For the NetBSD tools build we provide our own fts.h not FreeBSD's so the
comparator should have the NetBSD signature.

PR/57735: Jose Louis Duran: check if fts_read(3) had an error.
(FreeBSD commit https://github.com/freebsd/freebsd-src/commit/\
2dfa4b66b3d0caaaae6ce2df476b5615f8415a19)

Adjust test, we only want to return the structure member if the result is not NULL.

Heads up by Sascha Wildner @ DragonFly BSD.

branches: 1.75.2; 1.75.4;
factor out common code from Poul-Henning Kamp.

use uintmax_t for nlink_t from FreeBSD

branches: 1.73.2; 1.73.16;
Pass a file pointer to write data instead of writing to stdout.
No functional change.

branches: 1.72.2;
Our <sys/param.h> ends up calling header files that define intmax_t. This
should not be the case (but >sys/param.h> is not a standard header so all bets
are off). FreeBSD's does not, so explicitly include <stdint.h> to get it.

Restore old behavior of not printing sizes for directories, since it varies
depending on filesystem implementation. (Brooks Davis)

use intmax_t instead of long long. fixes wrong cast for time_t.

add the ability to only output paths included in a spec file when creating
from a directory (-c)

Implement the "flavor" output discussed in tech-userlevel@, by Brooks Davis

make us compatible with FreeBSD (for -n) (Brooks Davies)

-n is incorrectly supressing the ".." lines. (Brooks Davis)

Allow uid and uname or gid and gname to both be printed while
preserving the behavior of falling back to [ug]id when the name is not
resolvable. (brooks)

clarify seconds.09nanoseconds format (brooks)

add -j flag to indent directory blocks (from brooks)

add -n flag to disable comments (brook)

our fts compare routine is less constified.

more const
check if we support SHA384 before using
(brooks)

branches: 1.59.2;
resource leak (Coverity issues 274383 and 274384)

branches: 1.58.6;
Make "mtree -C" sort its output.

As the input is read from a specfile into a tree of linked lists,
keep each linked list sorted. The sort order is the same as that
already used by "mtree -c": directories sort after non-directories, but
otherwise names are sorted in the order used by strcmp().

branches: 1.57.2;
When {MD5,RMD160,SHA1,SHA256_,SHA384_,SHA512_}File fail, use errno
to print what's actually wrong

fix dev_t formats.

PR/37482: Malcolm Herbert: mtree does not check for getlogin() returning null
in spec create

Renamed dsort to dcmp, since it doesn't sort, it only compares.

Even on non-BSD systems, the "-R all" option should disable printing of
the "time" attribute. Noticed by Malcolm Herbert on tech-pkg.

cast to long explicitly for the compat build

branches: 1.51.4;
The hashing routines (MD5File() and co.) can do dynamic allocation, so
take advantage of that instead of an ugly hard-coded MAXHASHLEN limit
that needs updating.

I think this was suggested by cube@ years ago.

kill crypto/rmd160.h and crypto/sha2.h, and instead make symlinks to
/usr/include from /usr/include/sys. This makes all the one way hash
header handling identical.

Use crypto/rmd160.h.

Use MAXHASHLEN here too.

Add SHA2 support in mtree.

branches: 1.46.2;
strsvis(3) encode & decode the "link=" value (symlink target).

Completely rework how tools/compat is done. Purge all uses/references to
_NETBSD_SOURCE as this makes cross building from older/newer versions of
NetBSD harder, not easier (and also makes the resulting tools 'different')

Wrap all required code with the inclusion of nbtool_config.h, attempt to
only use POSIX code in all places (or when reasonable test w. configure and
provide definitions: ala u_int, etc).

Reviewed by lukem. Tested on FreeBSD 4.9, Redhat Linux ES3, NetBSD 1.6.2 x86
NetBSD current (x86 and amd64) and Solaris 9.

Fixes PR's: PR#17762 PR#25944

branches: 1.44.2;
Overhaul how `build.sh tools' are used:

* Rename "config.h" to "nbtool_config.h" and
HAVE_CONFIG_H to HAVE_NBTOOL_CONFIG_H.
This makes in more obvious in the source when we're using
tools/compat/config.h versus "standard autoconf" config.h

* Consistently move the inclusion of nbtool_config.h to before
<sys/cdefs.h> so that the former can provide __RCSID() (et al),
and there's no need to protect those macros any more.

These changes should make it easier to "tool-ify" a program by adding:
#if HAVE_NBTOOL_CONFIG_H
#include "nbtool_config.h"
#endif
to the top of the source files (for the general case).

Move UCB-licensed code from 4-clause to 3-clause licence.

Patches provided by Joel Baker in PR 22366, verified by myself.

Add -C, which dumps the specification with the full path name first,
and then all the other fields. (I.e, like -D except with the name
first instead of last).
Consistenly strsvis(3) encode path names (even for -C and -D).

avoid non-constant initializer errors by changing hack to workaround
gcc warning to a slightly different hack.

Remove extraneous \n's in {err,warn}{,x} that used to be printfs.

Protect __RCSID and __COPYRIGHT from being invoked if not defined.

Add hooks to make mtree compilable from src/tools.

provide the ability to remove support for generating or checking various
digests by #define-ing NO_MD5, NO_RMD160, or NO_SHA1. the keywords are
still parsed.

User interface changes from (or inspired by) FreeBSD:
- Add -L to walk the tree `logically', by following symbolic links in
the heirarchy.
- Add -P to walk the tree `physically'. This is the current behaviour,
and the default.
- Add "-X excludes-file" to give mtree the ability to exclude files and
directories from its traversal. excludes-file contains fnmatch(3)
patterns to exclude from the walk.
- Add "md5digest" synonym for "md5".
- Add "rmd160" keyword for RMD-160 message digest, and "rmd160digest" synonym.
- Add "sha1" keyword for SHA-1 message digest, and "sha1digest" synonym.
- Don't try to compare() other attributes if the type doesn't match;
it's nothing but trouble, and no use anyway.
- In -c, only emit "/set" records if something has changed since the
previous one.

