-/* List a tar archive.
- Copyright (C) 1988, 1992, 1993 Free Software Foundation
-
-This file is part of GNU Tar.
-
-GNU Tar 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 version.
-
-GNU Tar is distributed in the hope that it will be useful,
-but WITHOUT ANY WARRANTY; without even the implied warranty of
-MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-GNU General Public License for more details.
-
-You should have received a copy of the GNU General Public License
-along with GNU Tar; see the file COPYING. If not, write to
-the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. */
-
-/*
- * List a tar archive.
- *
- * Also includes support routines for reading a tar archive.
- *
- * this version written 26 Aug 1985 by John Gilmore (ihnp4!hoptoad!gnu).
- */
-
-#include <stdio.h>
-#include <ctype.h>
-#include <sys/types.h>
-#include <errno.h>
-#ifndef STDC_HEADERS
-extern int errno;
-#endif
-#include <time.h>
+/* List a tar archive, with support routines for reading a tar archive.
+
+ Copyright 1988, 1992-1994, 1996-2001, 2003-2007, 2010, 2012-2014 Free
+ Software Foundation, Inc.
+
+ This file is part of GNU tar.
+
+ GNU tar 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 3 of the License, or
+ (at your option) any later version.
+
+ GNU tar is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>.
+
+ Written by John Gilmore, on 1985-08-26. */
+
+#include <system.h>
+#include <inttostr.h>
+#include <quotearg.h>
+
+#include "common.h"
+
+union block *current_header; /* points to current archive header */
+enum archive_format current_format; /* recognized format */
+union block *recent_long_name; /* recent long name header and contents */
+union block *recent_long_link; /* likewise, for long link */
+size_t recent_long_name_blocks; /* number of blocks in recent_long_name */
+size_t recent_long_link_blocks; /* likewise, for long link */
+static union block *recent_global_header; /* Recent global header block */
+
+#define GID_FROM_HEADER(where) gid_from_header (where, sizeof (where))
+#define MAJOR_FROM_HEADER(where) major_from_header (where, sizeof (where))
+#define MINOR_FROM_HEADER(where) minor_from_header (where, sizeof (where))
+#define MODE_FROM_HEADER(where, hbits) \
+ mode_from_header (where, sizeof (where), hbits)
+#define TIME_FROM_HEADER(where) time_from_header (where, sizeof (where))
+#define UID_FROM_HEADER(where) uid_from_header (where, sizeof (where))
+
+static gid_t gid_from_header (const char *buf, size_t size);
+static major_t major_from_header (const char *buf, size_t size);
+static minor_t minor_from_header (const char *buf, size_t size);
+static mode_t mode_from_header (const char *buf, size_t size, bool *hbits);
+static time_t time_from_header (const char *buf, size_t size);
+static uid_t uid_from_header (const char *buf, size_t size);
+static intmax_t from_header (const char *, size_t, const char *,
+ intmax_t, uintmax_t, bool, bool);
+
+/* Base 64 digits; see Internet RFC 2045 Table 1. */
+static char const base_64_digits[64] =
+{
+ 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J', 'K', 'L', 'M',
+ 'N', 'O', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z',
+ 'a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j', 'k', 'l', 'm',
+ 'n', 'o', 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', 'x', 'y', 'z',
+ '0', '1', '2', '3', '4', '5', '6', '7', '8', '9', '+', '/'
+};
+
+/* Table of base-64 digit values indexed by unsigned chars.
+ The value is 64 for unsigned chars that are not base-64 digits. */
+static char base64_map[UCHAR_MAX + 1];
+
+static void
+base64_init (void)
+{
+ int i;
+ memset (base64_map, 64, sizeof base64_map);
+ for (i = 0; i < 64; i++)
+ base64_map[(int) base_64_digits[i]] = i;
+}
-#ifdef BSD42
-#include <sys/file.h>
-#else
-#ifndef V7
-#include <fcntl.h>
-#endif
-#endif
+static char *
+decode_xform (char *file_name, void *data)
+{
+ int type = *(int*)data;
+
+ switch (type)
+ {
+ case XFORM_SYMLINK:
+ /* FIXME: It is not quite clear how and to which extent are the symbolic
+ links subject to filename transformation. In the absence of another
+ solution, symbolic links are exempt from component stripping and
+ name suffix normalization, but subject to filename transformation
+ proper. */
+ return file_name;
+
+ case XFORM_LINK:
+ file_name = safer_name_suffix (file_name, true, absolute_names_option);
+ break;
+
+ case XFORM_REGFILE:
+ file_name = safer_name_suffix (file_name, false, absolute_names_option);
+ break;
+ }
+
+ if (strip_name_components)
+ {
+ size_t prefix_len = stripped_prefix_len (file_name,
+ strip_name_components);
+ if (prefix_len == (size_t) -1)
+ prefix_len = strlen (file_name);
+ file_name += prefix_len;
+ }
+ return file_name;
+}
+
+static bool
+transform_member_name (char **pinput, int type)
+{
+ return transform_name_fp (pinput, type, decode_xform, &type);
+}
-#define isodigit(c) ( ((c) >= '0') && ((c) <= '7') )
+static void
+enforce_one_top_level (char **pfile_name)
+{
+ char *file_name = *pfile_name;
+ char *p;
+
+ for (p = file_name; *p && (ISSLASH (*p) || *p == '.'); p++)
+ ;
-#include "tar.h"
-#include "port.h"
+ if (!*p)
+ return;
-extern FILE *msg_file;
+ if (strncmp (p, one_top_level_dir, strlen (one_top_level_dir)) == 0)
+ {
+ int pos = strlen (one_top_level_dir);
+ if (ISSLASH (p[pos]) || p[pos] == 0)
+ return;
+ }
-long from_oct (); /* Decode octal number */
-void demode (); /* Print file mode */
-void restore_saved_dir_info ();
-PTR ck_malloc ();
+ *pfile_name = new_name (one_top_level_dir, file_name);
+ normalize_filename_x (*pfile_name);
+ free (file_name);
+}
-union record *head; /* Points to current archive header */
-struct stat hstat; /* Stat struct corresponding */
-int head_standard; /* Tape header is in ANSI format */
+void
+transform_stat_info (int typeflag, struct tar_stat_info *stat_info)
+{
+ if (typeflag == GNUTYPE_VOLHDR)
+ /* Name transformations don't apply to volume headers. */
+ return;
-int check_exclude ();
-void close_archive ();
-void decode_header ();
-int findgid ();
-int finduid ();
-void name_gather ();
-int name_match ();
-void names_notfound ();
-void open_archive ();
-void print_header ();
-int read_header ();
-void saverec ();
-void skip_file ();
-void skip_extended_headers ();
+ transform_member_name (&stat_info->file_name, XFORM_REGFILE);
+ switch (typeflag)
+ {
+ case SYMTYPE:
+ transform_member_name (&stat_info->link_name, XFORM_SYMLINK);
+ break;
-extern char *quote_copy_string ();
+ case LNKTYPE:
+ transform_member_name (&stat_info->link_name, XFORM_LINK);
+ }
+ if (one_top_level_option)
+ enforce_one_top_level (¤t_stat_info.file_name);
+}
-/*
- * Main loop for reading an archive.
- */
+/* Main loop for reading an archive. */
void
-read_and (do_something)
- void (*do_something) ();
+read_and (void (*do_something) (void))
{
- int status = 3; /* Initial status at start of archive */
- int prev_status;
- extern time_t new_time;
- char save_linkflag;
+ enum read_header status = HEADER_STILL_UNREAD;
+ enum read_header prev_status;
+ struct timespec mtime;
- name_gather (); /* Gather all the names */
- open_archive (1); /* Open for reading */
+ base64_init ();
+ name_gather ();
- for (;;)
+ open_archive (ACCESS_READ);
+ do
{
prev_status = status;
- status = read_header ();
+ tar_stat_destroy (¤t_stat_info);
+
+ status = read_header (¤t_header, ¤t_stat_info,
+ read_header_auto);
switch (status)
{
-
- case 1: /* Valid header */
- /* We should decode next field (mode) first... */
- /* Ensure incoming names are null terminated. */
-
- if (!name_match (current_file_name)
- || (f_new_files && hstat.st_mtime < new_time)
- || (f_exclude && check_exclude (current_file_name)))
+ case HEADER_STILL_UNREAD:
+ case HEADER_SUCCESS_EXTENDED:
+ abort ();
+
+ case HEADER_SUCCESS:
+
+ /* Valid header. We should decode next field (mode) first.
