fix -fno-common issues; ok mortimer

Stop looking at current_cvsroot->cr_method to figure out if we're remote or not.

Instead use cvsroot_is_local() and cvsroot_is_remote().

Add missing getpw pledge

ok deraadt

opencvs can use pledge "stdio rpath wpath cpath fattr proc exec".
I have not looked into adding stricter pledges later in the code.

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)

An internal function called xrealloc() is actually a fail-hard
reallocarray()... so rename it.

unsigned char for ctype. ok deraadt

Lose unused variable cvs_msg, from Michael W Bombardieri.

Rename a bunch of variables and functions, removing the cvs_/rcs_
prefixes to reduce differences between cvs and rcs files.

"yes!" nicm

add a hash table mechanism based upon hcreate(3) but one that allows
us to maintain multiple hash tables concurrently.

immediatly start using it to keep track of what directories
we have already created and what CVS dirs we already created so
we do not recreate them when we do not need to.

we will be switching more internals to use this soon.
rejoice for cheaper lookups.

ok tobias@

unused vars

don't always re-open the same CVS/Entries file, instead keep the
last opened CVS/Entries around to operate on and close it whenever
we switch directory. gives us a small performance boost, obviously.

while doing this, switch the way we write revisions to disk by
using fwrite(3) so stuff can get written to disk in chunks
instead of writing everything line by line, another win.

with help from otto@
ok tobias@

make opencvs behave exactly like GNU cvs does when it comes to
verbosity, it's verbose by default and you can shut it up using -Q.

ok tobias@

Set up cmdp before first possible call of fatal to avoid NULL pointer
dereference in logging code. Spotted by Igor Zinovik.

ok joris

Don't limit ourselfs when it comes to arguments on command line or
arguments sent by client.

ok joris

Set umask (and cvs_umask) to system's umask in local and client mode.
Also zap umask(0);umask(mask); calls, we have cvs_umask for this.

ok joris

spacing

initial CVSROOT/modules support, only does aliases at the moment (-a).
"modules.c, hurray!" xsa@, ok tobias@

Zapped some variables which made the source harder to read (and to verify).

> Diff from Igor Zinovik

Simply the use of cmdp and cvs_command by zapping cvs_command and working
solely with cmdp. This remedies some reliability issues with invalid
commands supplied and also fixes a GNU cvs style incompatibility with
release command.

> Based on patch from Igor Zinovik

break after exit() is useless

ok ray@

Allow lines which do not specify any arguments for command (or cvs).
Sticks at GNU cvs behaviour.

OK joris@

parse global arguments (cvs) and command arguments only once, i.e. first
line of occurence. Nice side-effect: only parse arguments for current
command, not for all.

OK joris@

Print warning about invalid command in .cvsrc only if -V is in use.
OpenCVS defaults to -q, so man page should say "with -V", not "without -q
or -Q".

OK joris@

Next to ' ', '\t' is valid between command and arguments.

OK joris@

Weird, but our goal is compatibility: treat '\t' in front of keyword as '#'.

OK joris@

Usage of fgetln() instead of fgets() in .cvsrc parsing handles line
numbers better and allows longer command arguments.

OK joris@

Fixed segmentation fault that occured during .cvsrc parsing if an invalid
cvs command has been specified.

Spotted by Mathias De Belder <mathias at woei dot be> who also supplied a
patch, although I took his second suggestion.

OK joris@

Do not fatal if getpwuid() fails, instead ignore user configuration files.

OK joris@

Stick at GNU cvs behaviour with Root directive handling: it does not need
to be sent first, can occure multiple times (although this is ignored) but
should be supplied before Directory.

OK joris@

the ellipsis is not an optional argument; while here, sync the usage
and synopsis of commands

lots of good ideas by jmc@

ok jmc@

Add bounds check for cvs command arguments. Initial diff by Tobias
Stoeckmann, rearranged to "if (num_items >= BUFSIZ)" format for
readability.