User interface changes by me:
- Check a device's parameters before checking uid/gid/mode.
- If updating (-u), modify the following to match the specification:
- Device type (retaining existing ownership).
- Symlink target.

Fixes from (or inspired by) FreeBSD:
- Use p->ftslevel instead of own code to keep track of the level ourself.
The previous code got majorly confused if fts(3) couldn't descend
into a subdir, resulting in leaf nodes getting attached to the wrong
directory.
XXX: This new method is much much more robust, even though it's not 100%
perfect; it might result in a couple of following entries in the spec
to be incorrectly tagged as missing.
- Pass a useful pathname to rlink(), so that logical (-L) traversal
doesn't confuse symlink checking.
- Consistently use MAXPATHLEN+1 sized buffers for pathnames, so that
there's room for the NUL.
- Use mtree_err() and strerror(p->fts_errno) to report errors during
the fts(3) walk.

Fixes by me:
- Remove now-unused `const char *name' argument from compare().
- Change crc_total from an int to a u_int32_t, to match usr.bin/cksum/crc.c.
- Remove trailing whitespace.
- Remove unnecessary (void) casts on functions.
- Reorder entries in the getopt() switch.
- Replace strtoq() with strtoll(), and use strtoul() appropriately.
- Renumber F_ flags to be in alphabetical order.

- enforce the top-most entry must be "." and must be a directory; avoids
various potential problems when intermixing full and relative paths
- changes to make it much easier to use spec() - the specfile parser - in
other programs (via .PATHing spec.c and misc.c):
- move excludetags, includetags and keys from mtree.c to misc.c
- implement mtree_err() using vwarnx() instead of assuming name is
"mtree"
- move inotype() and nodetype() from compare.c to misc.c
- add nodetoinode(), to convert from an mtree F_* type to a
mode_t S_IF* type
- clean up #include use; don't assume "mtree.h" pulls in some
standard includes
- change spec() to take a FILE * arg (where the specfile is read from)

complete WARNS=2 cleanup (with a bit of ickiness for the fts_open argv issue)

make this mostly WARNS=2 safe

Add "device" keyword, which allows the device number to be specified,
to be later checked and possibly created. Uses parsing and encoding
routines from mknod(8).

features:
- add a new keyword - "tags" - which is a comma separated list of tags
associated with that file.
- add "-E tags"; exclude files in -D output with matching tags
- add "-I tags"; only include files in -D output with matching tags
(default is to list all)

fixes:
- move all extern variable references from individual files into extern.h
- `quote' some more user-specified strings in error messages
- man page: list an options arguments in the description, and sort xrefs.

- implement -D - dump the specfile out in a manner that's parseable
by various tools
- implement -R key - remove the given key(s) from the list of keys to print
(now it's possible with "-k type -R type" to remove all keys :)
- rename ftype() to nodetype(), and make it public
- use pwcache(3) functions instead of get{gr,pw}{nam,uid}(3)
- add dump_nodes(), which provides the guts for -D.

ANSIfy, KNF, in preperation for more work.

Remove all the "#ifdef __APPLE__" code, to make this actually
build again on Darwin. The code in question is outdated.

OK'd by Wilfredo Sanchez wsanchez@MIT.EDU.

LP64 format string cleanup. Problem reported by garbled.

More format string cleanups by sommerfeld.

branches: 1.25.4;
Now that we have strsvis, use it.

Encode file names using strvis(3) when creating a specification.
Escape white space characters and `#' characters. Do the inverse
when reading specs.

This is a temporary fix until we have strzouvis(3) (or some such)
which is an improved version of strvis (taking an additional list
of characters to encode).

Fixes PR 5006 and PR 7663

branches: 1.23.4;
portability fixes:
- don't use timespecs in struct stat on non 4.4BSD systems
- don't assume that tv_sec is a time_t

Actually generate MD5 checksums, as reported in PR 7928.
Approved by Dr. Lex Wennmacher <wennmach@geo.Uni-Koeln.DE>

branches: 1.21.2;
- add -m and -i to change file flags more usefully.
- fix bugs in fileflags checking

all of the above from dr. lex wennmacher <wennmach@geo.Uni-Koeln.DE>
with some error fixes from egcs via myself.

Appease gcc-2.8.1, make printf formats more portable.

Namespace protection against MAXUID etc.

add support for 4.4BSD file flags. we take a `flags=xxx' parameter,
using the same style of naming as chflags, ls & xinstall use, plus an
additional `none' to turn off a default setting.

Use u_int32_t rather than Sys-V u_long.
Rename err() to mtree_err(), to avoid conflict with libc.

Add braces to satisfy egcs.

Add {} pairs to shut up egcs.

- use an array MAXHOSTNAMELEN+1 size to hold hostnames
- ensure hostname from gethostname() is nul-terminated in all cases
- minor KNF
- use MAXHOSTNAMELEN over various other values/defines
- be safe will buffers that hold hostnames

WARNSify, getopt returns -1 not EOF

fix printf() formats, make explicit comparisons to NULL (gcc -Wall)

Only record sizes for regular files.

branches: 1.10.4;
Rename struct timespec fields to conform to POSIX.1b

add NetBSD id's

clean up import, fix a few bugs, etc.

off_t exp.

make things a bit more consistent

latest mtree from uunet. doesn't compile, but that'll be fixed soon.
(want to have 'pseudo-virgin' sources, for when fixed fnmatch stuff
is released...)

