Mercurial > hg > xemacs-beta
view nt/README @ 665:fdefd0186b75
[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.
| author | ben |
|---|---|
| date | Thu, 20 Sep 2001 06:31:11 +0000 |
| parents | 576fb035e263 |
| children | 943eaba38521 |
line wrap: on
line source
Building and Installing XEmacs on Windows 95/98/NT/2000 -*- mode:outline -*- David Hobley Marc Paquette Jonathan Harris Ben Wing This is a port of XEmacs to Windows 95/98/NT/2000. If you are looking for a port of GNU Emacs, see http://www.cs.washington.edu/homes/voelker/ntemacs.html. NT 3.51 or later is required for building on Windows NT. Note that the developers typically use NT 4.0 and Windows 2000, and there may possibly be problems under Windows 95/98 and NT 3.51. If so, please report them to xemacs-nt@xemacs.org; we are committed to maintaining compatibility with all systems listed. * Required tools and sources ============================ 1. You will need Visual C++ V4.0 or later to compile everything. Personally we have tested V4.0, V4.2, V5.0 and v6.0. Note that Visual C++ assumes that the environment variables INCLUDE and LIB are set to specify the location of the includes and libraries. Your PATH environment variable also needs to include the DevStudio vc\bin and sharedide\bin directories. Visual C++ V5.0 and later install a batch file called vcvars32.bat in c:\Program Files\DevStudio\VC\bin\ (or wherever you installed it) that you can run before building to set up all of these environment variables. Alternatively, you can choose at setup time to have these environment variables automatically set up in the registry, which is generally a good idea. 2. Grab the latest XEmacs source from ftp://ftp.xemacs.org/pub/xemacs/ or one of its mirrors listed at http://www.xemacs.org/Download/index.html. (NOTE: If you are behind a firewall and have problems with FTP access, the URL http://ftp.xemacs.org/pub/xemacs/ works just as well.) You'll also need the packages. You probably want to get the unified packages bundle from ftp://ftp.xemacs.org/pub/xemacs/packages/xemacs-sumo.tar.gz If you are building with international support, you also need ftp://ftp.xemacs.org/pub/xemacs/packages/xemacs-mule-sumo.tar.gz Although we don't recommend it, you can also retrieve just the packages you really need if you have an extremely slow net connection or are very short on disk space. You can find the various packages in ftp://ftp.xemacs.org/pub/xemacs/packages/. You will need the xemacs-base package (and mule-base, if building with international support). You'll also need the texinfo package unless you have a copy of makeinfo.exe on your machine. If you want to download additional or updated packages from within XEmacs you'll need the efs, dired and vm packages. You'll probably also want at least the edit-utils, text-modes, fsf-compat, cc-mode, prog-modes and xemacs-devel packages. Unpack the packages into "x:\your\choice\XEmacs\xemacs-packages", for example "c:\Program Files\XEmacs\xemacs-packages". 3. At this point you can choose to build for X and/or for Win32 native GUI. If you only want to build for the Win32 native GUI then skip the next section. ** Extra tools and sources required for X NOTE: XEmacs has not been tested with X support under the native Windows build for a long, long time! It may not even compile any more. If you are interested in X support, you're better off compiling the Cygwin version of XEmacs, which can handle both Win32 native and X frames (in the same binary, in fact, but not at the same time), and is actively tested with X support. If you want support for X you will also need: 1. An X server. XEmacs has been tested and runs well under MI/X, available from: http://www.microimages.com/mix/. (International aka "Mule" support even works under this X server!) Unfortunately, this is not free, but is trialware; you have to pay $25 if you want to use it for more than 15 days. XEmacs also runs (barely) under the free XWin server that comes as part of the Cygwin XFree86 package, available at ftp://sources.redhat.com/pub/cygwin/xfree/ or numerous mirrors, such as ftp://ftp.freesoftware.com/pub/sourceware/cygwin/xfree/ There are numerous other X servers available in the same package or at the same location, but unfortunately most of them behave even worse than XWin. If you have any luck with any of these, *PLEASE* email the maintainers at xemacs-nt@xemacs.org, and we'll add the info here. 2. Source for the MIT X11R6.3 libraries, available from ftp.x.org. 3. You'll need to compile the MIT libraries without multi-thread support. To do this, there is an example Win32.cf and site.def provided which set the relevant flags. You will also need to apply the patch in nt/X11.patch in the xc/lib/X11 directory which will fix the DLL definition file. Once compiled and installed, you will need to apply the patch in nt/Xmd.patch. This is messy and better solutions would be appreciated. 