+ Ensure incoming names are null terminated. */
+ decode_header (current_header, ¤t_stat_info,
+ ¤t_format, 1);
+ if (! name_match (current_stat_info.file_name)
+ || (NEWER_OPTION_INITIALIZED (newer_mtime_option)
+ /* FIXME: We get mtime now, and again later; this causes
+ duplicate diagnostics if header.mtime is bogus. */
+ && ((mtime.tv_sec
+ = TIME_FROM_HEADER (current_header->header.mtime)),
+ /* FIXME: Grab fractional time stamps from
+ extended header. */
+ mtime.tv_nsec = 0,
+ current_stat_info.mtime = mtime,
+ OLDER_TAR_STAT_TIME (current_stat_info, m)))
+ || excluded_name (current_stat_info.file_name))
{
-
- int isextended = 0;
-
- if (head->header.linkflag == LF_VOLHDR
- || head->header.linkflag == LF_MULTIVOL
- || head->header.linkflag == LF_NAMES)
+ switch (current_header->header.typeflag)
{
- (*do_something) ();
+ case GNUTYPE_VOLHDR:
+ case GNUTYPE_MULTIVOL:
+ break;
+
+ case DIRTYPE:
+ if (show_omitted_dirs_option)
+ WARN ((0, 0, _("%s: Omitting"),
+ quotearg_colon (current_stat_info.file_name)));
+ /* Fall through. */
+ default:
+ skip_member ();
continue;
}
- if (f_show_omitted_dirs
- && head->header.linkflag == LF_DIR)
- msg ("Omitting %s\n", current_file_name);
- /* Skip past it in the archive */
- if (head->header.isextended)
- isextended = 1;
- save_linkflag = head->header.linkflag;
- userec (head);
- if (isextended)
- {
- /* register union record *exhdr;
-
- for (;;) {
- exhdr = findrec();
- if (!exhdr->ext_hdr.isextended) {
- userec(exhdr);
- break;
- }
- }
- userec(exhdr);*/
- skip_extended_headers ();
- }
- /* Skip to the next header on the archive */
- if (save_linkflag != LF_DIR)
- skip_file ((long) hstat.st_size);
- continue;
-
}
+ transform_stat_info (current_header->header.typeflag,
+ ¤t_stat_info);
(*do_something) ();
continue;
- /*
- * If the previous header was good, tell them
- * that we are skipping bad ones.
- */
- case 0: /* Invalid header */
- userec (head);
- switch (prev_status)
+ case HEADER_ZERO_BLOCK:
+ if (block_number_option)
+ {
+ char buf[UINTMAX_STRSIZE_BOUND];
+ fprintf (stdlis, _("block %s: ** Block of NULs **\n"),
+ STRINGIFY_BIGINT (current_block_ordinal (), buf));
+ }
+
+ set_next_block_after (current_header);
+
+ if (!ignore_zeros_option)
{
- case 3: /* Error on first record */
- msg ("Hmm, this doesn't look like a tar archive.");
- /* FALL THRU */
- case 2: /* Error after record of zeroes */
- case 1: /* Error after header rec */
- msg ("Skipping to next file header...");
- case 0: /* Error after error */
+ char buf[UINTMAX_STRSIZE_BOUND];
+
+ status = read_header (¤t_header, ¤t_stat_info,
+ read_header_auto);
+ if (status == HEADER_ZERO_BLOCK)
+ break;
+ WARNOPT (WARN_ALONE_ZERO_BLOCK,
+ (0, 0, _("A lone zero block at %s"),
+ STRINGIFY_BIGINT (current_block_ordinal (), buf)));
break;
}
+ status = prev_status;
continue;
- case 2: /* Record of zeroes */
- userec (head);
- status = prev_status; /* If error after 0's */
- if (f_ignorez)
- continue;
- /* FALL THRU */
- case EOF: /* End of archive */
+ case HEADER_END_OF_FILE:
+ if (block_number_option)
+ {
+ char buf[UINTMAX_STRSIZE_BOUND];
+ fprintf (stdlis, _("block %s: ** End of File **\n"),
+ STRINGIFY_BIGINT (current_block_ordinal (), buf));
+ }
break;
+
+ case HEADER_FAILURE:
+ /* If the previous header was good, tell them that we are
+ skipping bad ones. */
+ set_next_block_after (current_header);
+ switch (prev_status)
+ {
+ case HEADER_STILL_UNREAD:
+ ERROR ((0, 0, _("This does not look like a tar archive")));
+ /* Fall through. */
+
+ case HEADER_ZERO_BLOCK:
+ case HEADER_SUCCESS:
+ if (block_number_option)
+ {
+ char buf[UINTMAX_STRSIZE_BOUND];
+ off_t block_ordinal = current_block_ordinal ();
+ block_ordinal -= recent_long_name_blocks;
+ block_ordinal -= recent_long_link_blocks;
+ fprintf (stdlis, _("block %s: "),
+ STRINGIFY_BIGINT (block_ordinal, buf));
+ }
+ ERROR ((0, 0, _("Skipping to next header")));
+ break;
+
+ case HEADER_END_OF_FILE:
+ case HEADER_FAILURE:
+ /* We are in the middle of a cascade of errors. */
+ break;
+
+ case HEADER_SUCCESS_EXTENDED:
+ abort ();
+ }
+ continue;
}
break;
- };
+ }
+ while (!all_names_found (¤t_stat_info));
- restore_saved_dir_info ();
close_archive ();
- names_notfound (); /* Print names not found */
+ names_notfound (); /* print names not found */
}
-
-/*
- * Print a header record, based on tar options.
- */
+/* Print a header block, based on tar options. */
void
-list_archive ()
+list_archive (void)
{
- extern char *save_name;
- int isextended = 0; /* Flag to remember if head is extended */
+ off_t block_ordinal = current_block_ordinal ();
+
+ /* Print the header block. */
+ if (verbose_option)
+ print_header (¤t_stat_info, current_header, block_ordinal);
+
+ if (incremental_option)
+ {
+ if (verbose_option > 2)
+ {
+ if (is_dumpdir (¤t_stat_info))
+ list_dumpdir (current_stat_info.dumpdir,
+ dumpdir_size (current_stat_info.dumpdir));
+ }
+ }
- /* Save the record */
- saverec (&head);
+ skip_member ();
+}
- /* Print the header record */
- if (f_verbose)
+/* Check header checksum */
+/* The standard BSD tar sources create the checksum by adding up the
+ bytes in the header as type char. I think the type char was unsigned
+ on the PDP-11, but it's signed on the Next and Sun. It looks like the
+ sources to BSD tar were never changed to compute the checksum
+ correctly, so both the Sun and Next add the bytes of the header as
+ signed chars. This doesn't cause a problem until you get a file with
+ a name containing characters with the high bit set. So tar_checksum
+ computes two checksums -- signed and unsigned. */
+
+enum read_header
+tar_checksum (union block *header, bool silent)
+{
+ size_t i;
+ int unsigned_sum = 0; /* the POSIX one :-) */
+ int signed_sum = 0; /* the Sun one :-( */
+ int recorded_sum;
+ int parsed_sum;
+ char *p;
+
+ p = header->buffer;
+ for (i = sizeof *header; i-- != 0;)
{
- if (f_verbose > 1)
- decode_header (head, &hstat, &head_standard, 0);
- print_header ();
+ unsigned_sum += (unsigned char) *p;
+ signed_sum += (signed char) (*p++);
}
- if (f_gnudump && head->header.linkflag == LF_DUMPDIR)
+ if (unsigned_sum == 0)
+ return HEADER_ZERO_BLOCK;
+
+ /* Adjust checksum to count the "chksum" field as blanks. */
+
+ for (i = sizeof header->header.chksum; i-- != 0;)
{
- size_t size, written, check;
- char *data;
- extern long save_totsize;
- extern long save_sizeleft;
+ unsigned_sum -= (unsigned char) header->header.chksum[i];
+ signed_sum -= (signed char) (header->header.chksum[i]);
+ }
+ unsigned_sum += ' ' * sizeof header->header.chksum;
+ signed_sum += ' ' * sizeof header->header.chksum;
+
+ parsed_sum = from_header (header->header.chksum,
+ sizeof header->header.chksum, 0,
+ 0, INT_MAX, true, silent);
+ if (parsed_sum < 0)
+ return HEADER_FAILURE;
+
+ recorded_sum = parsed_sum;
+
+ if (unsigned_sum != recorded_sum && signed_sum != recorded_sum)
+ return HEADER_FAILURE;
- userec (head);
- if (f_multivol)
+ return HEADER_SUCCESS;
+}
+
+/* Read a block that's supposed to be a header block. Return its
+ address in *RETURN_BLOCK, and if it is good, the file's size
+ and names (file name, link name) in *INFO.