I would have liked to allow our cvs client to accept as many arguments
as the system would allow, but the argument parsing code is surprising
hairy.

OK niallo@.

Revert, we dealt with the -x flag in the past. jmc suddenly
remembered we had created a separate section for compatibility
flags. Oh well, two free commits.

OK jmc@

Mention -x flag (for compatibility only). From Tobias Stoeckmann.

Input and OK jmc@

Call fatal() if extra arguments are given to "opencvs server". From
Tobias Stoeckmann.

OK niallo@

cvs_server_path is more appropriate to check than cvs_server_active.

From Tobias Stoeckmann.

OK niallo@

Note that the global -x option is kept for compatibility (for now) and has
no effect.

- we do not have -H, zap it.
- add missing `x' flag in getopt() and sort flags.
- sync usage().

Based on initial diff from Tobias Stoeckmann.

make usage() __dead; from Tobias Stoeckmann.

missing header; from Tobias Stoeckmann.

changes "Usage" to "usage"; sorts options list in cvs history;
the argument to -z is tz, not timezone; argument names to flags
should use lowercase letters; cvs's usage() output should fit
on a 80-column display

ok by jmc@, niallo@ and xsa@

fix comment; from jared r r spiegel

general includes cleanup sweep. ok joris@ niallo@

cvs_path_cat() removal since we can now easily handle that
functionality w/ xsnprintf(); Initial diff started by thib@.
OK thib@ joris@.

strtol(3) -> strtonum(3)

sort options;

in a remote setup:
do not connect to the remote server until we are sure all the options
passed to the commands are valid.

noticed by xsa@

global option -n implies -l

Permit checkout from a read-only repository using the -R global option
or using the CVSREADONLYFS environment variable.
Request/test/ok todd@. Man page bits courtesy of jmc@.

snprintf() -> cvs_path_cat()

Add support for the "Set" request.

Simplify in cvs_read_rcfile() by using cvs_path_cat();

handle SIGPIPE correctly.

first part of opencvs remote, fairly useable on existing trees
although i advise against using it on real development trees for now.

only a few commands work right so far:
- commit
- diff
- status
- log
- update (partially working)

if you feel like testing remote and run into bugs feel free to
contact me, and please include a full trace (-t).

use quiet mode (-q) by default and add a flag -V for the old "verbose"
mode. if you really need to know all the details (the old GNU'ish
behaviour), use -V or even -t for tracing.

ok joris@

be more accurate in describing what happened
when we cannot see the repository the user wants to work with.

bring back a working version of the `cvs init' command. OK joris@.

various spelling fixes

nuke cvs_repo_base variable

lets be nice and abort cvs when we find out there is
no CVSROOT directory where there is suppose to be one.

basic support for CVSROOT/config - only supporting the 'tag' and
'umask' keyword right now.

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.

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

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

OK joris@

bogus idiom

spread some fatal() here too;

strings cleanup; OK niallo@.

print complete command name in usage, not aliases.

- nuke remaining err(), errx() and warnx() instances.
- make cvs_log_init() return type void.
ok joris@.

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

use fatal() if getpwuid() fails; OK joris@.

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@

fix date handling in opencvs, this was broken since the very start.
now, when you checkout a tree with gnu cvs, opencvs no longer sees all
the files as modified, and visa versa.

this actually makes gnu cvs and opencvs dance together in working copies.

- remove strtab stuff. serves no useful purpose.

diff is from joris@, committing on his behalf because his net connection
is very dodgy right now.

knf and spacing, ok joris@

fgets() expects the 2nd arg to be an int. OK jfb@ joris@.

add global -w option and document it; ok jmc joris

use correct return codes when we exit(); ok joris

add missing [-T tmpdir] in usage(), spotted by jmc@.

handle TMPDIR environment variable as well as -T <tmpdir> global option;
Ok jfb@ joris@.

check only once for the HOME environment variable and reuse
its value (if any) to check for the .cvsrc and .cvsignore files;
ok jfb@ joris@.

respect $HOME when checking for .cvsrc; ok jfb@ joris@.

add missing -r in usage();

respect CVSREAD env. variable; ok jfb@.

