fix const related warnings; ok millert@

rcsnum_free() -> free() cleanups. From Michael W. Bombardieri.

OK millert@

Remove xfree(), like already done for RCS. From Michael W Bombardieri,
ok mmcc

Replace <sys/param.h> with <limits.h> and other less dirty headers where
possible. Annotate <sys/param.h> lines with their current reasons. Switch
to PATH_MAX, NGROUPS_MAX, HOST_NAME_MAX+1, LOGIN_NAME_MAX, etc. Change
MIN() and MAX() to local definitions of MINIMUM() and MAXIMUM() where
sensible to avoid pulling in the pollution. These are the files confirmed
through binary verification.
ok guenther, millert, doug (helped with the verification protocol)

unsigned char for ctype. ok deraadt

simplify rcsnum parsing by using a fixed array without all the realloc
ok stsp

correctly deal with non-zero depths specified in rcsnum_cmp();

from Stefan Sperling

goodbye rcs_errno - we didnt use you and you wont be missed

- always remove the additional .0 from magic branches, even when
running the log command.

- change getlog.c so it will add any .0 for magic branches in its output
if required


less confusing for rcs_translate_tag();

Handle magic branch numbers in OpenCVS as well.

OK joris@

Slap in -b support for tag command. It was missing for a full regress test
as of now.

Fix for an off-by-one.

OK fgsch@

add support to commit modified files to branches.

Replaced strlen() with sizeof() - 1 when #define'd strings are used.

OK ray@

Since xrealloc dies on failure it is safe to directly assign to the
original pointer. Theo agrees, and so does the rest of the tree
(ssh, etc. all do this already).

Saves a bunch of variables and assignments.

OK niallo@

More comment typos from Diego Casati. Including winners like funtion, allmost,
oustside, seqencer, toghether, nessissary, etc.

Change 0x30 to more readable '0'.

OK xsa@.

general includes cleanup sweep. ok joris@ niallo@

Replace a manual xrealloc with rcsnum_setsize.

From Charles Longeau <chl at tuxfamily dot org>.

OK joris@.

Check strlcpy/strlcat return values; from usr.bin/rcs/rcsnum.c

Pull in improper memory allocation from RCS:

> Not were we allocating memory for a pointer array, it wasn't even for
> the right variable! Fixes make regress with MALLOC_OPTIONS=PR.
>
> Found by and fix worked on with Joris.

be smarter when comparing 2 revisions to each other.

commit the new opencvs code, i have been hacking on
this for the past 2 weeks now and it should go in at
the start of the hackathon so others can help out.

this code is a lot safer, smarter, faster and best of
all it is actually doing what it is suppose to do!

basic checkout, update, status, diff and commit are
working in local mode only.
there is no support for any remote setups now.

- swap a hack for a slightly nicer hack using a global variable. this avoids
adding a really nasty hack to make some upcoming rlog fixes from joris work.

ok joris@

spacing

remove a metric buttload of excessive ()
no binary change; ok ray

rcsnum_setsize() will never return non-zero; change its return type to void.

Use variable names for sizeof, remove casts in front of xmalloc,
use xcalloc and xstrdup where appropriate, and other stuff.

OK joris@

first part of supporting branches in openrcs. right now we can only
check them out. commit is not working yet, but will be soon.

tested by myself and ray@
okay ray@

- Comment fixes.
- int rcsnum_cpy() -> void rcsnum_cpy().
- Check for overflow in rcsnum_cpy().

OK niallo@

Today is Integer Overflow Prevention Day:
- Sync xmalloc.? with ssh versions.
- Change all xrealloc() calls to new API.

``I really like this.'' niallo@

Change code to match comments, fixing potential off-by-one error.
According to the comments, the lowest number should be 1.

not okay, then okay niallo@

remove some bizzare idioms; ok ray

- implement GNU-compatible handling of revisions specified as single
digits. e.g. ci -l4 or ci -u5 will work like GNU now.

ok joris@

Fix rcsnum_tostr() description.

ok joris

- don't segfault on rcsnum overflow.

ok joris@

knf;

#include's cleanup; ok joris@ niallo@.

knf

- optimise rcsnum_tostr(). we call this function a LOT and using multiple
snprintf()s in it is SLOW.

second of a few important performance improvements.

ok joris@

rcsnum_alloc() and rcsnum_cpy() no longer can fail, so don't
bother checking return values for failure.

ok xsa@

switch to xmalloc stuff, me and xsa@ agreed on this a long
time ago, but we were being held back by jfb. too bad for him.

next step is to use fatal() through out the code for unrecoverable
errors instead of trying to be all nice and fluffy and reach main() again.

ok niallo@ and xsa@

add rcsnum_dec() api call, decreases a revision number by one.
ok niallo@

Add preprocessor directives needed for usr.bin/rcs programs to compile.

ok joris@

handle magic branch numbers.

ok xsa@, jfb@

KNF;

KNF;

headers ordering (alphabetically) and cleanup, from mbalmer@ and myself;

add a couple of functions to go from branch numbers to revisions and
vice versa, and fix RCSNUM_ISBRANCH()

make rcsnum_tostr() return an empty string if the RCSNUM is NULL

add RCS error codes and a global variable to hold the last error
and make the code less verbose at the same time

- switch to rcsnum_parse() where appropriate
- make rcsnum_parse() fail if it encounters an invalid character

add rcsnum_parse() to simplify the most common case

limit the maximum length of RCS numbers to 64 nums and avoid
potential integer overflow in case of ridiculously large RCSNUMs

proper typecast to get rid of another warning

less whitespace, more pretty. ok jfb

Do not error out when calling rcsnum_aton() on the string "0", which is
a valid RCSNUM

fix memleak on rcsnum reuse
initialize new members to 0

branches: 1.1.1;
Initial revision