does not
impose any limit on the length of a x header variable, this can confuse some
* 1.0
- Starting from this version, the exact sparse format version is specified explicitely
- in the header using the following variables:
+ Starting from this version, the exact sparse format version is specified
+ explicitely in the header using the following variables:
- GNU.sparse.major Major version
+ GNU.sparse.major Major version
GNU.sparse.minor Minor version
X header keeps the following variables:
-
+
GNU.sparse.name Real file name of the sparse file
GNU.sparse.realsize Real size of the stored file (corresponds to the old
GNU.sparse.size variable)
The name field of the ustar header is constructed using the pattern
"%d/GNUSparseFile.%p/%f".
-
+
The sparse map itself is stored in the file data block, preceding the actual
- file data. It consists of a series of octal numbers of arbitrary length, delimited
- by newlines. The map is padded with nulls to the nearest block boundary.
+ file data. It consists of a series of octal numbers of arbitrary length,
+ delimited by newlines. The map is padded with nulls to the nearest block
+ boundary.
- The first number gives the number of entries in the map. Following are map entries,
- each one consisting of two numbers giving the offset and size of the
- data block it describes.
+ The first number gives the number of entries in the map. Following are map
+ entries, each one consisting of two numbers giving the offset and size of
+ the data block it describes.
The format is designed in such a way that non-posix aware tars and tars not
- supporting GNU.sparse.* keywords will extract each sparse file in its condensed
- form with the file map attached and will place it into a separate directory.
- Then, using a simple program it would be possible to expand the file to its
- original form even without GNU tar.
-
- Bu default, v.1.0 archives are created. To use other formats, --sparse-version
- option is provided. Additionally, v.0.0 can be obtained by deleting GNU.sparse.map
- from 0.1 format: --sparse-version 0.1 --pax-option delete=GNU.sparse.map
+ supporting GNU.sparse.* keywords will extract each sparse file in its
+ condensed form with the file map attached and will place it into a separate
+ directory. Then, using a simple program it would be possible to expand the
+ file to its original form even without GNU tar.
+
+ Bu default, v.1.0 archives are created. To use other formats,
+ --sparse-version option is provided. Additionally, v.0.0 can be obtained
+ by deleting GNU.sparse.map from 0.1 format: --sparse-version 0.1
+ --pax-option delete=GNU.sparse.map
*/
static bool
|| file->stat_info->sparse_major > 0;
}
+/* Start a header that uses the effective (shrunken) file size. */
+static union block *
+pax_start_header (struct tar_stat_info *st)
+{
+ off_t realsize = st->stat.st_size;
+ union block *blk;
+ st->stat.st_size = st->archive_file_size;
+ blk = start_header (st);
+ st->stat.st_size = realsize;
+ return blk;
+}
+
static bool
pax_dump_header_0 (struct tar_sparse_file *file)
{
char nbuf[UINTMAX_STRSIZE_BOUND];
struct sp_array *map = file->stat_info->sparse_map;
char *save_file_name = NULL;
-
+
/* Store the real file size */
xheader_store ("GNU.sparse.size", file->stat_info, NULL);
xheader_store ("GNU.sparse.numblocks", file->stat_info, NULL);
-
+
if (xheader_keyword_deleted_p ("GNU.sparse.map")
|| tar_sparse_minor == 0)
{
file->stat_info->file_name = xheader_format_name (file->stat_info,
"%d/GNUSparseFile.%p/%f", 0);
- xheader_string_begin ();
+ xheader_string_begin (&file->stat_info->xhdr);
for (i = 0; i < file->stat_info->sparse_map_avail; i++)
{
if (i)
- xheader_string_add (",");
- xheader_string_add (umaxtostr (map[i].offset, nbuf));
- xheader_string_add (",");
- xheader_string_add (umaxtostr (map[i].numbytes, nbuf));
+ xheader_string_add (&file->stat_info->xhdr, ",");
+ xheader_string_add (&file->stat_info->xhdr,
+ umaxtostr (map[i].offset, nbuf));
+ xheader_string_add (&file->stat_info->xhdr, ",");
+ xheader_string_add (&file->stat_info->xhdr,
+ umaxtostr (map[i].numbytes, nbuf));
+ }
+ if (!xheader_string_end (&file->stat_info->xhdr,
+ "GNU.sparse.map"))
+ {