KNF;

some more `if (!var)' cleanup;

rewrite of the file code. the previous one was just
a mess and not clean.
this code is much cleaner, faster, and uses less memory overall.

tested by xsa@, brad@, Michael Knudsen, and myself.
okay xsa@

remove trailing whitespaces

from deraadt@

match gnu/usr.bin/cvs error msg output on bad CVSROOT error; ok joris

allow comments in the cvsrc file.
reported by david@

missed these in previous commits.

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

Merge the cvs_cmd and cvs_cmd_info structures and add the necessary
fields to hook local versions of the commands. This needs to go in
before it gets any bigger

ok joris

add -n flag to cvs usage

ok jfb@

first bits of the `cvs -n' option; ok jfb joris

- add missing flags for the log command
- enable the -h -N -R -t options
ok joris

add `dir' arg in release cmd synopsis;

enable `release' command; tested & ok joris@.

resurrect missing parts to generate rcsdiff(1)-style diffs with the -n
flag

ok xsa

* add missing 'R' flag in diff synopsis
* add the CVS_EX_BADROOT return code to indicate that there was no
CVSROOT given or it was invalid

ok joris, xsa

sync add command usage w/ man page; joris ok

missing '-R' (force recursion) option for diff.

noticed by brad@

let the `cvs remove' command work with no file(s) specified as args;
joris ok

* fix the tag command's synopsis
* prepend __progname and the command string when displaying command
usage

check the validity of tag names passed to the tag command, and print
an abort message if the tag is invalid. also, match gnu cvs behaviour
with regards to command usage output on usage errors.

ok joris

snprintf return value check; joris ok

don't show unneeded error messages.

tell about the CVS_EX_* error codes, as we do not use sysexits.h anymore;
ok joris@.

introduce our own set of error codes used by the commands to report
what exactly went wrong in case of an error.

ok jfb@

don't include sysexits.h now we don't use those error codes
anymore.

remove EX_USAGE error codes

ok jfb@

make cvs_findcmd() available to the rest of the program

ok joris@

first round of EX_* exit codes removal; ok joris@.

make sure the server command can still be reached with the
new framework.

ok xsa@

move all the client commands to the new command framework.
eliminates a lot of duplicate code.

ok jfb@

add the string table code and enable it on startup

add the first pieces of our new command framework.

ok jfb@

make cvs_findcmd() and cvs_read_rcfile() static.

ok xsa@, jfb@

initial support for the admin command. Not all options are working yet.

ok xsa@ jfb@

exit with an error message if cvs_file_init() fails

be more consistent everywhere with `-k mode' option;
ok jfb@ jmc@ Joris Vink;

sync usage of some of the commands, and start filling the cmd_opts
field, which will later be used as the argument to getopt() instead
of hardcoding the option strings. This will allow us to use the
same handler for commands which have a local and remote version
(i.e. rdiff, rlog, rtag)

* use the global cvs_msg instead of our own
* support the use of RCS keyword expansion flags in client mode

add missing command IDs

reviewed by xsa@ and Joris Vink

support the -D option to the tag command and sync usage for the tag
command

from Joris Vink, with modifications by me

resurrect the -N and -p options to the diff command

requested and tested by brad@, thanks!

more widespread use of cvs_msg, and free it when we are done

diff by Joris Vink and me

fix the annotate usage (there is no -F flag) and add the missing bits
so the -r and -f arguments are supported, and complain if both -D
and -r are given. GNU CVS silently accepts to have both specified,
but does not produce output as a result.

the status command now supports the same arguments as its gnu cvs
counterpart (-lRv)

Improved and working import code from Joris Vink with modifications
from me (client mode only for the moment)

spelling corrections; ok jmc@

s/files/file/ in `cvs remove' usage, like in the other commands;

gnu CVS has an undocumented -s command-line option to set a variable,
so let's support it too. This will also be required to implement the
`Set' request handler.

on the way to `cvs remove'; ok jfb

print line number of warnings in .cvsrc; ok jfb@.

