[xemacs-hg @ 2001-09-20 06:28:42 by ben] The great integral types renaming. The purpose of this is to rationalize the names used for various integral types, so that they match their intended uses and follow consist conventions, and eliminate types that were not semantically different from each other. The conventions are: -- All integral types that measure quantities of anything are signed. Some people disagree vociferously with this, but their arguments are mostly theoretical, and are vastly outweighed by the practical headaches of mixing signed and unsigned values, and more importantly by the far increased likelihood of inadvertent bugs: Because of the broken "viral" nature of unsigned quantities in C (operations involving mixed signed/unsigned are done unsigned, when exactly the opposite is nearly always wanted), even a single error in declaring a quantity unsigned that should be signed, or even the even more subtle error of comparing signed and unsigned values and forgetting the necessary cast, can be catastrophic, as comparisons will yield wrong results. -Wsign-compare is turned on specifically to catch this, but this tends to result in a great number of warnings when mixing signed and unsigned, and the casts are annoying. More has been written on this elsewhere. -- All such quantity types just mentioned boil down to EMACS_INT, which is 32 bits on 32-bit machines and 64 bits on 64-bit machines. This is guaranteed to be the same size as Lisp objects of type `int', and (as far as I can tell) of size_t (unsigned!) and ssize_t. The only type below that is not an EMACS_INT is Hashcode, which is an unsigned value of the same size as EMACS_INT. -- Type names should be relatively short (no more than 10 characters or so), with the first letter capitalized and no underscores if they can at all be avoided. -- "count" == a zero-based measurement of some quantity. Includes sizes, offsets, and indexes. -- "bpos" == a one-based measurement of a position in a buffer. "Charbpos" and "Bytebpos" count text in the buffer, rather than bytes in memory; thus Bytebpos does not directly correspond to the memory representation. Use "Membpos" for this. -- "Char" refers to internal-format characters, not to the C type "char", which is really a byte. -- For the actual name changes, see the script below. I ran the following script to do the conversion. (NOTE: This script is idempotent. You can safely run it multiple times and it will not screw up previous results -- in fact, it will do nothing if nothing has changed. Thus, it can be run repeatedly as necessary to handle patches coming in from old workspaces, or old branches.) There are two tags, just before and just after the change: `pre-integral-type-rename' and `post-integral-type-rename'. When merging code from the main trunk into a branch, the best thing to do is first merge up to `pre-integral-type-rename', then apply the script and associated changes, then merge from `post-integral-type-change' to the present. (Alternatively, just do the merging in one operation; but you may then have a lot of conflicts needing to be resolved by hand.) Script `fixtypes.sh' follows: ----------------------------------- cut ------------------------------------ files="*.[ch] s/*.h m/*.h config.h.in ../configure.in Makefile.in.in ../lib-src/*.[ch] ../lwlib/*.[ch]" gr Memory_Count Bytecount $files gr Lstream_Data_Count Bytecount $files gr Element_Count Elemcount $files gr Hash_Code Hashcode $files gr extcount bytecount $files gr bufpos charbpos $files gr bytind bytebpos $files gr memind membpos $files gr bufbyte intbyte $files gr Extcount Bytecount $files gr Bufpos Charbpos $files gr Bytind Bytebpos $files gr Memind Membpos $files gr Bufbyte Intbyte $files gr EXTCOUNT BYTECOUNT $files gr BUFPOS CHARBPOS $files gr BYTIND BYTEBPOS $files gr MEMIND MEMBPOS $files gr BUFBYTE INTBYTE $files gr MEMORY_COUNT BYTECOUNT $files gr LSTREAM_DATA_COUNT BYTECOUNT $files gr ELEMENT_COUNT ELEMCOUNT $files gr HASH_CODE HASHCODE $files ----------------------------------- cut ------------------------------------ `fixtypes.sh' is a Bourne-shell script; it uses 'gr': ----------------------------------- cut ------------------------------------ #!/bin/sh # Usage is like this: # gr FROM TO FILES ... # globally replace FROM with TO in FILES. FROM and TO are regular expressions. # backup files are stored in the `backup' directory. from="$1" to="$2" shift 2 echo ${1+"$@"} | xargs global-replace "s/$from/$to/g" ----------------------------------- cut ------------------------------------ `gr' in turn uses a Perl script to do its real work, `global-replace', which follows: ----------------------------------- cut ------------------------------------ : #-*- Perl -*- ### global-modify --- modify the contents of a file by a Perl expression ## Copyright (C) 1999 Martin Buchholz. ## Copyright (C) 2001 Ben Wing. ## Authors: Martin Buchholz <martin@xemacs.org>, Ben Wing <ben@xemacs.org> ## Maintainer: Ben Wing <ben@xemacs.org> ## Current Version: 1.0, May 5, 2001 # This program is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2, or (at your option) # any later version. # # This program 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 XEmacs; see the file COPYING. If not, write to the Free # Software Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA # 02111-1307, USA. eval 'exec perl -w -S $0 ${1+"$@"}' if 0; use strict; use FileHandle; use Carp; use Getopt::Long; use File::Basename; (my $myName = $0) =~ s@.*/@@; my $usage=" Usage: $myName [--help] [--backup-dir=DIR] [--line-mode] [--hunk-mode] PERLEXPR FILE ... Globally modify a file, either line by line or in one big hunk. Typical usage is like this: [with GNU print, GNU xargs: guaranteed to handle spaces, quotes, etc. in file names] find . -name '*.[ch]' -print0 | xargs -0 $0 's/\bCONST\b/const/g'\n [with non-GNU print, xargs] find . -name '*.