+ free (file->stat_info->file_name);
+ file->stat_info->file_name = save_file_name;
+ return false;
}
- xheader_string_end ("GNU.sparse.map");
}
- blk = start_header (file->stat_info);
- /* Store the effective (shrunken) file size */
- OFF_TO_CHARS (file->stat_info->archive_file_size, blk->header.size);
+ blk = pax_start_header (file->stat_info);
finish_header (file->stat_info, blk, block_ordinal);
if (save_file_name)
{
#define COPY_STRING(b,dst,src) do \
{ \
char *endp = b->buffer + BLOCKSIZE; \
- char *srcp = src; \
+ char const *srcp = src; \
while (*srcp) \
{ \
if (dst == endp) \
} \
*dst++ = *srcp++; \
} \
- } while (0)
+ } while (0)
/* Compute stored file size */
p = umaxtostr (file->stat_info->sparse_map_avail, nbuf);
}
size = (size + BLOCKSIZE - 1) / BLOCKSIZE;
file->stat_info->archive_file_size += size * BLOCKSIZE;
-
+ file->dumped_size += size * BLOCKSIZE;
+
/* Store sparse file identification */
xheader_store ("GNU.sparse.major", file->stat_info, NULL);
xheader_store ("GNU.sparse.minor", file->stat_info, NULL);
xheader_store ("GNU.sparse.name", file->stat_info, NULL);
xheader_store ("GNU.sparse.realsize", file->stat_info, NULL);
-
- file->stat_info->file_name = xheader_format_name (file->stat_info,
- "%d/GNUSparseFile.%p/%f", 0);
- blk = start_header (file->stat_info);
- /* Store the effective (shrunken) file size */
- OFF_TO_CHARS (file->stat_info->archive_file_size, blk->header.size);
+ file->stat_info->file_name =
+ xheader_format_name (file->stat_info, "%d/GNUSparseFile.%p/%f", 0);
+ /* Make sure the created header name is shorter than NAME_FIELD_SIZE: */
+ if (strlen (file->stat_info->file_name) > NAME_FIELD_SIZE)
+ file->stat_info->file_name[NAME_FIELD_SIZE] = 0;
+
+ blk = pax_start_header (file->stat_info);
finish_header (file->stat_info, blk, block_ordinal);
free (file->stat_info->file_name);
file->stat_info->file_name = save_file_name;
{
file->stat_info->sparse_major = tar_sparse_major;
file->stat_info->sparse_minor = tar_sparse_minor;
-
- if (file->stat_info->sparse_major == 0)
- pax_dump_header_0 (file);
- else
- pax_dump_header_1 (file);
+
+ return (file->stat_info->sparse_major == 0) ?
+ pax_dump_header_0 (file) : pax_dump_header_1 (file);
}
static bool
if (!ISDIGIT (*arg))
return false;
-
+
+ errno = 0;
u = strtoumax (arg, &arg_lim, 10);
if (! (u <= maxval && errno != ERANGE) || *arg_lim)
return false;
-
+
*num = u;
return true;
}
{ \
if (dst == buf + UINTMAX_STRSIZE_BOUND -1) \
{ \
- ERROR ((0, 0, _("%s: numeric overflow in sparse archive member"), \
+ ERROR ((0, 0, _("%s: numeric overflow in sparse archive member"), \
file->stat_info->orig_file_name)); \
return false; \
} \
- if (src == endp) \
+ if (src == endp) \
{ \
set_next_block_after (b); \
- b = find_next_block (); \
+ file->dumped_size += BLOCKSIZE; \
+ b = find_next_block (); \
src = b->buffer; \
endp = b->buffer + BLOCKSIZE; \
} \
} \
while (*dst++ != '\n'); \
dst[-1] = 0; \
- } while (0)
+ } while (0)
set_next_block_after (current_header);
+ file->dumped_size += BLOCKSIZE;
blk = find_next_block ();
p = blk->buffer;
COPY_BUF (blk,nbuf,p);
if (!decode_num (&u, nbuf, TYPE_MAXIMUM (size_t)))
{
- ERROR ((0, 0, _("%s: malformed sparse archive member"),
+ ERROR ((0, 0, _("%s: malformed sparse archive member"),
file->stat_info->orig_file_name));
return false;
}
for (i = 0; i < file->stat_info->sparse_map_size; i++)
{
struct sp_array sp;
-
+
COPY_BUF (blk,nbuf,p);
if (!decode_num (&u, nbuf, TYPE_MAXIMUM (off_t)))
{
- ERROR ((0, 0, _("%s: malformed sparse archive member"),
+ ERROR ((0, 0, _("%s: malformed sparse archive member"),
file->stat_info->orig_file_name));
return false;
}
sp.offset = u;
COPY_BUF (blk,nbuf,p);
- if (!decode_num (&u, nbuf, TYPE_MAXIMUM (size_t)))
+ if (!decode_num (&u, nbuf, TYPE_MAXIMUM (off_t)))
{
- ERROR ((0, 0, _("%s: malformed sparse archive member"),
+ ERROR ((0, 0, _("%s: malformed sparse archive member"),
file->stat_info->orig_file_name));
return false;
}
}
set_next_block_after (blk);
}
-
+
return true;
}
pax_sparse_member_p,
pax_dump_header,
NULL,
- pax_decode_header,
+ pax_decode_header,
NULL, /* No scan_block function */
sparse_dump_region,
sparse_extract_region,
projects / chaz / tar / blobdiff