Use *->fts_errno instead of errno after fts_read(). The fts manpage
indicates that the fts_errno will be set to the correct value, but there
are no guarentees about errno.

update for fts library change

Add RCS identifiers.

branches: 1.1.1;
Initial revision

PR/57735: Jose Louis Duran: check if fts_read(3) had an error.
(FreeBSD commit https://github.com/freebsd/freebsd-src/commit/\
2dfa4b66b3d0caaaae6ce2df476b5615f8415a19)

Adjust test, we only want to return the structure member if the result is not NULL.

Heads up by Sascha Wildner @ DragonFly BSD.

branches: 1.75.2; 1.75.4;
factor out common code from Poul-Henning Kamp.

use uintmax_t for nlink_t from FreeBSD

branches: 1.73.2; 1.73.16;
Pass a file pointer to write data instead of writing to stdout.
No functional change.

branches: 1.72.2;
Our <sys/param.h> ends up calling header files that define intmax_t. This
should not be the case (but >sys/param.h> is not a standard header so all bets
are off). FreeBSD's does not, so explicitly include <stdint.h> to get it.

Restore old behavior of not printing sizes for directories, since it varies
depending on filesystem implementation. (Brooks Davis)

use intmax_t instead of long long. fixes wrong cast for time_t.

add the ability to only output paths included in a spec file when creating
from a directory (-c)

Implement the "flavor" output discussed in tech-userlevel@, by Brooks Davis

make us compatible with FreeBSD (for -n) (Brooks Davies)

-n is incorrectly supressing the ".." lines. (Brooks Davis)

Allow uid and uname or gid and gname to both be printed while
preserving the behavior of falling back to [ug]id when the name is not
resolvable. (brooks)

clarify seconds.09nanoseconds format (brooks)

add -j flag to indent directory blocks (from brooks)

add -n flag to disable comments (brook)

our fts compare routine is less constified.

more const
check if we support SHA384 before using
(brooks)

branches: 1.59.2;
resource leak (Coverity issues 274383 and 274384)

branches: 1.58.6;
Make "mtree -C" sort its output.

As the input is read from a specfile into a tree of linked lists,
keep each linked list sorted. The sort order is the same as that
already used by "mtree -c": directories sort after non-directories, but
otherwise names are sorted in the order used by strcmp().

branches: 1.57.2;
When {MD5,RMD160,SHA1,SHA256_,SHA384_,SHA512_}File fail, use errno
to print what's actually wrong

fix dev_t formats.

PR/37482: Malcolm Herbert: mtree does not check for getlogin() returning null
in spec create

Renamed dsort to dcmp, since it doesn't sort, it only compares.

Even on non-BSD systems, the "-R all" option should disable printing of
the "time" attribute. Noticed by Malcolm Herbert on tech-pkg.

cast to long explicitly for the compat build

branches: 1.51.4;
The hashing routines (MD5File() and co.) can do dynamic allocation, so
take advantage of that instead of an ugly hard-coded MAXHASHLEN limit
that needs updating.

I think this was suggested by cube@ years ago.

kill crypto/rmd160.h and crypto/sha2.h, and instead make symlinks to
/usr/include from /usr/include/sys. This makes all the one way hash
header handling identical.

Use crypto/rmd160.h.

Use MAXHASHLEN here too.

Add SHA2 support in mtree.

branches: 1.46.2;
strsvis(3) encode & decode the "link=" value (symlink target).

Completely rework how tools/compat is done. Purge all uses/references to
_NETBSD_SOURCE as this makes cross building from older/newer versions of
NetBSD harder, not easier (and also makes the resulting tools 'different')

Wrap all required code with the inclusion of nbtool_config.h, attempt to
only use POSIX code in all places (or when reasonable test w. configure and
provide definitions: ala u_int, etc).

Reviewed by lukem. Tested on FreeBSD 4.9, Redhat Linux ES3, NetBSD 1.6.2 x86
NetBSD current (x86 and amd64) and Solaris 9.

Fixes PR's: PR#17762 PR#25944

branches: 1.44.2;
Overhaul how `build.sh tools' are used:

* Rename "config.h" to "nbtool_config.h" and
HAVE_CONFIG_H to HAVE_NBTOOL_CONFIG_H.
This makes in more obvious in the source when we're using
tools/compat/config.h versus "standard autoconf" config.h

* Consistently move the inclusion of nbtool_config.h to before
<sys/cdefs.h> so that the former can provide __RCSID() (et al),
and there's no need to protect those macros any more.

These changes should make it easier to "tool-ify" a program by adding:
#if HAVE_NBTOOL_CONFIG_H
#include "nbtool_config.h"
#endif
to the top of the source files (for the general case).

Move UCB-licensed code from 4-clause to 3-clause licence.

Patches provided by Joel Baker in PR 22366, verified by myself.

Add -C, which dumps the specification with the full path name first,
and then all the other fields. (I.e, like -D except with the name
first instead of last).
Consistenly strsvis(3) encode path names (even for -C and -D).

avoid non-constant initializer errors by changing hack to workaround
gcc warning to a slightly different hack.

Remove extraneous \n's in {err,warn}{,x} that used to be printfs.

Protect __RCSID and __COPYRIGHT from being invoked if not defined.

Add hooks to make mtree compilable from src/tools.

provide the ability to remove support for generating or checking various
digests by #define-ing NO_MD5, NO_RMD160, or NO_SHA1. the keywords are
still parsed.