Permit the client to generate trace output when the -t option is given
by using the LP_TRACE level to cvs_log(). Traces are filtered by
default.

be more verbose for .cvsrc errors/warnings; ok jfb@.

support the tag command

from Joris Vink

print full path of .cvsrc in log messages; jfb@ ok.

sync SYNOPSIS and usage();
ok jfb@

support the 'annotate' command, will need some polishing

less whitespace, more pretty. ok jfb

add missing parts for support of .cvsrc

tested, tweaked by and ok djm@

add missing aliases for the status, tag, update and version commands

original patch from Peter Hessler, with additions by me

tighten vertical spacing for else; jfb ok

* don't use the import code just yet, it uses new functions that aren't
tested enough yet

* don't setup a signal handler for SIGCHLD yet
* ignore SIGPIPE
* add the getopt() argument format to the command structure

After the command is done executing, if cvs_files isn't NULL, free the
file structure

Handle the '-b' and '-x' global options even if we don't support them,
and add support for case insensitivity

Have one global hierarchy of files that are being affected. This will
allow us to build the tree in memory as well as on disk for operations
such as checkout and update. It will also allow us to write all Entries
in a single disk write and to avoid creating empty directories on updates
when pruning is requested

No more global cvs_root

missing include

Plug in the handler for `cvs status'

Move to the new API for the client-server protocol. All functions now
take a cvs root structure as parameter. This will allow for much easier
management of CVS trees that make use of multiple roots.

undefine DEBUG

Add the handler for the checkout command

add a usage string for cvs diff options

cvsignore support and simpler API for file access

branches: 1.1.1;
Initial revision

Stop looking at current_cvsroot->cr_method to figure out if we're remote or not.

Instead use cvsroot_is_local() and cvsroot_is_remote().

Add missing getpw pledge

ok deraadt

opencvs can use pledge "stdio rpath wpath cpath fattr proc exec".
I have not looked into adding stricter pledges later in the code.

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)

An internal function called xrealloc() is actually a fail-hard
reallocarray()... so rename it.

unsigned char for ctype. ok deraadt

Lose unused variable cvs_msg, from Michael W Bombardieri.

Rename a bunch of variables and functions, removing the cvs_/rcs_
prefixes to reduce differences between cvs and rcs files.

"yes!" nicm

add a hash table mechanism based upon hcreate(3) but one that allows
us to maintain multiple hash tables concurrently.

immediatly start using it to keep track of what directories
we have already created and what CVS dirs we already created so
we do not recreate them when we do not need to.

we will be switching more internals to use this soon.
rejoice for cheaper lookups.

ok tobias@

unused vars

don't always re-open the same CVS/Entries file, instead keep the
last opened CVS/Entries around to operate on and close it whenever
we switch directory. gives us a small performance boost, obviously.

while doing this, switch the way we write revisions to disk by
using fwrite(3) so stuff can get written to disk in chunks
instead of writing everything line by line, another win.

with help from otto@
ok tobias@

make opencvs behave exactly like GNU cvs does when it comes to
verbosity, it's verbose by default and you can shut it up using -Q.

ok tobias@

Set up cmdp before first possible call of fatal to avoid NULL pointer
dereference in logging code. Spotted by Igor Zinovik.

ok joris

Don't limit ourselfs when it comes to arguments on command line or
arguments sent by client.

ok joris

Set umask (and cvs_umask) to system's umask in local and client mode.
Also zap umask(0);umask(mask); calls, we have cvs_umask for this.

ok joris

spacing

initial CVSROOT/modules support, only does aliases at the moment (-a).
"modules.c, hurray!" xsa@, ok tobias@

Zapped some variables which made the source harder to read (and to verify).