+
+ Return one of enum read_header describing the status of the
+ operation.
+
+ The MODE parameter instructs read_header what to do with special
+ header blocks, i.e.: extended POSIX, GNU long name or long link,
+ etc.:
+
+ read_header_auto process them automatically,
+ read_header_x_raw when a special header is read, return
+ HEADER_SUCCESS_EXTENDED without actually
+ processing the header,
+ read_header_x_global when a POSIX global header is read,
+ decode it and return HEADER_SUCCESS_EXTENDED.
+
+ You must always set_next_block_after(*return_block) to skip past
+ the header which this routine reads. */
+
+enum read_header
+read_header (union block **return_block, struct tar_stat_info *info,
+ enum read_header_mode mode)
+{
+ union block *header;
+ union block *header_copy;
+ char *bp;
+ union block *data_block;
+ size_t size, written;
+ union block *next_long_name = 0;
+ union block *next_long_link = 0;
+ size_t next_long_name_blocks = 0;
+ size_t next_long_link_blocks = 0;
+
+ while (1)
+ {
+ enum read_header status;
+
+ header = find_next_block ();
+ *return_block = header;
+ if (!header)
+ return HEADER_END_OF_FILE;
+
+ if ((status = tar_checksum (header, false)) != HEADER_SUCCESS)
+ return status;
+
+ /* Good block. Decode file size and return. */
+
+ if (header->header.typeflag == LNKTYPE)
+ info->stat.st_size = 0; /* links 0 size on tape */
+ else
{
- save_name = current_file_name;
- save_totsize = hstat.st_size;
+ info->stat.st_size = OFF_FROM_HEADER (header->header.size);
+ if (info->stat.st_size < 0)
+ return HEADER_FAILURE;
}
- for (size = hstat.st_size; size > 0; size -= written)
+
+ if (header->header.typeflag == GNUTYPE_LONGNAME
+ || header->header.typeflag == GNUTYPE_LONGLINK
+ || header->header.typeflag == XHDTYPE
+ || header->header.typeflag == XGLTYPE
+ || header->header.typeflag == SOLARIS_XHDTYPE)
{
- if (f_multivol)
- save_sizeleft = size;
- data = findrec ()->charptr;
- if (data == NULL)
+ if (mode == read_header_x_raw)
+ return HEADER_SUCCESS_EXTENDED;
+ else if (header->header.typeflag == GNUTYPE_LONGNAME
+ || header->header.typeflag == GNUTYPE_LONGLINK)
{
- msg ("EOF in archive file?");
- break;
+ size_t name_size = info->stat.st_size;
+ size_t n = name_size % BLOCKSIZE;
+ size = name_size + BLOCKSIZE;
+ if (n)
+ size += BLOCKSIZE - n;
+
+ if (name_size != info->stat.st_size || size < name_size)
+ xalloc_die ();
+
+ header_copy = xmalloc (size + 1);
+
+ if (header->header.typeflag == GNUTYPE_LONGNAME)
+ {
+ free (next_long_name);
+ next_long_name = header_copy;
+ next_long_name_blocks = size / BLOCKSIZE;
+ }
+ else
+ {
+ free (next_long_link);
+ next_long_link = header_copy;
+ next_long_link_blocks = size / BLOCKSIZE;
+ }
+
+ set_next_block_after (header);
+ *header_copy = *header;
+ bp = header_copy->buffer + BLOCKSIZE;
+
+ for (size -= BLOCKSIZE; size > 0; size -= written)
+ {
+ data_block = find_next_block ();
+ if (! data_block)
+ {
+ ERROR ((0, 0, _("Unexpected EOF in archive")));
+ break;
+ }
+ written = available_space_after (data_block);
+ if (written > size)
+ written = size;
+
+ memcpy (bp, data_block->buffer, written);
+ bp += written;
+ set_next_block_after ((union block *)
+ (data_block->buffer + written - 1));
+ }
+
+ *bp = '\0';
}
- written = endofrecs ()->charptr - data;
- if (written > size)
- written = size;
- errno = 0;
- check = fwrite (data, sizeof (char), written, msg_file);
- userec ((union record *) (data + written - 1));
- if (check != written)
+ else if (header->header.typeflag == XHDTYPE
+ || header->header.typeflag == SOLARIS_XHDTYPE)
+ xheader_read (&info->xhdr, header,
+ OFF_FROM_HEADER (header->header.size));
+ else if (header->header.typeflag == XGLTYPE)
{
- msg_perror ("only wrote %ld of %ld bytes to file %s", check, written, current_file_name);
- skip_file ((long) (size) - written);
- break;
+ struct xheader xhdr;
+
+ if (!recent_global_header)
+ recent_global_header = xmalloc (sizeof *recent_global_header);
+ memcpy (recent_global_header, header,
+ sizeof *recent_global_header);
+ memset (&xhdr, 0, sizeof xhdr);
+ xheader_read (&xhdr, header,
+ OFF_FROM_HEADER (header->header.size));
+ xheader_decode_global (&xhdr);
+ xheader_destroy (&xhdr);
+ if (mode == read_header_x_global)
+ return HEADER_SUCCESS_EXTENDED;
}
- }
- if (f_multivol)
- save_name = 0;
- saverec ((union record **) 0); /* Unsave it */
- fputc ('\n', msg_file);
- fflush (msg_file);
- return;
- }
- saverec ((union record **) 0);/* Unsave it */
- /* Check to see if we have an extended header to skip over also */
- if (head->header.isextended)
- isextended = 1;
+ /* Loop! */
- /* Skip past the header in the archive */
- userec (head);
+ }
+ else
+ {
+ char const *name;
+ struct posix_header const *h = &header->header;
+ char namebuf[sizeof h->prefix + 1 + NAME_FIELD_SIZE + 1];
- /*
- * If we needed to skip any extended headers, do so now, by
- * reading extended headers and skipping past them in the
- * archive.
- */
- if (isextended)
- {
- /* register union record *exhdr;
+ free (recent_long_name);
- for (;;) {
- exhdr = findrec();
+ if (next_long_name)
+ {
+ name = next_long_name->buffer + BLOCKSIZE;
+ recent_long_name = next_long_name;
+ recent_long_name_blocks = next_long_name_blocks;
+ }
+ else
+ {
+ /* Accept file names as specified by POSIX.1-1996
+ section 10.1.1. */
+ char *np = namebuf;
- if (!exhdr->ext_hdr.isextended) {
- userec(exhdr);
- break;
- }
- userec(exhdr);
- }*/
- skip_extended_headers ();
- }
+ if (h->prefix[0] && strcmp (h->magic, TMAGIC) == 0)
+ {
+ memcpy (np, h->prefix, sizeof h->prefix);
+ np[sizeof h->prefix] = '\0';
+ np += strlen (np);
+ *np++ = '/';
+ }
+ memcpy (np, h->name, sizeof h->name);
+ np[sizeof h->name] = '\0';
+ name = namebuf;
+ recent_long_name = 0;
+ recent_long_name_blocks = 0;
+ }
+ assign_string (&info->orig_file_name, name);
+ assign_string (&info->file_name, name);
+ info->had_trailing_slash = strip_trailing_slashes (info->file_name);
- if (f_multivol)
- save_name = current_file_name;
- /* Skip to the next header on the archive */
+ free (recent_long_link);
- skip_file ((long) hstat.st_size);
+ if (next_long_link)
+ {
+ name = next_long_link->buffer + BLOCKSIZE;
+ recent_long_link = next_long_link;
+ recent_long_link_blocks = next_long_link_blocks;
+ }
+ else
+ {
+ memcpy (namebuf, h->linkname, sizeof h->linkname);
+ namebuf[sizeof h->linkname] = '\0';
+ name = namebuf;
+ recent_long_link = 0;
+ recent_long_link_blocks = 0;
+ }
+ assign_string (&info->link_name, name);
- if (f_multivol)
- save_name = 0;
+ return HEADER_SUCCESS;
+ }
+ }
}
+#define ISOCTAL(c) ((c)>='0'&&(c)<='7')
+
+/* Decode things from a file HEADER block into STAT_INFO, also setting
+ *FORMAT_POINTER depending on the header block format. If
+ DO_USER_GROUP, decode the user/group information (this is useful
+ for extraction, but waste time when merely listing).
+
+ read_header() has already decoded the checksum and length, so we don't.
-/*
- * Read a record that's supposed to be a header record.
- * Return its address in "head", and if it is good, the file's
- * size in hstat.st_size.
- *
- * Return 1 for success, 0 if the checksum is bad, EOF on eof,
- * 2 for a record full of zeros (EOF marker).
- *
- * You must always userec(head) to skip past the header which this
- * routine reads.