User interface changes from (or inspired by) FreeBSD:
- Add -L to walk the tree `logically', by following symbolic links in
the heirarchy.
- Add -P to walk the tree `physically'. This is the current behaviour,
and the default.
- Add "-X excludes-file" to give mtree the ability to exclude files and
directories from its traversal. excludes-file contains fnmatch(3)
patterns to exclude from the walk.
- Add "md5digest" synonym for "md5".
- Add "rmd160" keyword for RMD-160 message digest, and "rmd160digest" synonym.
- Add "sha1" keyword for SHA-1 message digest, and "sha1digest" synonym.
- Don't try to compare() other attributes if the type doesn't match;
it's nothing but trouble, and no use anyway.
- In -c, only emit "/set" records if something has changed since the
previous one.

User interface changes by me:
- Check a device's parameters before checking uid/gid/mode.
- If updating (-u), modify the following to match the specification:
- Device type (retaining existing ownership).
- Symlink target.

Fixes from (or inspired by) FreeBSD:
- Use p->ftslevel instead of own code to keep track of the level ourself.
The previous code got majorly confused if fts(3) couldn't descend
into a subdir, resulting in leaf nodes getting attached to the wrong
directory.
XXX: This new method is much much more robust, even though it's not 100%
perfect; it might result in a couple of following entries in the spec
to be incorrectly tagged as missing.
- Pass a useful pathname to rlink(), so that logical (-L) traversal
doesn't confuse symlink checking.
- Consistently use MAXPATHLEN+1 sized buffers for pathnames, so that
there's room for the NUL.
- Use mtree_err() and strerror(p->fts_errno) to report errors during
the fts(3) walk.

Fixes by me:
- Remove now-unused `const char *name' argument from compare().
- Change crc_total from an int to a u_int32_t, to match usr.bin/cksum/crc.c.
- Remove trailing whitespace.
- Remove unnecessary (void) casts on functions.
- Reorder entries in the getopt() switch.
- Replace strtoq() with strtoll(), and use strtoul() appropriately.
- Renumber F_ flags to be in alphabetical order.

- enforce the top-most entry must be "." and must be a directory; avoids
various potential problems when intermixing full and relative paths
- changes to make it much easier to use spec() - the specfile parser - in
other programs (via .PATHing spec.c and misc.c):
- move excludetags, includetags and keys from mtree.c to misc.c
- implement mtree_err() using vwarnx() instead of assuming name is
"mtree"
- move inotype() and nodetype() from compare.c to misc.c
- add nodetoinode(), to convert from an mtree F_* type to a
mode_t S_IF* type
- clean up #include use; don't assume "mtree.h" pulls in some
standard includes
- change spec() to take a FILE * arg (where the specfile is read from)

complete WARNS=2 cleanup (with a bit of ickiness for the fts_open argv issue)

make this mostly WARNS=2 safe

Add "device" keyword, which allows the device number to be specified,
to be later checked and possibly created. Uses parsing and encoding
routines from mknod(8).

features:
- add a new keyword - "tags" - which is a comma separated list of tags
associated with that file.
- add "-E tags"; exclude files in -D output with matching tags
- add "-I tags"; only include files in -D output with matching tags
(default is to list all)

fixes:
- move all extern variable references from individual files into extern.h
- `quote' some more user-specified strings in error messages
- man page: list an options arguments in the description, and sort xrefs.

- implement -D - dump the specfile out in a manner that's parseable
by various tools
- implement -R key - remove the given key(s) from the list of keys to print
(now it's possible with "-k type -R type" to remove all keys :)
- rename ftype() to nodetype(), and make it public
- use pwcache(3) functions instead of get{gr,pw}{nam,uid}(3)
- add dump_nodes(), which provides the guts for -D.

ANSIfy, KNF, in preperation for more work.

Remove all the "#ifdef __APPLE__" code, to make this actually
build again on Darwin. The code in question is outdated.

OK'd by Wilfredo Sanchez wsanchez@MIT.EDU.

LP64 format string cleanup. Problem reported by garbled.

More format string cleanups by sommerfeld.

branches: 1.25.4;
Now that we have strsvis, use it.

Encode file names using strvis(3) when creating a specification.
Escape white space characters and `#' characters. Do the inverse
when reading specs.

This is a temporary fix until we have strzouvis(3) (or some such)
which is an improved version of strvis (taking an additional list
of characters to encode).

Fixes PR 5006 and PR 7663

branches: 1.23.4;
portability fixes:
- don't use timespecs in struct stat on non 4.4BSD systems
- don't assume that tv_sec is a time_t

Actually generate MD5 checksums, as reported in PR 7928.
Approved by Dr. Lex Wennmacher <wennmach@geo.Uni-Koeln.DE>

branches: 1.21.2;
- add -m and -i to change file flags more usefully.
- fix bugs in fileflags checking

all of the above from dr. lex wennmacher <wennmach@geo.Uni-Koeln.DE>
with some error fixes from egcs via myself.

Appease gcc-2.8.1, make printf formats more portable.

Namespace protection against MAXUID etc.

add support for 4.4BSD file flags. we take a `flags=xxx' parameter,
using the same style of naming as chflags, ls & xinstall use, plus an
additional `none' to turn off a default setting.

Use u_int32_t rather than Sys-V u_long.
Rename err() to mtree_err(), to avoid conflict with libc.

Add braces to satisfy egcs.

Add {} pairs to shut up egcs.

- use an array MAXHOSTNAMELEN+1 size to hold hostnames
- ensure hostname from gethostname() is nul-terminated in all cases
- minor KNF
- use MAXHOSTNAMELEN over various other values/defines
- be safe will buffers that hold hostnames

WARNSify, getopt returns -1 not EOF

fix printf() formats, make explicit comparisons to NULL (gcc -Wall)

Only record sizes for regular files.

branches: 1.10.4;
Rename struct timespec fields to conform to POSIX.1b

add NetBSD id's

clean up import, fix a few bugs, etc.

off_t exp.

make things a bit more consistent

latest mtree from uunet. doesn't compile, but that'll be fixed soon.
(want to have 'pseudo-virgin' sources, for when fixed fnmatch stuff
is released...)