> Diff from Igor Zinovik

Simply the use of cmdp and cvs_command by zapping cvs_command and working
solely with cmdp. This remedies some reliability issues with invalid
commands supplied and also fixes a GNU cvs style incompatibility with
release command.

> Based on patch from Igor Zinovik

break after exit() is useless

ok ray@

Allow lines which do not specify any arguments for command (or cvs).
Sticks at GNU cvs behaviour.

OK joris@

parse global arguments (cvs) and command arguments only once, i.e. first
line of occurence. Nice side-effect: only parse arguments for current
command, not for all.

OK joris@

Print warning about invalid command in .cvsrc only if -V is in use.
OpenCVS defaults to -q, so man page should say "with -V", not "without -q
or -Q".

OK joris@

Next to ' ', '\t' is valid between command and arguments.

OK joris@

Weird, but our goal is compatibility: treat '\t' in front of keyword as '#'.

OK joris@

Usage of fgetln() instead of fgets() in .cvsrc parsing handles line
numbers better and allows longer command arguments.

OK joris@

Fixed segmentation fault that occured during .cvsrc parsing if an invalid
cvs command has been specified.

Spotted by Mathias De Belder <mathias at woei dot be> who also supplied a
patch, although I took his second suggestion.

OK joris@

Do not fatal if getpwuid() fails, instead ignore user configuration files.

OK joris@

Stick at GNU cvs behaviour with Root directive handling: it does not need
to be sent first, can occure multiple times (although this is ignored) but
should be supplied before Directory.

OK joris@

the ellipsis is not an optional argument; while here, sync the usage
and synopsis of commands

lots of good ideas by jmc@

ok jmc@

Add bounds check for cvs command arguments. Initial diff by Tobias
Stoeckmann, rearranged to "if (num_items >= BUFSIZ)" format for
readability.

I would have liked to allow our cvs client to accept as many arguments
as the system would allow, but the argument parsing code is surprising
hairy.

OK niallo@.

Revert, we dealt with the -x flag in the past. jmc suddenly
remembered we had created a separate section for compatibility
flags. Oh well, two free commits.

OK jmc@

Mention -x flag (for compatibility only). From Tobias Stoeckmann.

Input and OK jmc@

Call fatal() if extra arguments are given to "opencvs server". From
Tobias Stoeckmann.

OK niallo@

cvs_server_path is more appropriate to check than cvs_server_active.

From Tobias Stoeckmann.

OK niallo@

Note that the global -x option is kept for compatibility (for now) and has
no effect.

- we do not have -H, zap it.
- add missing `x' flag in getopt() and sort flags.
- sync usage().

Based on initial diff from Tobias Stoeckmann.

make usage() __dead; from Tobias Stoeckmann.

missing header; from Tobias Stoeckmann.

changes "Usage" to "usage"; sorts options list in cvs history;
the argument to -z is tz, not timezone; argument names to flags
should use lowercase letters; cvs's usage() output should fit
on a 80-column display

ok by jmc@, niallo@ and xsa@

fix comment; from jared r r spiegel

general includes cleanup sweep. ok joris@ niallo@

cvs_path_cat() removal since we can now easily handle that
functionality w/ xsnprintf(); Initial diff started by thib@.
OK thib@ joris@.

strtol(3) -> strtonum(3)

sort options;

in a remote setup:
do not connect to the remote server until we are sure all the options
passed to the commands are valid.

noticed by xsa@

global option -n implies -l

Permit checkout from a read-only repository using the -R global option
or using the CVSREADONLYFS environment variable.
Request/test/ok todd@. Man page bits courtesy of jmc@.

snprintf() -> cvs_path_cat()

Add support for the "Set" request.