- */
-int
-read_header ()
+ This routine should *not* be called twice for the same block, since
+ the two calls might use different DO_USER_GROUP values and thus
+ might end up with different uid/gid for the two calls. If anybody
+ wants the uid/gid they should decode it first, and other callers
+ should decode it without uid/gid before calling a routine,
+ e.g. print_header, that assumes decoded data. */
+void
+decode_header (union block *header, struct tar_stat_info *stat_info,
+ enum archive_format *format_pointer, int do_user_group)
{
- register int i;
- register long sum, signed_sum, recsum;
- register char *p;
- register union record *header;
- long from_oct ();
- char **longp;
- char *bp, *data;
- int size, written;
- static char *next_long_name, *next_long_link;
- char *name;
+ enum archive_format format;
+ bool hbits;
+ mode_t mode = MODE_FROM_HEADER (header->header.mode, &hbits);
-recurse:
+ if (strcmp (header->header.magic, TMAGIC) == 0)
+ {
+ if (header->star_header.prefix[130] == 0
+ && ISOCTAL (header->star_header.atime[0])
+ && header->star_header.atime[11] == ' '
+ && ISOCTAL (header->star_header.ctime[0])
+ && header->star_header.ctime[11] == ' ')
+ format = STAR_FORMAT;
+ else if (stat_info->xhdr.size)
+ format = POSIX_FORMAT;
+ else
+ format = USTAR_FORMAT;
+ }
+ else if (strcmp (header->buffer + offsetof (struct posix_header, magic),
+ OLDGNU_MAGIC)
+ == 0)
+ format = hbits ? OLDGNU_FORMAT : GNU_FORMAT;
+ else
+ format = V7_FORMAT;
+ *format_pointer = format;
- header = findrec ();
- head = header; /* This is our current header */
- if (NULL == header)
- return EOF;
+ stat_info->stat.st_mode = mode;
+ stat_info->mtime.tv_sec = TIME_FROM_HEADER (header->header.mtime);
+ stat_info->mtime.tv_nsec = 0;
+ assign_string (&stat_info->uname,
+ header->header.uname[0] ? header->header.uname : NULL);
+ assign_string (&stat_info->gname,
+ header->header.gname[0] ? header->header.gname : NULL);
- recsum = from_oct (8, header->header.chksum);
+ xheader_xattr_init (stat_info);
- sum = 0;
- p = header->charptr;
- for (i = sizeof (*header); --i >= 0;)
+ if (format == OLDGNU_FORMAT && incremental_option)
{
- /*
- * We can't use unsigned char here because of old compilers,
- * e.g. V7.
- */
- signed_sum += *p;
- sum += 0xFF & *p++;
+ stat_info->atime.tv_sec = TIME_FROM_HEADER (header->oldgnu_header.atime);
+ stat_info->ctime.tv_sec = TIME_FROM_HEADER (header->oldgnu_header.ctime);
+ stat_info->atime.tv_nsec = stat_info->ctime.tv_nsec = 0;
}
-
- /* Adjust checksum to count the "chksum" field as blanks. */
- for (i = sizeof (header->header.chksum); --i >= 0;)
+ else if (format == STAR_FORMAT)
{
- sum -= 0xFF & header->header.chksum[i];
- signed_sum -= (char) header->header.chksum[i];
+ stat_info->atime.tv_sec = TIME_FROM_HEADER (header->star_header.atime);
+ stat_info->ctime.tv_sec = TIME_FROM_HEADER (header->star_header.ctime);
+ stat_info->atime.tv_nsec = stat_info->ctime.tv_nsec = 0;
}
- sum += ' ' * sizeof header->header.chksum;
- signed_sum += ' ' * sizeof header->header.chksum;
+ else
+ stat_info->atime = stat_info->ctime = start_time;
- if (sum == 8 * ' ')
+ if (format == V7_FORMAT)
{
- /*
- * This is a zeroed record...whole record is 0's except
- * for the 8 blanks we faked for the checksum field.
- */
- return 2;
+ stat_info->stat.st_uid = UID_FROM_HEADER (header->header.uid);
+ stat_info->stat.st_gid = GID_FROM_HEADER (header->header.gid);
+ stat_info->stat.st_rdev = 0;
}
-
- if (sum != recsum && signed_sum != recsum)
- return 0;
-
- /*
- * Good record. Decode file size and return.
- */
- if (header->header.linkflag == LF_LINK)
- hstat.st_size = 0; /* Links 0 size on tape */
else
- hstat.st_size = from_oct (1 + 12, header->header.size);
-
- header->header.arch_name[NAMSIZ - 1] = '\0';
- if (header->header.linkflag == LF_LONGNAME
- || header->header.linkflag == LF_LONGLINK)
{
- longp = ((header->header.linkflag == LF_LONGNAME)
- ? &next_long_name
- : &next_long_link);
+ if (do_user_group)
+ {
+ /* FIXME: Decide if this should somewhat depend on -p. */
+
+ if (numeric_owner_option
+ || !*header->header.uname
+ || !uname_to_uid (header->header.uname, &stat_info->stat.st_uid))
+ stat_info->stat.st_uid = UID_FROM_HEADER (header->header.uid);
- userec (header);
- if (*longp)
- free (*longp);
- bp = *longp = (char *) ck_malloc (hstat.st_size);
+ if (numeric_owner_option
+ || !*header->header.gname
+ || !gname_to_gid (header->header.gname, &stat_info->stat.st_gid))
+ stat_info->stat.st_gid = GID_FROM_HEADER (header->header.gid);
+ }
- for (size = hstat.st_size;
- size > 0;
- size -= written)
+ switch (header->header.typeflag)
{
- data = findrec ()->charptr;
- if (data == NULL)
- {
- msg ("Unexpected EOF on archive file");
- break;
- }
- written = endofrecs ()->charptr - data;
- if (written > size)
- written = size;
+ case BLKTYPE:
+ case CHRTYPE:
+ stat_info->stat.st_rdev =
+ makedev (MAJOR_FROM_HEADER (header->header.devmajor),
+ MINOR_FROM_HEADER (header->header.devminor));
+ break;
- bcopy (data, bp, written);
- bp += written;
- userec ((union record *) (data + written - 1));
+ default:
+ stat_info->stat.st_rdev = 0;
}
- goto recurse;
+ }
+
+ stat_info->archive_file_size = stat_info->stat.st_size;
+ xheader_decode (stat_info);
+
+ if (sparse_member_p (stat_info))
+ {
+ sparse_fixup_header (stat_info);
+ stat_info->is_sparse = true;
}
else
{
- name = (next_long_name
- ? next_long_name
- : head->header.arch_name);
- if (current_file_name)
- free (current_file_name);
- current_file_name = ck_malloc (strlen (name) + 1);
- strcpy (current_file_name, name);
-
- name = (next_long_link
- ? next_long_link
- : head->header.arch_linkname);
- if (current_link_name)
- free (current_link_name);
- current_link_name = ck_malloc (strlen (name) + 1);
- strcpy (current_link_name, name);
-
- next_long_link = next_long_name = 0;
- return 1;
+ stat_info->is_sparse = false;
+ if (((current_format == GNU_FORMAT
+ || current_format == OLDGNU_FORMAT)
+ && current_header->header.typeflag == GNUTYPE_DUMPDIR)
+ || stat_info->dumpdir)
+ stat_info->is_dumpdir = true;
}
}
-/*
- * Decode things from a file header record into a "struct stat".
- * Also set "*stdp" to !=0 or ==0 depending whether header record is "Unix
- * Standard" tar format or regular old tar format.
- *
- * read_header() has already decoded the checksum and length, so we don't.
- *
- * If wantug != 0, we want the uid/group info decoded from Unix Standard
- * tapes (for extraction). If == 0, we are just printing anyway, so save time.
- *
- * decode_header should NOT be called twice for the same record, since the
- * two calls might use different "wantug" values and thus might end up with
- * different uid/gid for the two calls. If anybody wants the uid/gid they
- * should decode it first, and other callers should decode it without uid/gid
- * before calling a routine, e.g. print_header, that assumes decoded data.