Use *->fts_errno instead of errno after fts_read(). The fts manpage
indicates that the fts_errno will be set to the correct value, but there
are no guarentees about errno.

update for fts library change

Add RCS identifiers.

branches: 1.1.1;
Initial revision

Adjust test, we only want to return the structure member if the result is not NULL.

Heads up by Sascha Wildner @ DragonFly BSD.

branches: 1.75.2;
factor out common code from Poul-Henning Kamp.

use uintmax_t for nlink_t from FreeBSD

branches: 1.73.2; 1.73.16;
Pass a file pointer to write data instead of writing to stdout.
No functional change.

branches: 1.72.2;
Our <sys/param.h> ends up calling header files that define intmax_t. This
should not be the case (but >sys/param.h> is not a standard header so all bets
are off). FreeBSD's does not, so explicitly include <stdint.h> to get it.

Restore old behavior of not printing sizes for directories, since it varies
depending on filesystem implementation. (Brooks Davis)

use intmax_t instead of long long. fixes wrong cast for time_t.

add the ability to only output paths included in a spec file when creating
from a directory (-c)

Implement the "flavor" output discussed in tech-userlevel@, by Brooks Davis

make us compatible with FreeBSD (for -n) (Brooks Davies)

-n is incorrectly supressing the ".." lines. (Brooks Davis)

Allow uid and uname or gid and gname to both be printed while
preserving the behavior of falling back to [ug]id when the name is not
resolvable. (brooks)

clarify seconds.09nanoseconds format (brooks)

add -j flag to indent directory blocks (from brooks)

add -n flag to disable comments (brook)

our fts compare routine is less constified.

more const
check if we support SHA384 before using
(brooks)

branches: 1.59.2;
resource leak (Coverity issues 274383 and 274384)

branches: 1.58.6;
Make "mtree -C" sort its output.

As the input is read from a specfile into a tree of linked lists,
keep each linked list sorted. The sort order is the same as that
already used by "mtree -c": directories sort after non-directories, but
otherwise names are sorted in the order used by strcmp().

branches: 1.57.2;
When {MD5,RMD160,SHA1,SHA256_,SHA384_,SHA512_}File fail, use errno
to print what's actually wrong

fix dev_t formats.

PR/37482: Malcolm Herbert: mtree does not check for getlogin() returning null
in spec create

Renamed dsort to dcmp, since it doesn't sort, it only compares.

Even on non-BSD systems, the "-R all" option should disable printing of
the "time" attribute. Noticed by Malcolm Herbert on tech-pkg.

cast to long explicitly for the compat build

branches: 1.51.4;
The hashing routines (MD5File() and co.) can do dynamic allocation, so
take advantage of that instead of an ugly hard-coded MAXHASHLEN limit
that needs updating.

I think this was suggested by cube@ years ago.

kill crypto/rmd160.h and crypto/sha2.h, and instead make symlinks to
/usr/include from /usr/include/sys. This makes all the one way hash
header handling identical.

Use crypto/rmd160.h.

Use MAXHASHLEN here too.

Add SHA2 support in mtree.

branches: 1.46.2;
strsvis(3) encode & decode the "link=" value (symlink target).

Completely rework how tools/compat is done. Purge all uses/references to
_NETBSD_SOURCE as this makes cross building from older/newer versions of
NetBSD harder, not easier (and also makes the resulting tools 'different')

Wrap all required code with the inclusion of nbtool_config.h, attempt to
only use POSIX code in all places (or when reasonable test w. configure and
provide definitions: ala u_int, etc).

Reviewed by lukem. Tested on FreeBSD 4.9, Redhat Linux ES3, NetBSD 1.6.2 x86
NetBSD current (x86 and amd64) and Solaris 9.

Fixes PR's: PR#17762 PR#25944

branches: 1.44.2;
Overhaul how `build.sh tools' are used:

* Rename "config.h" to "nbtool_config.h" and
HAVE_CONFIG_H to HAVE_NBTOOL_CONFIG_H.
This makes in more obvious in the source when we're using
tools/compat/config.h versus "standard autoconf" config.h

* Consistently move the inclusion of nbtool_config.h to before
<sys/cdefs.h> so that the former can provide __RCSID() (et al),
and there's no need to protect those macros any more.

These changes should make it easier to "tool-ify" a program by adding:
#if HAVE_NBTOOL_CONFIG_H
#include "nbtool_config.h"
#endif
to the top of the source files (for the general case).

Move UCB-licensed code from 4-clause to 3-clause licence.

Patches provided by Joel Baker in PR 22366, verified by myself.

Add -C, which dumps the specification with the full path name first,
and then all the other fields. (I.e, like -D except with the name
first instead of last).
Consistenly strsvis(3) encode path names (even for -C and -D).

avoid non-constant initializer errors by changing hack to workaround
gcc warning to a slightly different hack.

Remove extraneous \n's in {err,warn}{,x} that used to be printfs.

Protect __RCSID and __COPYRIGHT from being invoked if not defined.

Add hooks to make mtree compilable from src/tools.

provide the ability to remove support for generating or checking various
digests by #define-ing NO_MD5, NO_RMD160, or NO_SHA1. the keywords are
still parsed.

User interface changes from (or inspired by) FreeBSD:
- Add -L to walk the tree `logically', by following symbolic links in
the heirarchy.
- Add -P to walk the tree `physically'. This is the current behaviour,
and the default.
- Add "-X excludes-file" to give mtree the ability to exclude files and
directories from its traversal. excludes-file contains fnmatch(3)
patterns to exclude from the walk.
- Add "md5digest" synonym for "md5".
- Add "rmd160" keyword for RMD-160 message digest, and "rmd160digest" synonym.
- Add "sha1" keyword for SHA-1 message digest, and "sha1digest" synonym.
- Don't try to compare() other attributes if the type doesn't match;
it's nothing but trouble, and no use anyway.
- In -c, only emit "/set" records if something has changed since the
previous one.