[ch]' -print | xargs $0 's/\bCONST\b/const/g'\n The file is read in, either line by line (with --line-mode specified) or in one big hunk (with --hunk-mode specified; it's the default), and the Perl expression is then evalled with \$_ set to the line or hunk of text, including the terminating newline if there is one. It should destructively modify the value there, storing the changed result in \$_. Files in which any modifications are made are backed up to the directory specified using --backup-dir, or to `backup' by default. To disable this, use --backup-dir= with no argument. Hunk mode is the default because it is MUCH MUCH faster than line-by-line. Use line-by-line only when it matters, e.g. you want to do a replacement only once per line (the default without the `g' argument). Conversely, when using hunk mode, *ALWAYS* use `g'; otherwise, you will only make one replacement in the entire file! "; my %options = (); $Getopt::Long::ignorecase = 0; &GetOptions ( \%options, 'help', 'backup-dir=s', 'line-mode', 'hunk-mode', ); die $usage if $options{"help"} or @ARGV <= 1; my $code = shift; die $usage if grep (-d || ! -w, @ARGV); sub SafeOpen { open ((my $fh = new FileHandle), $_[0]); confess "Can't open $_[0]: $!" if ! defined $fh; return $fh; } sub SafeClose { close $_[0] or confess "Can't close $_[0]: $!"; } sub FileContents { my $fh = SafeOpen ("< $_[0]"); my $olddollarslash = $/; local $/ = undef; my $contents = <$fh>; $/ = $olddollarslash; return $contents; } sub WriteStringToFile { my $fh = SafeOpen ("> $_[0]"); binmode $fh; print $fh $_[1] or confess "$_[0]: $!\n"; SafeClose $fh; } foreach my $file (@ARGV) { my $changed_p = 0; my $new_contents = ""; if ($options{"line-mode"}) { my $fh = SafeOpen $file; while (<$fh>) { my $save_line = $_; eval $code; $changed_p = 1 if $save_line ne $_; $new_contents .= $_; } } else { my $orig_contents = $_ = FileContents $file; eval $code; if ($_ ne $orig_contents) { $changed_p = 1; $new_contents = $_; } } if ($changed_p) { my $backdir = $options{"backup-dir"}; $backdir = "backup" if !defined ($backdir); if ($backdir) { my ($name, $path, $suffix) = fileparse ($file, ""); my $backfulldir = $path . $backdir; my $backfile = "$backfulldir/$name"; mkdir $backfulldir, 0755 unless -d $backfulldir; print "modifying $file (original saved in $backfile)\n"; rename $file, $backfile; } WriteStringToFile ($file, $new_contents); } } ----------------------------------- cut ------------------------------------ In addition to those programs, I needed to fix up a few other things, particularly relating to the duplicate definitions of types, now that some types merged with others. Specifically: 1. in lisp.h, removed duplicate declarations of Bytecount. The changed code should now look like this: (In each code snippet below, the first and last lines are the same as the original, as are all lines outside of those lines. That allows you to locate the section to be replaced, and replace the stuff in that section, verifying that there isn't anything new added that would need to be kept.) --------------------------------- snip ------------------------------------- /* Counts of bytes or chars */ typedef EMACS_INT Bytecount; typedef EMACS_INT Charcount; /* Counts of elements */ typedef EMACS_INT Elemcount; /* Hash codes */ typedef unsigned long Hashcode; /* ------------------------ dynamic arrays ------------------- */ --------------------------------- snip ------------------------------------- 2. in lstream.h, removed duplicate declaration of Bytecount. Rewrote the comment about this type. The changed code should now look like this: --------------------------------- snip ------------------------------------- #endif /* The have been some arguments over the what the type should be that specifies a count of bytes in a data block to be written out or read in, using Lstream_read(), Lstream_write(), and related functions. Originally it was long, which worked fine; Martin "corrected" these to size_t and ssize_t on the grounds that this is theoretically cleaner and is in keeping with the C standards. Unfortunately, this practice is horribly error-prone due to design flaws in the way that mixed signed/unsigned arithmetic happens. In fact, by doing this change, Martin introduced a subtle but fatal error that caused the operation of sending large mail messages to the SMTP server under Windows to fail. By putting all values back to be signed, avoiding any signed/unsigned mixing, the bug immediately went away. The type then in use was Lstream_Data_Count, so that it be reverted cleanly if a vote came to that. Now it is Bytecount. Some earlier comments about why the type must be signed: This MUST BE SIGNED, since it also is used in functions that return the number of bytes actually read to or written from in an operation, and these functions can return -1 to signal error. Note that the standard Unix read() and write() functions define the count going in as a size_t, which is UNSIGNED, and the count going out as an ssize_t, which is SIGNED. This is a horrible design flaw. Not only is it highly likely to lead to logic errors when a -1 gets interpreted as a large positive number, but operations are bound to fail in all sorts of horrible ways when a number in the upper-half of the size_t range is passed in -- this number is unrepresentable as an ssize_t, so code that checks to see how many bytes are actually written (which is mandatory if you are dealing with certain types of devices) will get completely screwed up. --ben */ typedef enum lstream_buffering --------------------------------- snip ------------------------------------- 3. in dumper.c, there are four places, all inside of switch() statements, where XD_BYTECOUNT appears twice as a case tag. In each case, the two case blocks contain identical code, and you should *REMOVE THE SECOND* and leave the first.