Simplify in cvs_read_rcfile() by using cvs_path_cat();

handle SIGPIPE correctly.

first part of opencvs remote, fairly useable on existing trees
although i advise against using it on real development trees for now.

only a few commands work right so far:
- commit
- diff
- status
- log
- update (partially working)

if you feel like testing remote and run into bugs feel free to
contact me, and please include a full trace (-t).

use quiet mode (-q) by default and add a flag -V for the old "verbose"
mode. if you really need to know all the details (the old GNU'ish
behaviour), use -V or even -t for tracing.

ok joris@

be more accurate in describing what happened
when we cannot see the repository the user wants to work with.

bring back a working version of the `cvs init' command. OK joris@.

various spelling fixes

nuke cvs_repo_base variable

lets be nice and abort cvs when we find out there is
no CVSROOT directory where there is suppose to be one.

basic support for CVSROOT/config - only supporting the 'tag' and
'umask' keyword right now.

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.

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

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

OK joris@

bogus idiom

spread some fatal() here too;

strings cleanup; OK niallo@.

print complete command name in usage, not aliases.

- nuke remaining err(), errx() and warnx() instances.
- make cvs_log_init() return type void.
ok joris@.

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

use fatal() if getpwuid() fails; OK joris@.

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@

fix date handling in opencvs, this was broken since the very start.
now, when you checkout a tree with gnu cvs, opencvs no longer sees all
the files as modified, and visa versa.

this actually makes gnu cvs and opencvs dance together in working copies.

- remove strtab stuff. serves no useful purpose.

diff is from joris@, committing on his behalf because his net connection
is very dodgy right now.

knf and spacing, ok joris@

fgets() expects the 2nd arg to be an int. OK jfb@ joris@.

add global -w option and document it; ok jmc joris

use correct return codes when we exit(); ok joris

add missing [-T tmpdir] in usage(), spotted by jmc@.

handle TMPDIR environment variable as well as -T <tmpdir> global option;
Ok jfb@ joris@.

check only once for the HOME environment variable and reuse
its value (if any) to check for the .cvsrc and .cvsignore files;
ok jfb@ joris@.

respect $HOME when checking for .cvsrc; ok jfb@ joris@.

add missing -r in usage();

respect CVSREAD env. variable; ok jfb@.

KNF;

some more `if (!var)' cleanup;

rewrite of the file code. the previous one was just
a mess and not clean.
this code is much cleaner, faster, and uses less memory overall.

tested by xsa@, brad@, Michael Knudsen, and myself.
okay xsa@

remove trailing whitespaces

from deraadt@

match gnu/usr.bin/cvs error msg output on bad CVSROOT error; ok joris

allow comments in the cvsrc file.
reported by david@

missed these in previous commits.

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

Merge the cvs_cmd and cvs_cmd_info structures and add the necessary
fields to hook local versions of the commands. This needs to go in
before it gets any bigger

ok joris

add -n flag to cvs usage

ok jfb@

first bits of the `cvs -n' option; ok jfb joris

- add missing flags for the log command
- enable the -h -N -R -t options
ok joris

add `dir' arg in release cmd synopsis;

enable `release' command; tested & ok joris@.

resurrect missing parts to generate rcsdiff(1)-style diffs with the -n
flag

ok xsa

* add missing 'R' flag in diff synopsis
* add the CVS_EX_BADROOT return code to indicate that there was no
CVSROOT given or it was invalid

ok joris, xsa

sync add command usage w/ man page; joris ok

missing '-R' (force recursion) option for diff.

noticed by brad@

let the `cvs remove' command work with no file(s) specified as args;
joris ok

* fix the tag command's synopsis
* prepend __progname and the command string when displaying command
usage

check the validity of tag names passed to the tag command, and print
an abort message if the tag is invalid. also, match gnu cvs behaviour
with regards to command usage output on usage errors.

ok joris

snprintf return value check; joris ok

don't show unneeded error messages.

tell about the CVS_EX_* error codes, as we do not use sysexits.h anymore;
ok joris@.