User interface changes by me:
- Check a device's parameters before checking uid/gid/mode.
- If updating (-u), modify the following to match the specification:
- Device type (retaining existing ownership).
- Symlink target.

Fixes from (or inspired by) FreeBSD:
- Use p->ftslevel instead of own code to keep track of the level ourself.
The previous code got majorly confused if fts(3) couldn't descend
into a subdir, resulting in leaf nodes getting attached to the wrong
directory.
XXX: This new method is much much more robust, even though it's not 100%
perfect; it might result in a couple of following entries in the spec
to be incorrectly tagged as missing.
- Pass a useful pathname to rlink(), so that logical (-L) traversal
doesn't confuse symlink checking.
- Consistently use MAXPATHLEN+1 sized buffers for pathnames, so that
there's room for the NUL.
- Use mtree_err() and strerror(p->fts_errno) to report errors during
the fts(3) walk.

Fixes by me:
- Remove now-unused `const char *name' argument from compare().
- Change crc_total from an int to a u_int32_t, to match usr.bin/cksum/crc.c.
- Remove trailing whitespace.
- Remove unnecessary (void) casts on functions.
- Reorder entries in the getopt() switch.
- Replace strtoq() with strtoll(), and use strtoul() appropriately.
- Renumber F_ flags to be in alphabetical order.

- enforce the top-most entry must be "." and must be a directory; avoids
various potential problems when intermixing full and relative paths
- changes to make it much easier to use spec() - the specfile parser - in
other programs (via .PATHing spec.c and misc.c):
- move excludetags, includetags and keys from mtree.c to misc.c
- implement mtree_err() using vwarnx() instead of assuming name is
"mtree"
- move inotype() and nodetype() from compare.c to misc.c
- add nodetoinode(), to convert from an mtree F_* type to a
mode_t S_IF* type
- clean up #include use; don't assume "mtree.h" pulls in some
standard includes
- change spec() to take a FILE * arg (where the specfile is read from)

complete WARNS=2 cleanup (with a bit of ickiness for the fts_open argv issue)

make this mostly WARNS=2 safe

Add "device" keyword, which allows the device number to be specified,
to be later checked and possibly created. Uses parsing and encoding
routines from mknod(8).

features:
- add a new keyword - "tags" - which is a comma separated list of tags
associated with that file.
- add "-E tags"; exclude files in -D output with matching tags
- add "-I tags"; only include files in -D output with matching tags
(default is to list all)

fixes:
- move all extern variable references from individual files into extern.h
- `quote' some more user-specified strings in error messages
- man page: list an options arguments in the description, and sort xrefs.

- implement -D - dump the specfile out in a manner that's parseable
by various tools
- implement -R key - remove the given key(s) from the list of keys to print
(now it's possible with "-k type -R type" to remove all keys :)
- rename ftype() to nodetype(), and make it public
- use pwcache(3) functions instead of get{gr,pw}{nam,uid}(3)
- add dump_nodes(), which provides the guts for -D.

ANSIfy, KNF, in preperation for more work.

Remove all the "#ifdef __APPLE__" code, to make this actually
build again on Darwin. The code in question is outdated.

OK'd by Wilfredo Sanchez wsanchez@MIT.EDU.

LP64 format string cleanup. Problem reported by garbled.

More format string cleanups by sommerfeld.

branches: 1.25.4;
Now that we have strsvis, use it.

Encode file names using strvis(3) when creating a specification.
Escape white space characters and `#' characters. Do the inverse
when reading specs.

This is a temporary fix until we have strzouvis(3) (or some such)
which is an improved version of strvis (taking an additional list
of characters to encode).

Fixes PR 5006 and PR 7663

branches: 1.23.4;
portability fixes:
- don't use timespecs in struct stat on non 4.4BSD systems
- don't assume that tv_sec is a time_t

Actually generate MD5 checksums, as reported in PR 7928.
Approved by Dr. Lex Wennmacher <wennmach@geo.Uni-Koeln.DE>

branches: 1.21.2;
- add -m and -i to change file flags more usefully.
- fix bugs in fileflags checking

all of the above from dr. lex wennmacher <wennmach@geo.Uni-Koeln.DE>
with some error fixes from egcs via myself.

Appease gcc-2.8.1, make printf formats more portable.

Namespace protection against MAXUID etc.

add support for 4.4BSD file flags. we take a `flags=xxx' parameter,
using the same style of naming as chflags, ls & xinstall use, plus an
additional `none' to turn off a default setting.

Use u_int32_t rather than Sys-V u_long.
Rename err() to mtree_err(), to avoid conflict with libc.

Add braces to satisfy egcs.

Add {} pairs to shut up egcs.

- use an array MAXHOSTNAMELEN+1 size to hold hostnames
- ensure hostname from gethostname() is nul-terminated in all cases
- minor KNF
- use MAXHOSTNAMELEN over various other values/defines
- be safe will buffers that hold hostnames

WARNSify, getopt returns -1 not EOF

fix printf() formats, make explicit comparisons to NULL (gcc -Wall)

Only record sizes for regular files.

branches: 1.10.4;
Rename struct timespec fields to conform to POSIX.1b

add NetBSD id's

clean up import, fix a few bugs, etc.

off_t exp.

make things a bit more consistent

latest mtree from uunet. doesn't compile, but that'll be fixed soon.
(want to have 'pseudo-virgin' sources, for when fixed fnmatch stuff
is released...)