.TH GNUSERV 1 "" "XEmacs Server"
.UC 4
.SH NAME
gnuserv, gnuclient \- Server and Clients for XEmacs
.SH SYNOPSIS
.B gnuclient
[-nw] [-display display] [-q] [-v] [-l library] [-batch] [-f function] [-eval form] 
[-h hostname] [-p port] [-r remote-pathname] [[+line] file] ...
.br
.B gnudoit [-q] 
form
.br
.B gnuserv
.br
.B gnuattach   
Removed as of gnuserv 3.x
.SH DESCRIPTION

.PP
\fIgnuclient\fP allows the user to request a running XEmacs process to
edit the named files or directories and/or evaluate lisp forms.
Depending on your environment, it can be an X frame or a TTY frame.
One typical use for this is with a dialup connection to a machine on
which an XEmacs process is currently running.
.PP
\fIgnudoit\fP is a shell script frontend to ``gnuclient -batch -eval form''.
Its use is deprecated. Try to get used to calling gnuclient directly.
.PP
\fIgnuserv\fP is the server program that is set running by XEmacs to
handle all incoming and outgoing requests. It is not usually invoked
directly, but is started from XEmacs by loading the \fIgnuserv\fP
package and evaluating the Lisp form (gnuserv-start).
.PP
\fIgnuattach\fP no longer exists. Its functionality has been replaced by
\fIgnuclient -nw\fP.
.SH OPTIONS
.PP 
\fIgnuclient\fP supports as much of the command line options of Emacs as
makes sense in this context. In addition it adds a few of its own. 
.br
Options with long names can also be specified using a double
hyphen instead of a single one.
.TP 8
.BI \-nw
This option makes \fIgnuclient\fP act as a frontend such that XEmacs
can attach to the current TTY. XEmacs will then open a new TTY frame.
The effect is similar to having started a new XEmacs on this TTY with
the ``-nw'' option. It currently only works if XEmacs is running on
the same machine as gnuclient. This is the default if the `DISPLAY'
environment variable is not set.
.TP 8
.BI \-display " display, " \--display " display" 
If this option is given or the `DISPLAY' environment variable is set
then gnuclient will tell XEmacs to edit files in a frame on the
specified X device.
.TP 8
.BI \-q
This option informs \fIgnuclient\fP to exit once connection has been
made with the XEmacs process.  Normally \fIgnuclient\fP waits until
all of the files on the command line have been finished with (their
buffers killed) by the XEmacs process, and all the forms have been
evaluated.
.TP 8
.BI \-v
When this option is specified \fIgnuclient\fP will request for the
specified files to be viewed instead of edited.
.TP 8
.BI \-l " library"
Tell Emacs to load the specified library.
.TP 8
.BI \-batch
Tell Emacs not to open any frames. Just load libraries and evaluate
lisp code.  If no files to execute, functions to call or forms to eval 
are given using the
.BR \-l ,
.BR \-f ,
or
.B \-eval
options, then forms to eval are read from STDIN.
.TP 8
.BI \-f " function," 
Make Emacs execute the lisp function.
.TP 8
.BI \-eval " form"
Make Emacs execute the lisp form.
.TP 8
.BI \-h " hostname"
Used only with Internet-domain sockets, this option specifies the host
machine which should be running \fIgnuserv\fP. If this option is not
specified then the value of the environment variable GNU_HOST is used
if set. If no hostname is specified, and the GNU_HOST variable is not
set, an internet connection will not be attempted. N\.B.:
\fIgnuserv\fP does NOT allow internet connections unless XAUTH
authentication is used or the GNU_SECURE variable has been specified
and points at a file listing all trusted hosts. (See SECURITY below.)