- */
-void
-decode_header (header, st, stdp, wantug)
- register union record *header;
- register struct stat *st;
- int *stdp;
- int wantug;
+/* Convert buffer at WHERE0 of size DIGS from external format to
+ intmax_t. DIGS must be positive. If TYPE is nonnull, the data are
+ of type TYPE. The buffer must represent a value in the range
+ MINVAL through MAXVAL; if the mathematically correct result V would
+ be greater than INTMAX_MAX, return a negative integer V such that
+ (uintmax_t) V yields the correct result. If OCTAL_ONLY, allow only octal
+ numbers instead of the other GNU extensions. Return -1 on error,
+ diagnosing the error if TYPE is nonnull and if !SILENT. */
+#if ! (INTMAX_MAX <= UINTMAX_MAX && - (INTMAX_MIN + 1) <= UINTMAX_MAX)
+# error "from_header internally represents intmax_t as uintmax_t + sign"
+#endif
+#if ! (UINTMAX_MAX / 2 <= INTMAX_MAX)
+# error "from_header returns intmax_t to represent uintmax_t"
+#endif
+static intmax_t
+from_header (char const *where0, size_t digs, char const *type,
+ intmax_t minval, uintmax_t maxval,
+ bool octal_only, bool silent)
{
- long from_oct ();
-
- st->st_mode = from_oct (8, header->header.mode);
- st->st_mode &= 07777;
- st->st_mtime = from_oct (1 + 12, header->header.mtime);
- if (f_gnudump)
+ uintmax_t value;
+ uintmax_t uminval = minval;
+ uintmax_t minus_minval = - uminval;
+ char const *where = where0;
+ char const *lim = where + digs;
+ bool negative = false;
+
+ /* Accommodate buggy tar of unknown vintage, which outputs leading
+ NUL if the previous field overflows. */
+ where += !*where;
+
+ /* Accommodate older tars, which output leading spaces. */
+ for (;;)
{
- st->st_atime = from_oct (1 + 12, header->header.atime);
- st->st_ctime = from_oct (1 + 12, header->header.ctime);
+ if (where == lim)
+ {
+ if (type && !silent)
+ ERROR ((0, 0,
+ /* TRANSLATORS: %s is type of the value (gid_t, uid_t,
+ etc.) */
+ _("Blanks in header where numeric %s value expected"),
+ type));
+ return -1;
+ }
+ if (!isspace ((unsigned char) *where))
+ break;
+ where++;
}
- if (0 == strcmp (header->header.magic, TMAGIC))
+ value = 0;
+ if (ISODIGIT (*where))
{
- /* Unix Standard tar archive */
- *stdp = 1;
- if (wantug)
+ char const *where1 = where;
+ bool overflow = false;
+
+ for (;;)
{
-#ifdef NONAMES
- st->st_uid = from_oct (8, header->header.uid);
- st->st_gid = from_oct (8, header->header.gid);
-#else
- st->st_uid =
- (*header->header.uname
- ? finduid (header->header.uname)
- : from_oct (8, header->header.uid));
- st->st_gid =
- (*header->header.gname
- ? findgid (header->header.gname)
- : from_oct (8, header->header.gid));
-#endif
+ value += *where++ - '0';
+ if (where == lim || ! ISODIGIT (*where))
+ break;
+ overflow |= value != (value << LG_8 >> LG_8);
+ value <<= LG_8;
}
-#if defined(S_IFBLK) || defined(S_IFCHR)
- switch (header->header.linkflag)
+
+ /* Parse the output of older, unportable tars, which generate
+ negative values in two's complement octal. If the leading
+ nonzero digit is 1, we can't recover the original value
+ reliably; so do this only if the digit is 2 or more. This
+ catches the common case of 32-bit negative time stamps. */
+ if ((overflow || maxval < value) && '2' <= *where1 && type)
{
- case LF_BLK:
- case LF_CHR:
- st->st_rdev = makedev (from_oct (8, header->header.devmajor),
- from_oct (8, header->header.devminor));
+ /* Compute the negative of the input value, assuming two's
+ complement. */
+ int digit = (*where1 - '0') | 4;
+ overflow = 0;
+ value = 0;
+ where = where1;
+ for (;;)
+ {
+ value += 7 - digit;
+ where++;
+ if (where == lim || ! ISODIGIT (*where))
+ break;
+ digit = *where - '0';
+ overflow |= value != (value << LG_8 >> LG_8);
+ value <<= LG_8;
+ }
+ value++;
+ overflow |= !value;
+
+ if (!overflow && value <= minus_minval)
+ {
+ if (!silent)
+ WARN ((0, 0,
+ /* TRANSLATORS: Second %s is a type name (gid_t,uid_t,etc.) */
+ _("Archive octal value %.*s is out of %s range; assuming two's complement"),
+ (int) (where - where1), where1, type));
+ negative = true;
+ }
+ }
+
+ if (overflow)
+ {
+ if (type && !silent)
+ ERROR ((0, 0,
+ /* TRANSLATORS: Second %s is a type name (gid_t,uid_t,etc.) */
+ _("Archive octal value %.*s is out of %s range"),
+ (int) (where - where1), where1, type));
+ return -1;
}
-#endif
}
- else
+ else if (octal_only)
{
- /* Old fashioned tar archive */
- *stdp = 0;
- st->st_uid = from_oct (8, header->header.uid);
- st->st_gid = from_oct (8, header->header.gid);
- st->st_rdev = 0;
+ /* Suppress the following extensions. */
}
+ else if (*where == '-' || *where == '+')
+ {
+ /* Parse base-64 output produced only by tar test versions
+ 1.13.6 (1999-08-11) through 1.13.11 (1999-08-23).
+ Support for this will be withdrawn in future releases. */
+ int dig;
+ if (!silent)
+ {
+ static bool warned_once;
+ if (! warned_once)
+ {
+ warned_once = true;
+ WARN ((0, 0, _("Archive contains obsolescent base-64 headers")));
+ }
+ }
+ negative = *where++ == '-';
+ while (where != lim
+ && (dig = base64_map[(unsigned char) *where]) < 64)
+ {
+ if (value << LG_64 >> LG_64 != value)
+ {
+ char *string = alloca (digs + 1);
+ memcpy (string, where0, digs);
+ string[digs] = '\0';
+ if (type && !silent)
+ ERROR ((0, 0,
+ _("Archive signed base-64 string %s is out of %s range"),
+ quote (string), type));
+ return -1;
+ }
+ value = (value << LG_64) | dig;
+ where++;
+ }
+ }
+ else if (*where == '\200' /* positive base-256 */
+ || *where == '\377' /* negative base-256 */)
+ {
+ /* Parse base-256 output. A nonnegative number N is
+ represented as (256**DIGS)/2 + N; a negative number -N is
+ represented as (256**DIGS) - N, i.e. as two's complement.
+ The representation guarantees that the leading bit is
+ always on, so that we don't confuse this format with the
+ others (assuming ASCII bytes of 8 bits or more). */
+ int signbit = *where & (1 << (LG_256 - 2));
+ uintmax_t topbits = (((uintmax_t) - signbit)
+ << (CHAR_BIT * sizeof (uintmax_t)
+ - LG_256 - (LG_256 - 2)));
+ value = (*where++ & ((1 << (LG_256 - 2)) - 1)) - signbit;
+ for (;;)
+ {
+ value = (value << LG_256) + (unsigned char) *where++;
+ if (where == lim)
+ break;
+ if (((value << LG_256 >> LG_256) | topbits) != value)
+ {
+ if (type && !silent)
+ ERROR ((0, 0,
+ _("Archive base-256 value is out of %s range"),
+ type));
+ return -1;
+ }
+ }
+ negative = signbit != 0;
+ if (negative)
+ value = -value;
+ }
+
+ if (where != lim && *where && !isspace ((unsigned char) *where))
+ {
+ if (type)
+ {
+ char buf[1000]; /* Big enough to represent any header. */
+ static struct quoting_options *o;
+
+ if (!o)
+ {
+ o = clone_quoting_options (0);
+ set_quoting_style (o, locale_quoting_style);
+ }
+
+ while (where0 != lim && ! lim[-1])
+ lim--;
+ quotearg_buffer (buf, sizeof buf, where0, lim - where0, o);
+ if (!silent)
+ ERROR ((0, 0,
+ /* TRANSLATORS: Second %s is a type name (gid_t,uid_t,etc.) */
+ _("Archive contains %.*s where numeric %s value expected"),
+ (int) sizeof buf, buf, type));
+ }
+
+ return -1;
+ }
+
+ if (value <= (negative ? minus_minval : maxval))
+ return represent_uintmax (negative ? -value : value);
+
+ if (type && !silent)
+ {
+ char minval_buf[UINTMAX_STRSIZE_BOUND + 1];
+ char maxval_buf[UINTMAX_STRSIZE_BOUND];
+ char value_buf[UINTMAX_STRSIZE_BOUND + 1];
+ char *minval_string = STRINGIFY_BIGINT (minus_minval, minval_buf + 1);
+ char *value_string = STRINGIFY_BIGINT (value, value_buf + 1);
+ if (negative)
+ *--value_string = '-';
+ if (minus_minval)
+ *--minval_string = '-';
+ /* TRANSLATORS: Second %s is type name (gid_t,uid_t,etc.) */
+ ERROR ((0, 0, _("Archive value %s is out of %s range %s..%s"),
+ value_string, type,
+ minval_string, STRINGIFY_BIGINT (maxval, maxval_buf)));
+ }
+
+ return -1;
}
+static gid_t
+gid_from_header (const char *p, size_t s)
+{
+ return from_header (p, s, "gid_t",
+ TYPE_MINIMUM (gid_t), TYPE_MAXIMUM (gid_t),
+ false, false);
+}
-/*
- * Quick and dirty octal conversion.