Use *->fts_errno instead of errno after fts_read(). The fts manpage
indicates that the fts_errno will be set to the correct value, but there
are no guarentees about errno.

update for fts library change

Add RCS identifiers.

branches: 1.1.1;
Initial revision

factor out common code from Poul-Henning Kamp.

use uintmax_t for nlink_t from FreeBSD

Pass a file pointer to write data instead of writing to stdout.
No functional change.

branches: 1.72.2;
Our <sys/param.h> ends up calling header files that define intmax_t. This
should not be the case (but >sys/param.h> is not a standard header so all bets
are off). FreeBSD's does not, so explicitly include <stdint.h> to get it.

Restore old behavior of not printing sizes for directories, since it varies
depending on filesystem implementation. (Brooks Davis)

use intmax_t instead of long long. fixes wrong cast for time_t.

add the ability to only output paths included in a spec file when creating
from a directory (-c)

Implement the "flavor" output discussed in tech-userlevel@, by Brooks Davis

make us compatible with FreeBSD (for -n) (Brooks Davies)

-n is incorrectly supressing the ".." lines. (Brooks Davis)

Allow uid and uname or gid and gname to both be printed while
preserving the behavior of falling back to [ug]id when the name is not
resolvable. (brooks)

clarify seconds.09nanoseconds format (brooks)

add -j flag to indent directory blocks (from brooks)

add -n flag to disable comments (brook)

our fts compare routine is less constified.

more const
check if we support SHA384 before using
(brooks)

branches: 1.59.2;
resource leak (Coverity issues 274383 and 274384)

branches: 1.58.6;
Make "mtree -C" sort its output.

As the input is read from a specfile into a tree of linked lists,
keep each linked list sorted. The sort order is the same as that
already used by "mtree -c": directories sort after non-directories, but
otherwise names are sorted in the order used by strcmp().

branches: 1.57.2;
When {MD5,RMD160,SHA1,SHA256_,SHA384_,SHA512_}File fail, use errno
to print what's actually wrong

fix dev_t formats.

PR/37482: Malcolm Herbert: mtree does not check for getlogin() returning null
in spec create

Renamed dsort to dcmp, since it doesn't sort, it only compares.

Even on non-BSD systems, the "-R all" option should disable printing of
the "time" attribute. Noticed by Malcolm Herbert on tech-pkg.

cast to long explicitly for the compat build

branches: 1.51.4;
The hashing routines (MD5File() and co.) can do dynamic allocation, so
take advantage of that instead of an ugly hard-coded MAXHASHLEN limit
that needs updating.

I think this was suggested by cube@ years ago.

kill crypto/rmd160.h and crypto/sha2.h, and instead make symlinks to
/usr/include from /usr/include/sys. This makes all the one way hash
header handling identical.

Use crypto/rmd160.h.

Use MAXHASHLEN here too.

Add SHA2 support in mtree.

branches: 1.46.2;
strsvis(3) encode & decode the "link=" value (symlink target).

Completely rework how tools/compat is done. Purge all uses/references to
_NETBSD_SOURCE as this makes cross building from older/newer versions of
NetBSD harder, not easier (and also makes the resulting tools 'different')

Wrap all required code with the inclusion of nbtool_config.h, attempt to
only use POSIX code in all places (or when reasonable test w. configure and
provide definitions: ala u_int, etc).

Reviewed by lukem. Tested on FreeBSD 4.9, Redhat Linux ES3, NetBSD 1.6.2 x86
NetBSD current (x86 and amd64) and Solaris 9.

Fixes PR's: PR#17762 PR#25944

branches: 1.44.2;
Overhaul how `build.sh tools' are used:

* Rename "config.h" to "nbtool_config.h" and
HAVE_CONFIG_H to HAVE_NBTOOL_CONFIG_H.
This makes in more obvious in the source when we're using
tools/compat/config.h versus "standard autoconf" config.h

* Consistently move the inclusion of nbtool_config.h to before
<sys/cdefs.h> so that the former can provide __RCSID() (et al),
and there's no need to protect those macros any more.

These changes should make it easier to "tool-ify" a program by adding:
#if HAVE_NBTOOL_CONFIG_H
#include "nbtool_config.h"
#endif
to the top of the source files (for the general case).

Move UCB-licensed code from 4-clause to 3-clause licence.

Patches provided by Joel Baker in PR 22366, verified by myself.

Add -C, which dumps the specification with the full path name first,
and then all the other fields. (I.e, like -D except with the name
first instead of last).
Consistenly strsvis(3) encode path names (even for -C and -D).

avoid non-constant initializer errors by changing hack to workaround
gcc warning to a slightly different hack.

Remove extraneous \n's in {err,warn}{,x} that used to be printfs.

Protect __RCSID and __COPYRIGHT from being invoked if not defined.

Add hooks to make mtree compilable from src/tools.

provide the ability to remove support for generating or checking various
digests by #define-ing NO_MD5, NO_RMD160, or NO_SHA1. the keywords are
still parsed.

User interface changes from (or inspired by) FreeBSD:
- Add -L to walk the tree `logically', by following symbolic links in
the heirarchy.
- Add -P to walk the tree `physically'. This is the current behaviour,
and the default.
- Add "-X excludes-file" to give mtree the ability to exclude files and
directories from its traversal. excludes-file contains fnmatch(3)
patterns to exclude from the walk.
- Add "md5digest" synonym for "md5".
- Add "rmd160" keyword for RMD-160 message digest, and "rmd160digest" synonym.
- Add "sha1" keyword for SHA-1 message digest, and "sha1digest" synonym.
- Don't try to compare() other attributes if the type doesn't match;
it's nothing but trouble, and no use anyway.
- In -c, only emit "/set" records if something has changed since the
previous one.