.br
Note that an internet address may be specified instead of a hostname
which can speed up connections to the server by quite a bit,
especially if the client machine is running YP.

.br
Note also that a hostname of \fBunix\fP can be used to specify that
the connection to the server should use a Unix-domain socket (if
supported) rather than an Internet-domain socket.
.TP 8
.BI \-p " port"
Used only with Internet-domain sockets, this option specifies the
service port used to communicate between server and clients.  If this
option is not specified, then the value of the environment variable
GNU_PORT is used, if set, otherwise a service called ``gnuserv'' is
looked up in the services database.  Finally, if no other value can be
found for the port, then a default port is used which is usually 21490
+ uid.
.br
Note that since \fIgnuserv\fP doesn't allow command-line options, the port for
it will have to be specified via one of the alternative methods.
.TP 8
.BI \-r " pathname"
Used only with Internet-domain sockets, the pathname argument may be
needed to inform XEmacs how to reach the root directory of a remote
machine.  \fIgnuclient\fP prepends this string to each path argument
given.  For example, if you were trying to edit a file on a client
machine called otter, whose root directory was accessible from the
server machine via the path /net/otter, then this argument should be
set to '/net/otter'.  If this option is omitted, then the value is
taken from the environment variable GNU_NODE, if set, or the empty
string otherwise.
.TP 8
.BI "[+n] file"
This is the path of the file to be edited.  If the file is a directory, then
the directory browsers dired or monkey are usually invoked instead.
The cursor is put at line number 'n' if specified.

.SH SETUP
\fIgnuserv\fP is packaged standardly with recent versions of XEmacs.
Therefore, you should be able to start the server simply by evaluating
the XEmacs Lisp form (gnuserv-start), or equivalently by typing
`M-x gnuserv-start'.

.SH CONFIGURATION
The behavior of this suite of program is mostly controlled on the lisp 
side in Emacs and its behavior can be customized to a large extent.
Type `M-x customize-group RET gnuserv RET' for easy access. More
documentation can be found in the file `gnuserv.el'

.SH EXAMPLE
.RS 4
gnuclient -q -f mh-smail
.br
gnuclient -h cuckoo -r /ange@otter: /tmp/*
.br
gnuclient -nw ../src/listproc.c
.RE
.br

.br
More examples and sample wrapper scripts are provided in the
etc/gnuserv directory of the Emacs installation.