4. Goto step 2 under 'Optional libraries' below. * Optional libraries ==================== 1. You really want the XPM library. Grab the latest version of the xpm sources (xpm-3.4k.tar.gz at time of writing) from ftp://ftp.xemacs.org/pub/xemacs/aux/ and unpack them somewhere. Copy nt\xpm.mak from the xemacs sources to the lib subdirectory of the xpm sources, cd to that directory and build xpm with 'nmake -f xpm.mak'. 2. You probably also want PNG image support. Grab the latest versions of zlib and libpng (zlib-1.1.3 and libpng-1.0.2 at time of writing) from ftp://ftp.xemacs.org/pub/xemacs/aux/, unpack them somewhere and read the respective READMEs for details on how to build them. The following build procedure works for zlib-1.1.3 and libpng-1.0.2: cd to the zlib directory, type 'copy msdos\makefile.w32 Makefile' and then type 'nmake'. cd to the libpng directory, rename or move the zlib directory to ..\zlib and type 'nmake -f scripts\makefile.w32'. 3. If you want TIFF support, grap the latest version of libtiff (tiff-v3.4 at time of writing) from ftp://ftp.xemacs.org/pub/xemacs/aux/ and unpack it somewhere. Copy nt\tiff.mak from the xemacs sources to the contrib\winnt subdirectory of the tiff sources, cd to that directory and build libtiff with 'nmake -f tiff.mak'. Note: tiff.mak has only been verified to work under WinNT, not Win95 or 98. However, the latest distribution of libtiff includes a contrib\win95\makefile.w95; that might work. 4. If you want JPEG support grab the latest version of jpegsrc (jpeg-6b at time of writing) from ftp://ftp.xemacs.org/pub/xemacs/aux/ and read the README for details on how to build it. 5. If you want X-Face support, grab the compface distribution from ftp://ftp.xemacs.org/pub/xemacs/aux/ and unpack it somewhere. Copy nt\compface.mak from xemacs sources to the compface directory. cd to that directory and build libcompface with 'nmake -f compface.mak'. * Building ========== 1. cd to the nt subdirectory of the xemacs distribution and copy the file config.inc.samp to config.inc. Make any necessary modifications. This file controls the options that XEmacs is built with: -- If you want international (aka "Mule") support, modify the appropriate line in config.inc as follows: HAVE_MULE=1 NOTE: This support is still quite raw under the Win32 native GUI, but works well if you compile the Cygwin version with X support and disable native Win32 support (--with-msw=no). -- If you're building with XPM support, modify the appropriate lines in config.inc as follows: HAVE_XPM=1 XPM_DIR="x:\location\of\your\xpm\sources" and similarly for JPEG and TIFF support. -- If you're building with PNG support, modify the appropriate lines in config.inc as follows: HAVE_PNG=1 PNG_DIR="x:\location\of\your\png\sources" ZLIB_DIR="x:\location\of\your\zlib\sources" -- If you're building with GIF support, modify the appropriate lines in config.inc as follows: HAVE_GIF=1 -- If you're building with X-Face support, modify the appropriate lines in config.inc as follows: HAVE_XFACE=1 COMPFACE_DIR="x:\location\of\your\compface\sources" -- If you're building for X, modify the appropriate lines in config.inc as follows: HAVE_X_WINDOWS=1 X11_DIR=x:\root\directory\of\your\X11\installation -- By default, XEmacs will expect to find its packages in the subdirectories "site-packages", "mule-packages" and "xemacs-packages" under the directory "c:\Program Files\XEmacs". If you want it to look for these subdirectories elsewhere, modify the appropriate lines in config.inc as follows: PACKAGE_PREFIX="x:\your\package\directory" Make sure that the directory pointed to by PACKAGE_PREFIX contains the xemacs-packages directory into which you installed the packages. -- XEmacs can build its info files more quickly if you have a copy of the makeinfo program. If you have a copy, modify the appropriate lines in config.inc as follows: MAKEINFO="x:\location\of\makeinfo.exe" If you don't have a copy of makeinfo then you'll need to have installed the XEmacs texinfo package. 