User interface changes by me:
- Check a device's parameters before checking uid/gid/mode.
- If updating (-u), modify the following to match the specification:
- Device type (retaining existing ownership).
- Symlink target.

Fixes from (or inspired by) FreeBSD:
- Use p->ftslevel instead of own code to keep track of the level ourself.
The previous code got majorly confused if fts(3) couldn't descend
into a subdir, resulting in leaf nodes getting attached to the wrong
directory.
XXX: This new method is much much more robust, even though it's not 100%
perfect; it might result in a couple of following entries in the spec
to be incorrectly tagged as missing.
- Pass a useful pathname to rlink(), so that logical (-L) traversal
doesn't confuse symlink checking.
- Consistently use MAXPATHLEN+1 sized buffers for pathnames, so that
there's room for the NUL.
- Use mtree_err() and strerror(p->fts_errno) to report errors during
the fts(3) walk.

Fixes by me:
- Remove now-unused `const char *name' argument from compare().
- Change crc_total from an int to a u_int32_t, to match usr.bin/cksum/crc.c.
- Remove trailing whitespace.
- Remove unnecessary (void) casts on functions.
- Reorder entries in the getopt() switch.
- Replace strtoq() with strtoll(), and use strtoul() appropriately.
- Renumber F_ flags to be in alphabetical order.

- enforce the top-most entry must be "." and must be a directory; avoids
various potential problems when intermixing full and relative paths
- changes to make it much easier to use spec() - the specfile parser - in
other programs (via .PATHing spec.c and misc.c):
- move excludetags, includetags and keys from mtree.c to misc.c
- implement mtree_err() using vwarnx() instead of assuming name is
"mtree"
- move inotype() and nodetype() from compare.c to misc.c
- add nodetoinode(), to convert from an mtree F_* type to a
mode_t S_IF* type
- clean up #include use; don't assume "mtree.h" pulls in some
standard includes
- change spec() to take a FILE * arg (where the specfile is read from)

complete WARNS=2 cleanup (with a bit of ickiness for the fts_open argv issue)

make this mostly WARNS=2 safe

Add "device" keyword, which allows the device number to be specified,
to be later checked and possibly created. Uses parsing and encoding
routines from mknod(8).

features:
- add a new keyword - "tags" - which is a comma separated list of tags
associated with that file.
- add "-E tags"; exclude files in -D output with matching tags
- add "-I tags"; only include files in -D output with matching tags
(default is to list all)

fixes:
- move all extern variable references from individual files into extern.h
- `quote' some more user-specified strings in error messages
- man page: list an options arguments in the description, and sort xrefs.

- implement -D - dump the specfile out in a manner that's parseable
by various tools
- implement -R key - remove the given key(s) from the list of keys to print
(now it's possible with "-k type -R type" to remove all keys :)
- rename ftype() to nodetype(), and make it public
- use pwcache(3) functions instead of get{gr,pw}{nam,uid}(3)
- add dump_nodes(), which provides the guts for -D.

ANSIfy, KNF, in preperation for more work.

Remove all the "#ifdef __APPLE__" code, to make this actually
build again on Darwin. The code in question is outdated.

OK'd by Wilfredo Sanchez wsanchez@MIT.EDU.

LP64 format string cleanup. Problem reported by garbled.

More format string cleanups by sommerfeld.

branches: 1.25.4;
Now that we have strsvis, use it.

Encode file names using strvis(3) when creating a specification.
Escape white space characters and `#' characters. Do the inverse
when reading specs.

This is a temporary fix until we have strzouvis(3) (or some such)
which is an improved version of strvis (taking an additional list
of characters to encode).

Fixes PR 5006 and PR 7663

branches: 1.23.4;
portability fixes:
- don't use timespecs in struct stat on non 4.4BSD systems
- don't assume that tv_sec is a time_t

Actually generate MD5 checksums, as reported in PR 7928.
Approved by Dr. Lex Wennmacher <wennmach@geo.Uni-Koeln.DE>

branches: 1.21.2;
- add -m and -i to change file flags more usefully.
- fix bugs in fileflags checking

all of the above from dr. lex wennmacher <wennmach@geo.Uni-Koeln.DE>
with some error fixes from egcs via myself.

Appease gcc-2.8.1, make printf formats more portable.

Namespace protection against MAXUID etc.

add support for 4.4BSD file flags. we take a `flags=xxx' parameter,
using the same style of naming as chflags, ls & xinstall use, plus an
additional `none' to turn off a default setting.

Use u_int32_t rather than Sys-V u_long.
Rename err() to mtree_err(), to avoid conflict with libc.

Add braces to satisfy egcs.

Add {} pairs to shut up egcs.

- use an array MAXHOSTNAMELEN+1 size to hold hostnames
- ensure hostname from gethostname() is nul-terminated in all cases
- minor KNF
- use MAXHOSTNAMELEN over various other values/defines
- be safe will buffers that hold hostnames

WARNSify, getopt returns -1 not EOF

fix printf() formats, make explicit comparisons to NULL (gcc -Wall)

Only record sizes for regular files.

branches: 1.10.4;
Rename struct timespec fields to conform to POSIX.1b

add NetBSD id's

clean up import, fix a few bugs, etc.

off_t exp.

make things a bit more consistent

latest mtree from uunet. doesn't compile, but that'll be fixed soon.
(want to have 'pseudo-virgin' sources, for when fixed fnmatch stuff
is released...)

Use *->fts_errno instead of errno after fts_read(). The fts manpage
indicates that the fts_errno will be set to the correct value, but there
are no guarentees about errno.

update for fts library change

Add RCS identifiers.

branches: 1.1.1;
Initial revision