introduce our own set of error codes used by the commands to report
what exactly went wrong in case of an error.

ok jfb@

don't include sysexits.h now we don't use those error codes
anymore.

remove EX_USAGE error codes

ok jfb@

make cvs_findcmd() available to the rest of the program

ok joris@

first round of EX_* exit codes removal; ok joris@.

make sure the server command can still be reached with the
new framework.

ok xsa@

move all the client commands to the new command framework.
eliminates a lot of duplicate code.

ok jfb@

add the string table code and enable it on startup

add the first pieces of our new command framework.

ok jfb@

make cvs_findcmd() and cvs_read_rcfile() static.

ok xsa@, jfb@

initial support for the admin command. Not all options are working yet.

ok xsa@ jfb@

exit with an error message if cvs_file_init() fails

be more consistent everywhere with `-k mode' option;
ok jfb@ jmc@ Joris Vink;

sync usage of some of the commands, and start filling the cmd_opts
field, which will later be used as the argument to getopt() instead
of hardcoding the option strings. This will allow us to use the
same handler for commands which have a local and remote version
(i.e. rdiff, rlog, rtag)

* use the global cvs_msg instead of our own
* support the use of RCS keyword expansion flags in client mode

add missing command IDs

reviewed by xsa@ and Joris Vink

support the -D option to the tag command and sync usage for the tag
command

from Joris Vink, with modifications by me

resurrect the -N and -p options to the diff command

requested and tested by brad@, thanks!

more widespread use of cvs_msg, and free it when we are done

diff by Joris Vink and me

fix the annotate usage (there is no -F flag) and add the missing bits
so the -r and -f arguments are supported, and complain if both -D
and -r are given. GNU CVS silently accepts to have both specified,
but does not produce output as a result.

the status command now supports the same arguments as its gnu cvs
counterpart (-lRv)

Improved and working import code from Joris Vink with modifications
from me (client mode only for the moment)

spelling corrections; ok jmc@

s/files/file/ in `cvs remove' usage, like in the other commands;

gnu CVS has an undocumented -s command-line option to set a variable,
so let's support it too. This will also be required to implement the
`Set' request handler.

on the way to `cvs remove'; ok jfb

print line number of warnings in .cvsrc; ok jfb@.

Permit the client to generate trace output when the -t option is given
by using the LP_TRACE level to cvs_log(). Traces are filtered by
default.

be more verbose for .cvsrc errors/warnings; ok jfb@.

support the tag command

from Joris Vink

print full path of .cvsrc in log messages; jfb@ ok.

sync SYNOPSIS and usage();
ok jfb@

support the 'annotate' command, will need some polishing

less whitespace, more pretty. ok jfb

add missing parts for support of .cvsrc

tested, tweaked by and ok djm@

add missing aliases for the status, tag, update and version commands

original patch from Peter Hessler, with additions by me

tighten vertical spacing for else; jfb ok

* don't use the import code just yet, it uses new functions that aren't
tested enough yet

* don't setup a signal handler for SIGCHLD yet
* ignore SIGPIPE
* add the getopt() argument format to the command structure

After the command is done executing, if cvs_files isn't NULL, free the
file structure

Handle the '-b' and '-x' global options even if we don't support them,
and add support for case insensitivity

Have one global hierarchy of files that are being affected. This will
allow us to build the tree in memory as well as on disk for operations
such as checkout and update. It will also allow us to write all Entries
in a single disk write and to avoid creating empty directories on updates
when pruning is requested

No more global cvs_root

missing include

Plug in the handler for `cvs status'

Move to the new API for the client-server protocol. All functions now
take a cvs root structure as parameter. This will allow for much easier
management of CVS trees that make use of multiple roots.

undefine DEBUG

Add the handler for the checkout command

add a usage string for cvs diff options

cvsignore support and simpler API for file access

branches: 1.1.1;
Initial revision