- *
- * Result is -1 if the field is invalid (all blank, or nonoctal).
- */
-long
-from_oct (digs, where)
- register int digs;
- register char *where;
+static major_t
+major_from_header (const char *p, size_t s)
{
- register long value;
+ return from_header (p, s, "major_t",
+ TYPE_MINIMUM (major_t), TYPE_MAXIMUM (major_t),
+ false, false);
+}
- while (isspace (*where))
- { /* Skip spaces */
- where++;
- if (--digs <= 0)
- return -1; /* All blank field */
- }
- value = 0;
- while (digs > 0 && isodigit (*where))
- { /* Scan til nonoctal */
- value = (value << 3) | (*where++ - '0');
- --digs;
+static minor_t
+minor_from_header (const char *p, size_t s)
+{
+ return from_header (p, s, "minor_t",
+ TYPE_MINIMUM (minor_t), TYPE_MAXIMUM (minor_t),
+ false, false);
+}
+
+/* Convert P to the file mode, as understood by tar.
+ Set *HBITS if there are any unrecognized bits. */
+static mode_t
+mode_from_header (const char *p, size_t s, bool *hbits)
+{
+ intmax_t u = from_header (p, s, "mode_t",
+ INTMAX_MIN, UINTMAX_MAX,
+ false, false);
+ mode_t mode = ((u & TSUID ? S_ISUID : 0)
+ | (u & TSGID ? S_ISGID : 0)
+ | (u & TSVTX ? S_ISVTX : 0)
+ | (u & TUREAD ? S_IRUSR : 0)
+ | (u & TUWRITE ? S_IWUSR : 0)
+ | (u & TUEXEC ? S_IXUSR : 0)
+ | (u & TGREAD ? S_IRGRP : 0)
+ | (u & TGWRITE ? S_IWGRP : 0)
+ | (u & TGEXEC ? S_IXGRP : 0)
+ | (u & TOREAD ? S_IROTH : 0)
+ | (u & TOWRITE ? S_IWOTH : 0)
+ | (u & TOEXEC ? S_IXOTH : 0));
+ *hbits = (u & ~07777) != 0;
+ return mode;
+}
+
+off_t
+off_from_header (const char *p, size_t s)
+{
+ /* Negative offsets are not allowed in tar files, so invoke
+ from_header with minimum value 0, not TYPE_MINIMUM (off_t). */
+ return from_header (p, s, "off_t",
+ 0, TYPE_MAXIMUM (off_t),
+ false, false);
+}
+
+static time_t
+time_from_header (const char *p, size_t s)
+{
+ return from_header (p, s, "time_t",
+ TYPE_MINIMUM (time_t), TYPE_MAXIMUM (time_t),
+ false, false);
+}
+
+static uid_t
+uid_from_header (const char *p, size_t s)
+{
+ return from_header (p, s, "uid_t",
+ TYPE_MINIMUM (uid_t), TYPE_MAXIMUM (uid_t),
+ false, false);
+}
+
+uintmax_t
+uintmax_from_header (const char *p, size_t s)
+{
+ return from_header (p, s, "uintmax_t", 0, UINTMAX_MAX, false, false);
+}
+
+
+/* Return a printable representation of T. The result points to
+ static storage that can be reused in the next call to this
+ function, to ctime, or to asctime. If FULL_TIME, then output the
+ time stamp to its full resolution; otherwise, just output it to
+ 1-minute resolution. */
+char const *
+tartime (struct timespec t, bool full_time)
+{
+ enum { fraclen = sizeof ".FFFFFFFFF" - 1 };
+ static char buffer[max (UINTMAX_STRSIZE_BOUND + 1,
+ INT_STRLEN_BOUND (int) + 16)
+ + fraclen];
+ struct tm *tm;
+ time_t s = t.tv_sec;
+ int ns = t.tv_nsec;
+ bool negative = s < 0;
+ char *p;
+
+ if (negative && ns != 0)
+ {
+ s++;
+ ns = 1000000000 - ns;
}
- if (digs > 0 && *where && !isspace (*where))
- return -1; /* Ended on non-space/nul */
+ tm = utc_option ? gmtime (&s) : localtime (&s);
+ if (tm)
+ {
+ if (full_time)
+ {
+ sprintf (buffer, "%04ld-%02d-%02d %02d:%02d:%02d",
+ tm->tm_year + 1900L, tm->tm_mon + 1, tm->tm_mday,
+ tm->tm_hour, tm->tm_min, tm->tm_sec);
+ code_ns_fraction (ns, buffer + strlen (buffer));
+ }
+ else
+ sprintf (buffer, "%04ld-%02d-%02d %02d:%02d",
+ tm->tm_year + 1900L, tm->tm_mon + 1, tm->tm_mday,
+ tm->tm_hour, tm->tm_min);
+ return buffer;
+ }
- return value;
+ /* The time stamp cannot be broken down, most likely because it
+ is out of range. Convert it as an integer,
+ right-adjusted in a field with the same width as the usual
+ 4-year ISO time format. */
+ p = umaxtostr (negative ? - (uintmax_t) s : s,
+ buffer + sizeof buffer - UINTMAX_STRSIZE_BOUND - fraclen);
+ if (negative)
+ *--p = '-';
+ while ((buffer + sizeof buffer - sizeof "YYYY-MM-DD HH:MM"
+ + (full_time ? sizeof ":SS.FFFFFFFFF" - 1 : 0))
+ < p)
+ *--p = ' ';
+ if (full_time)
+ code_ns_fraction (ns, buffer + sizeof buffer - 1 - fraclen);
+ return p;
}
+/* Actually print it.
-/*
- * Actually print it.
- *
- * Plain and fancy file header block logging.
- * Non-verbose just prints the name, e.g. for "tar t" or "tar x".
- * This should just contain file names, so it can be fed back into tar
- * with xargs or the "-T" option. The verbose option can give a bunch
- * of info, one line per file. I doubt anybody tries to parse its
- * format, or if they do, they shouldn't. Unix tar is pretty random here
- * anyway.
- *
- * Note that print_header uses the globals <head>, <hstat>, and
- * <head_standard>, which must be set up in advance. This is not very clean
- * and should be cleaned up. FIXME.