2. If you want to install XEmacs when you build it, modify the appropriate lines in config.inc as follows (you can also run XEmacs from its build directory): INSTALL_DIR="x:\your\installation\directory" (By default, XEmacs will be installed in directories under the directory "c:\Program Files\XEmacs\XEmacs-21.2".) 3. If you want to build xemacs on the command line, use `nmake install -f xemacs.mak', or just `nmake -f xemacs.mak' if you want to run XEmacs from its build directory. nmake will build temacs, the DOC file, update the elc's, dump xemacs and (optionally) install the relevant files in the directories under the installation directory. If you chose to install XEmacs, the file that you should run to start XEmacs will be installed (by default) as "c:\Program Files\XEmacs\XEmacs-21.2\i586-pc-win32\xemacs.exe". To run from the build directory, run the file "nt\xemacs.exe" off of the root of the build directory. You may want to create a shortcut to the file from your Desktop or Start Menu. 4. To build using MS Developer Studio, you can use the workspace file `nt/xemacs.dsw'. This was prepared for Visual C++ 6.0. If you are using Visual C++ 5.0, you can use the workspace file `nt/xemacs-vc50.dsw'. If you have a different version and neither file works, just open up `nt/xemacs.mak' from within MS Developer Studio and it will offer to wrap this Makefile in a workspace file, from which you can build. Assuming you want to run from the build directory (which you will want to do if you are planning on doing any development work on XEmacs), use the following settings in Project/Settings...: Under the General tab: Build command line: NMAKE /f xemacs.mak Output file name: ..\src\xemacs.exe Browse info file name: ..\src\temacs.bsc Under the Debug tab: Executable for debug session: ..\src\xemacs.exe If you want to install XEmacs when it's built, change the build command line to "NMAKE install /f xemacs.mak". (You will have to make the same change even if you use the provided workspace nt/xemacs.dsw.) * Debugging under MS Developer Studio ===================================== The build process always creates debugging and "Source Browser" information in the source tree for use with DevStudio. However that information is not very useful unless you build a debug version of XEmacs: 1. Set DEBUG_XEMACS=1 and DEPEND=1 in config.inc and rebuild. 2. See instructions above for obtaining a workspace file for use with MS Developer Studio. Build and debug your XEmacs this way. 3. To display the contents of a lisp variable, type Shift-F9 (or use the menu) to bring up the QuickWatch window, type debug_print(variable) and click Recalculate. The output will appear in a console window, as well as in the Debug window in MS Developer Studio. 4. To view Lisp variables in the "Watch" window wrap the variable in one of the helper functions from the file src\console-msw.c, for example type DSTRING(variable) in the "Watch" window to inspect a Lisp string. * Known Problems ================ Please look at the PROBLEMS file for known problems. Any other problems you need clarified, please email us and we will endeavour to provide any assistance we can: The XEmacs NT Mailing List: xemacs-nt@xemacs.org Subscribe address: xemacs-nt-request@xemacs.org Ben Wing (current primary MS Windows maintainer; author of the MS Windows Mule code and some of the dialog box code) Andy Piper (MS Windows contributor; author of the Cygwin support and the MS Windows glyph and widget code) Jonathan Harris (MS Windows contributor; author of the MS Windows redisplay and underlying GUI code) Kirill Katsnelson (MS Windows contributor; author of the MS Windows process and printing code and some of the dialog box code; general guru on obscure MS Windows programming topics) David Hobley (early MS Windows contributor) Marc Paquette (early MS Windows contributor) August Hill (early MS Windows contributor) and others.