.SH SYSV IPC
SysV IPC is used to communicate between \fIgnuclient\fP and
\fIgnuserv\fP if the symbol SYSV_IPC is defined at the top of
gnuserv.h. This is incompatible with both Unix-domain and
Internet-domain socket communication as described below. A file called
/tmp/gsrv??? is created as a key for the message queue, and if removed
will cause the communication between server and client to fail until
the server is restarted.
.SH UNIX-DOMAIN SOCKETS
A Unix-domain socket is used to communicate between \fIgnuclient\fP
and \fIgnuserv\fP if the symbol UNIX_DOMAIN_SOCKETS is defined at the
top of gnuserv.h.  A file called /tmp/gsrvdir????/gsrv is created for
communication.  If the symbol USE_TMPDIR is set at the top of gnuserv.h,
$TMPDIR, when set, is used instead of /tmp.  If that file is deleted,
or TMPDIR has different values for the server and the client, communication
between server and client will fail.  Only the user running gnuserv will be
able to connect to the socket.
.SH INTERNET-DOMAIN SOCKETS
Internet-domain sockets are used to communicate between
\fIgnuclient\fP and \fIgnuserv\fP if the symbol
INTERNET_DOMAIN_SOCKETS is defined at the top of gnuserv.h. Both
Internet-domain and Unix-domain sockets can be used at the same
time. If a hostname is specified via -h or via the GNU_HOST
environment variable, \fIgnuclient\fP establish connections using an
internet domain socket. If not, a local connection is attempted via
either a unix-domain socket or SYSV IPC.
.SH SECURITY
Using Internet-domain sockets, a more robust form of security is
needed that wasn't necessary with either Unix-domain sockets or SysV
IPC. Currently, two authentication protocols are supported to provide
this: MIT-MAGIC-COOKIE-1 (based on the X11 xauth(1) program) and a
simple host-based access control mechanism, hereafter called
GNUSERV-1. The GNUSERV-1 protocol is always available, whereas support
for MIT-MAGIC-COOKIE-1 may or may not have been enabled (via a #define
at the top of gnuserv.h) at compile-time.
.PP
\fIgnuserv\fP, using GNUSERV-1, performs a limited form of access
control at the machine level. By default no internet-domain socket is
opened.  If the variable GNU_SECURE can be found in \fIgnuserv\fP's
environment, and it names a readable filename, then this file is
opened and assumed to be a list of hosts, one per line, from which the
server will allow requests. Connections from any other host will be
rejected. Even the machine on which \fIgnuserv\fP is running is not
permitted to make connections via the internet socket unless its
hostname is explicitly specified in this file.  Note that a host may
be either a numeric IP address or a hostname, and that
.I any
user on an approved host may connect to your gnuserv and execute arbitrary
elisp (e.g., delete all your files).
If this file contains a lot of
hostnames then the server may take quite a time to start up.
.PP
When the MIT-MAGIC-COOKIE-1 protocol is enabled, an internet socket
\fIis\fP opened by default. \fIgnuserv\fP will accept a connection from
any host, and will wait for a "magic cookie" (essentially, a password)
to be presented by the client. If the client doesn't present the
cookie, or if the cookie is wrong, the authentication of the client is
considered to have failed. At this point. \fIgnuserv\fP falls back to
the GNUSERV-1 protocol; If the client is calling from a host listed in
the GNU_SECURE file, the connection will be accepted, otherwise it
will be rejected. 
.TP 4
.I  Using MIT-MAGIC-COOKIE-1 authentication
When the \fIgnuserv\fP server is started, it looks for a cookie
defined for display 999 on the machine where it is running. If the
cookie is found, it will be stored for use as the authentication
cookie. These cookies are defined in an authorization file (usually
~/.Xauthority) that is manipulated by the X11 xauth(1) program. For
example, a machine "kali" which runs an emacs that invokes
\fIgnuserv\fP should respond as follows (at the shell prompt) when set
up correctly.
.PP
.RS 8
kali% xauth list
.br
GS65.SP.CS.CMU.EDU:0  MIT-MAGIC-COOKIE-1  11223344
.br
KALI.FTM.CS.CMU.EDU:999  MIT-MAGIC-COOKIE-1  1234
.RE
.PP
.RS 4
In the above case, the authorization file defines two cookies. The
second one, defined for screen 999 on the server machine, is used for
gnuserv authentication. 
.PP
On the client machine's side, the authorization file must contain an
identical line, specifying the 
.I server's 
cookie. In other words, on a machine "foobar" which wishes to connect
to "kali,"  the `xauth list' output should contain the line:
.PP
.RS 4
KALI.FTM.CS.CMU.EDU:999  MIT-MAGIC-COOKIE-1  1234
.RE
.PP
For more information on authorization files, take a look at the
xauth(1X11) man page, or invoke xauth interactively (without any
arguments) and type "help" at the prompt. Remember that case in the
name of the authorization protocol (i.e.`MIT-MAGIC-COOKIE-1') 
.I is
significant!
.RE


.SH ENVIRONMENT
.PP
.TP 8
.B DISPLAY
Default X device to put edit frame.

.SH FILES
.PP
.TP 8
.B /tmp/gsrv???
(SYSV_IPC only)
.TP 8
.B /tmp/gsrvdir???/gsrv
(unix domain sockets only)
.TP 8
.B ~/.emacs
XEmacs customization file, see xemacs(1).
.SH SEE ALSO
.PP
.TP 8
xauth(1X11), Xsecurity(1X11), gnuserv.el
.SH BUGS
.PP 
NULs occurring in result strings don't get passed back to gnudoit properly.

.SH AUTHOR.
Andy Norman (ange@hplb.hpl.hp.com), based heavily upon
etc/emacsclient.c, etc/server.c and lisp/server.el from the GNU Emacs
18.52 distribution.  Various modifications from Bob Weiner (weiner@mot.com),
Darrell Kindred (dkindred@cmu.edu), Arup Mukherjee (arup@cmu.edu), Ben
Wing (ben@xemacs.org) and Hrvoje Niksic (hniksic@xemacs.org).