- */
-#define UGSWIDTH 18 /* min width of User, group, size */
-/* UGSWIDTH of 18 means that with user and group names <= 8 chars the columns
- never shift during the listing. */
-#define DATEWIDTH 19 /* Last mod date */
-static int ugswidth = UGSWIDTH; /* Max width encountered so far */
+ Plain and fancy file header block logging. Non-verbose just prints
+ the name, e.g. for "tar t" or "tar x". This should just contain
+ file names, so it can be fed back into tar with xargs or the "-T"
+ option. The verbose option can give a bunch of info, one line per
+ file. I doubt anybody tries to parse its format, or if they do,
+ they shouldn't. Unix tar is pretty random here anyway. */
-void
-print_header ()
+
+/* Width of "user/group size", with initial value chosen
+ heuristically. This grows as needed, though this may cause some
+ stairstepping in the output. Make it too small and the output will
+ almost always look ragged. Make it too large and the output will
+ be spaced out too far. */
+static int ugswidth = 19;
+
+/* Width of printed time stamps. It grows if longer time stamps are
+ found (typically, those with nanosecond resolution). Like
+ USGWIDTH, some stairstepping may occur. */
+static int datewidth = sizeof "YYYY-MM-DD HH:MM" - 1;
+
+static bool volume_label_printed = false;
+
+static void
+simple_print_header (struct tar_stat_info *st, union block *blk,
+ off_t block_ordinal)
{
- char modes[11];
- char *timestamp;
- char uform[11], gform[11]; /* These hold formatted ints */
+ char modes[12];
+ char const *time_stamp;
+ int time_stamp_len;
+ char *temp_name;
+
+ /* These hold formatted ints. */
+ char uform[max (INT_BUFSIZE_BOUND (intmax_t), UINTMAX_STRSIZE_BOUND)];
+ char gform[sizeof uform];
char *user, *group;
- char size[24]; /* Holds a formatted long or maj, min */
- time_t longie; /* To make ctime() call portable */
+ char size[2 * UINTMAX_STRSIZE_BOUND];
+ /* holds formatted size or major,minor */
+ char uintbuf[UINTMAX_STRSIZE_BOUND];
int pad;
- char *name;
- extern long baserec;
+ int sizelen;
- if (f_sayblock)
- fprintf (msg_file, "rec %10d: ", baserec + (ar_record - ar_block));
- /* annofile(msg_file, (char *)NULL); */
+ if (show_transformed_names_option)
+ temp_name = st->file_name ? st->file_name : st->orig_file_name;
+ else
+ temp_name = st->orig_file_name ? st->orig_file_name : st->file_name;
- if (f_verbose <= 1)
+ if (block_number_option)
{
- /* Just the fax, mam. */
- char *name;
+ char buf[UINTMAX_STRSIZE_BOUND];
+ if (block_ordinal < 0)
+ block_ordinal = current_block_ordinal ();
+ block_ordinal -= recent_long_name_blocks;
+ block_ordinal -= recent_long_link_blocks;
+ fprintf (stdlis, _("block %s: "),
+ STRINGIFY_BIGINT (block_ordinal, buf));
+ }
- name = quote_copy_string (current_file_name);
- if (name == 0)
- name = current_file_name;
- fprintf (msg_file, "%s\n", name);
- if (name != current_file_name)
- free (name);
+ if (verbose_option <= 1)
+ {
+ /* Just the fax, mam. */
+ fputs (quotearg (temp_name), stdlis);
+ if (show_transformed_names_option && st->had_trailing_slash)
+ fputc ('/', stdlis);
+ fputc ('\n', stdlis);
}
else
{
- /* File type and modes */
+ /* File type and modes. */
+
modes[0] = '?';
- switch (head->header.linkflag)
+ switch (blk->header.typeflag)
{
- case LF_VOLHDR:
+ case GNUTYPE_VOLHDR:
+ volume_label_printed = true;
modes[0] = 'V';
break;
- case LF_MULTIVOL:
+ case GNUTYPE_MULTIVOL:
modes[0] = 'M';
break;
- case LF_NAMES:
- modes[0] = 'N';
+ case GNUTYPE_LONGNAME:
+ case GNUTYPE_LONGLINK:
+ modes[0] = 'L';
+ ERROR ((0, 0, _("Unexpected long name header")));
break;
- case LF_LONGNAME:
- case LF_LONGLINK:
- msg ("Visible longname error\n");
+ case GNUTYPE_SPARSE:
+ case REGTYPE:
+ case AREGTYPE:
+ modes[0] = st->had_trailing_slash ? 'd' : '-';
break;
-
- case LF_SPARSE:
- case LF_NORMAL:
- case LF_OLDNORMAL:
- case LF_LINK:
- modes[0] = '-';
- if ('/' == current_file_name[strlen (current_file_name) - 1])
- modes[0] = 'd';
+ case LNKTYPE:
+ modes[0] = 'h';
break;
- case LF_DUMPDIR:
+ case GNUTYPE_DUMPDIR:
modes[0] = 'd';
break;
- case LF_DIR:
+ case DIRTYPE:
modes[0] = 'd';
break;
- case LF_SYMLINK:
+ case SYMTYPE:
modes[0] = 'l';
break;
- case LF_BLK:
+ case BLKTYPE:
modes[0] = 'b';
break;
- case LF_CHR:
+ case CHRTYPE:
modes[0] = 'c';
break;
- case LF_FIFO:
+ case FIFOTYPE:
modes[0] = 'p';
break;
- case LF_CONTIG:
+ case CONTTYPE:
modes[0] = 'C';
break;
}
- demode ((unsigned) hstat.st_mode, modes + 1);
+ pax_decode_mode (st->stat.st_mode, modes + 1);
- /* Timestamp */
- longie = hstat.st_mtime;
- timestamp = ctime (&longie);
- timestamp[16] = '\0';
- timestamp[24] = '\0';
+ /* extended attributes: GNU `ls -l'-like preview */
+ xattrs_print_char (st, modes + 10);
- /* User and group names */
- if (*head->header.uname && head_standard)
- {
- user = head->header.uname;
- }
+ /* Time stamp. */
+
+ time_stamp = tartime (st->mtime, full_time_option);
+ time_stamp_len = strlen (time_stamp);
+ if (datewidth < time_stamp_len)
+ datewidth = time_stamp_len;
+
+ /* User and group names. */
+
+ if (st->uname
+ && st->uname[0]
+ && current_format != V7_FORMAT
+ && !numeric_owner_option)
+ user = st->uname;
else
{
- user = uform;
- (void) sprintf (uform, "%d",
- from_oct (8, head->header.uid));
- }
- if (*head->header.gname && head_standard)
- {
- group = head->header.gname;
+ /* Try parsing it as an unsigned integer first, and as a
+ uid_t if that fails. This method can list positive user
+ ids that are too large to fit in a uid_t. */
+ uintmax_t u = from_header (blk->header.uid,
+ sizeof blk->header.uid, 0,
+ 0, UINTMAX_MAX,
+ false, false);
+ user = (u != -1
+ ? STRINGIFY_BIGINT (u, uform)
+ : imaxtostr (UID_FROM_HEADER (blk->header.uid), uform));
}
+
+ if (st->gname
+ && st->gname[0]
+ && current_format != V7_FORMAT
+ && !numeric_owner_option)
+ group = st->gname;
else
{
- group = gform;
- (void) sprintf (gform, "%d",
- from_oct (8, head->header.gid));
+ /* Try parsing it as an unsigned integer first, and as a
+ gid_t if that fails. This method can list positive group
+ ids that are too large to fit in a gid_t. */
+ uintmax_t g = from_header (blk->header.gid,
+ sizeof blk->header.gid, 0,
+ 0, UINTMAX_MAX,
+ false, false);
+ group = (g != -1
+ ? STRINGIFY_BIGINT (g, gform)
+ : imaxtostr (GID_FROM_HEADER (blk->header.gid), gform));
}
- /* Format the file size or major/minor device numbers */
- switch (head->header.linkflag)
+ /* Format the file size or major/minor device numbers. */
+
+ switch (blk->header.typeflag)
{
-#if defined(S_IFBLK) || defined(S_IFCHR)
- case LF_CHR:
- case LF_BLK:
- (void) sprintf (size, "%d,%d",
- major (hstat.st_rdev),
- minor (hstat.st_rdev));
- break;
-#endif
- case LF_SPARSE:
- (void) sprintf (size, "%ld",
- from_oct (1 + 12, head->header.realsize));
+ case CHRTYPE:
+ case BLKTYPE:
+ strcpy (size,
+ STRINGIFY_BIGINT (major (st->stat.st_rdev), uintbuf));
+ strcat (size, ",");
+ strcat (size,
+ STRINGIFY_BIGINT (minor (st->stat.st_rdev), uintbuf));
break;
+
default:
- (void) sprintf (size, "%ld", (long) hstat.st_size);
+ /* st->stat.st_size keeps stored file size */
+ strcpy (size, STRINGIFY_BIGINT (st->stat.st_size, uintbuf));
+ break;
}
- /* Figure out padding and print the whole line. */
- pad = strlen (user) + strlen (group) + strlen (size) + 1;
+ /* Figure out padding and print the whole line. */
+
+ sizelen = strlen (size);
+ pad = strlen (user) + 1 + strlen (group) + 1 + sizelen;
if (pad > ugswidth)
ugswidth = pad;
- name = quote_copy_string (current_file_name);
- if (!name)
- name = current_file_name;
- fprintf (msg_file, "%s %s/%s %*s%s %s %s %s",
- modes,
- user,
- group,
- ugswidth - pad,
- "",
- size,
- timestamp + 4, timestamp + 20,
- name);
-
- if (name != current_file_name)
- free (name);
- switch (head->header.linkflag)
+ fprintf (stdlis, "%s %s/%s %*s %-*s",
+ modes, user, group, ugswidth - pad + sizelen, size,
+ datewidth, time_stamp);
+
+ fprintf (stdlis, " %s", quotearg (temp_name));
+ if (show_transformed_names_option && st->had_trailing_slash)
+ fputc ('/', stdlis);
+
+ switch (blk->header.typeflag)
{
- case LF_SYMLINK:
- name = quote_copy_string (current_link_name);
- if (!name)
- name = current_link_name;
- fprintf (msg_file, " -> %s\n", name);
- if (name != current_link_name)
- free (name);
+ case SYMTYPE:
+ fprintf (stdlis, " -> %s\n", quotearg (st->link_name));
break;
- case LF_LINK:
- name = quote_copy_string (current_link_name);
- if (!name)
- name = current_link_name;
- fprintf (msg_file, " link to %s\n", current_link_name);
- if (name != current_link_name)
- free (name);
+ case LNKTYPE:
+ fprintf (stdlis, _(" link to %s\n"), quotearg (st->link_name));
break;
default:
- fprintf (msg_file, " unknown file type '%c'\n",
- head->header.linkflag);
+ {
+ char type_string[2];
+ type_string[0] = blk->header.typeflag;
+ type_string[1] = '\0';
+ fprintf (stdlis, _(" unknown file type %s\n"),
+ quote (type_string));
+ }
+ break;
+
+ case AREGTYPE:
+ case REGTYPE:
+ case GNUTYPE_SPARSE:
+ case CHRTYPE:
+ case BLKTYPE:
+ case DIRTYPE:
+ case FIFOTYPE:
+ case CONTTYPE:
+ case GNUTYPE_DUMPDIR:
+ putc ('\n', stdlis);
break;
- case LF_OLDNORMAL:
- case LF_NORMAL:
- case LF_SPARSE:
- case LF_CHR:
- case LF_BLK:
- case LF_DIR:
- case LF_FIFO:
- case LF_CONTIG:
- case LF_DUMPDIR:
- putc ('\n', msg_file);
+ case GNUTYPE_LONGLINK:
+ fprintf (stdlis, _("--Long Link--\n"));
break;
- case LF_VOLHDR:
- fprintf (msg_file, "--Volume Header--\n");
+ case GNUTYPE_LONGNAME:
+ fprintf (stdlis, _("--Long Name--\n"));
break;
- case LF_MULTIVOL:
- fprintf (msg_file, "--Continued at byte %ld--\n", from_oct (1 + 12, head->header.offset));
+ case GNUTYPE_VOLHDR:
+ fprintf (stdlis, _("--Volume Header--\n"));
break;
- case LF_NAMES:
- fprintf (msg_file, "--Mangled file names--\n");
+ case GNUTYPE_MULTIVOL:
+ strcpy (size,
+ STRINGIFY_BIGINT
+ (UINTMAX_FROM_HEADER (blk->oldgnu_header.offset),
+ uintbuf));
+ fprintf (stdlis, _("--Continued at byte %s--\n"), size);
break;
}
}
- fflush (msg_file);
+ fflush (stdlis);
+ xattrs_print (st);
+}
+
+
+static void
+print_volume_label (void)
+{
+ struct tar_stat_info vstat;
+ union block vblk;
+ enum archive_format dummy;
+
+ memset (&vblk, 0, sizeof (vblk));
+ vblk.header.typeflag = GNUTYPE_VOLHDR;
+ if (recent_global_header)
+ memcpy (vblk.header.mtime, recent_global_header->header.mtime,
+ sizeof vblk.header.mtime);
+ tar_stat_init (&vstat);
+ assign_string (&vstat.file_name, ".");
+ decode_header (&vblk, &vstat, &dummy, 0);
+ assign_string (&vstat.file_name, volume_label);
+ simple_print_header (&vstat, &vblk, 0);
+ tar_stat_destroy (&vstat);
}
-/*
- * Print a similar line when we make a directory automatically.
- */
void
-pr_mkdir (pathname, length, mode)
- char *pathname;
- int length;
- int mode;
+print_header (struct tar_stat_info *st, union block *blk,
+ off_t block_ordinal)
+{
+ if (current_format == POSIX_FORMAT && !volume_label_printed && volume_label)
+ {
+ print_volume_label ();
+ volume_label_printed = true;
+ }
+
+ simple_print_header (st, blk, block_ordinal);
+}
+
+/* Print a similar line when we make a directory automatically. */
+void
+print_for_mkdir (char *dirname, int length, mode_t mode)
{
char modes[11];
- char *name;
- extern long baserec;
- if (f_verbose > 1)
+ if (verbose_option > 1)
{
- /* File type and modes */
+ /* File type and modes. */
+
modes[0] = 'd';
- demode ((unsigned) mode, modes + 1);
-
- if (f_sayblock)
- fprintf (msg_file, "rec %10d: ", baserec + (ar_record - ar_block));
- /* annofile(msg_file, (char *)NULL); */
- name = quote_copy_string (pathname);
- if (!name)
- name = pathname;
- fprintf (msg_file, "%s %*s %.*s\n",
- modes,
- ugswidth + DATEWIDTH,
- "Creating directory:",
- length,
- pathname);
- if (name != pathname)
- free (name);
+ pax_decode_mode (mode, modes + 1);
+
+ if (block_number_option)
+ {
+ char buf[UINTMAX_STRSIZE_BOUND];
+ fprintf (stdlis, _("block %s: "),
+ STRINGIFY_BIGINT (current_block_ordinal (), buf));
+ }
+
+ fprintf (stdlis, "%s %*s %s\n", modes, ugswidth + 1 + datewidth,
+ _("Creating directory:"), quotearg (dirname));
}
}
-
-/*
- * Skip over <size> bytes of data in records in the archive.
- */
+/* Skip over SIZE bytes of data in blocks in the archive. */
void
-skip_file (size)
- register long size;
+skip_file (off_t size)
{
- union record *x;
- extern long save_totsize;
- extern long save_sizeleft;
+ union block *x;
+
+ /* FIXME: Make sure mv_begin_read is always called before it */
- if (f_multivol)
+ if (seekable_archive)
{
- save_totsize = size;
- save_sizeleft = size;
+ off_t nblk = seek_archive (size);
+ if (nblk >= 0)
+ size -= nblk * BLOCKSIZE;
+ else
+ seekable_archive = false;
}
+ mv_size_left (size);
+
while (size > 0)
{
- x = findrec ();
- if (x == NULL)
- { /* Check it... */
- msg ("Unexpected EOF on archive file");
- exit (EX_BADARCH);
- }
- userec (x);
- size -= RECORDSIZE;
- if (f_multivol)
- save_sizeleft -= RECORDSIZE;
+ x = find_next_block ();
+ if (! x)
+ FATAL_ERROR ((0, 0, _("Unexpected EOF in archive")));
+
+ set_next_block_after (x);
+ size -= BLOCKSIZE;
+ mv_size_left (size);
}
}
+/* Skip the current member in the archive.
+ NOTE: Current header must be decoded before calling this function. */
void
-skip_extended_headers ()
+skip_member (void)
{
- register union record *exhdr;
-
- for (;;)
+ if (!current_stat_info.skipped)
{
- exhdr = findrec ();
- if (!exhdr->ext_hdr.isextended)
- {
- userec (exhdr);
- break;
- }
- userec (exhdr);
+ char save_typeflag = current_header->header.typeflag;
+ set_next_block_after (current_header);
+
+ mv_begin_read (¤t_stat_info);
+
+ if (current_stat_info.is_sparse)
+ sparse_skip_file (¤t_stat_info);
+ else if (save_typeflag != DIRTYPE)
+ skip_file (current_stat_info.stat.st_size);
+
+ mv_end ();
}
}
-/*
- * Decode the mode string from a stat entry into a 9-char string and a null.
- */
void
-demode (mode, string)
- register unsigned mode;
- register char *string;
+test_archive_label (void)
{
- register unsigned mask;
- register char *rwx = "rwxrwxrwx";
+ base64_init ();
+ name_gather ();
- for (mask = 0400; mask != 0; mask >>= 1)
+ open_archive (ACCESS_READ);
+ if (read_header (¤t_header, ¤t_stat_info, read_header_auto)
+ == HEADER_SUCCESS)
{
- if (mode & mask)
- *string++ = *rwx++;
- else
+ decode_header (current_header,
+ ¤t_stat_info, ¤t_format, 0);
+ if (current_header->header.typeflag == GNUTYPE_VOLHDR)
+ assign_string (&volume_label, current_header->header.name);
+
+ if (volume_label)
{
- *string++ = '-';
- rwx++;
+ if (verbose_option)
+ print_volume_label ();
+ if (!name_match (volume_label) && multi_volume_option)
+ {
+ char *s = drop_volume_label_suffix (volume_label);
+ name_match (s);
+ free (s);
+ }
}
}
-
- if (mode & S_ISUID)
- if (string[-7] == 'x')
- string[-7] = 's';
- else
- string[-7] = 'S';
- if (mode & S_ISGID)
- if (string[-4] == 'x')
- string[-4] = 's';
- else
- string[-4] = 'S';
- if (mode & S_ISVTX)
- if (string[-1] == 'x')
- string[-1] = 't';
- else
- string[-1] = 'T';
- *string = '\0';
+ close_archive ();
+ label_notfound ();
}
projects / chaz / tar / blobdiff