Mercurial > emacs
view configure1.in @ 1167:a9aeeaa9da8f
(scan_lists): When searching back for comment:
if comment-end is 2 chars, assume it does end a comment.
Otherwise, scan back to previous comment-end to see if there's
a comment starter between. Also record whether the string quotes
between the start and the end are paired and uniform.
If so, skip to comment starter. If not, scan from start of
defun to find comment starter if any.
(find_defun_start): New function.
author | Richard M. Stallman <rms@gnu.org> |
---|---|
date | Sat, 19 Sep 1992 17:52:07 +0000 |
parents | 4875b01ff377 |
children | 93ee9c5e7eec |
line wrap: on
line source
#!/bin/sh # Configuration script for GNU Emacs # Copyright (C) 1992 Free Software Foundation, Inc. #This file is part of GNU Emacs. #GNU Emacs 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 1, or (at your option) #any later version. #GNU Emacs 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 GNU Emacs; see the file COPYING. If not, write to #the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. # Shell script to edit files and make symlinks in preparation for # compiling Emacs. # # Usage: configure machine # # If configure succeeds, it leaves its status in config.status. # If configure fails after disturbing the status quo, # config.status is removed. # # Remove any leading "." elements from the path name. If we don't # remove them, then another "./" will be prepended to the file name # each time we use config.status, and the program name will get larger # and larger. This wouldn't be a problem, except that since progname # gets recorded in all the Makefiles this script produces, # move-if-changed thinks they're different when they're not. # # It would be nice if we could put the ./ in a \( \) group and then # apply the * operator to that, so we remove as many leading ./././'s # as are present, but some seds (like Ultrix's sed) don't allow you to # apply * to a \( \) group. Bleah. progname="`echo $0 | sed 's:^\./::'`" short_usage="Type \`${progname} -usage' for more information about options." usage_message="Usage: ${progname} MACHINENAME [-OPTION[=VALUE] ...] This message needs to be updated. Set compilation and installation parameters for GNU Emacs, and report. MACHINENAME is the machine to build for. See \`etc/MACHINES'. Options are: --opsystem=SYSTEM - operating system to build for; see \`etc/MACHINES'. --prefix=DIR - where to install Emacs's library files These options have reasonable defaults (in []s), and may not be needed: -g, -O - Passed to the compiler. Default is -g, plus -O if using gcc. --cc=COMPILER - Which compiler to use. Defaults to gcc if available. --libdir=DIR - where to look for arch-dependent library files --datadir=DIR - where to look for architecture-independent library files --bindir=DIR - where to install the Emacs executable, and some friends --lisppath=PATH - colon-separated list of Emacs Lisp directories --lockdir=DIR - where Emacs should do its file-locking stuff --with-x or --with-x10 - what window system to use; default is to use X11 if present If successful, ${progname} leaves its status in config.status. If unsuccessful after disturbing the status quo, config.status is removed." # These are omitted since users should not mess with them. # --gnu-malloc=[yes] or no - use the GNU memory allocator # --rel-alloc=[yes] or no - use compacting allocator for buffers # --lisp-float-type=[yes] or no - Support floating point in Emacs Lisp. # --window-system is omitted because --with... follow the conventions. if [ ! -r ./src/lisp.h ]; then echo "${progname}: Can't find Emacs sources in \`./src'. Run this config script in the top directory of the Emacs source tree." >&2 exit 1 fi # The option names defined here are actually the shell variable names. # They should have `_' in place of `-'. options=":\ usage:help:\ machine:opsystem:\ g:O:cc:\ prefix:bindir:emacsdir:datadir:lispdir:locallisppath:\ lisppath:buildlisppath:statedir:lockdir:libdir:mandir:infodir:\ gnu_malloc:rel_alloc:lisp_float_type:\ window_system:\ " boolean_opts=":\ g:O:\ gnu_malloc:rel_alloc:lisp_float_type:have_x_menu:with_x:with_x11:with_x10:\ " config_h_opts=":\ gnu_malloc:rel_alloc:lisp_float_type:\ have_x_windows:have_x11:\ c_switch_site:sigtype:\ " prefix= bindir=/usr/local/bin gnu_malloc=yes lisp_float_type=yes # The default values for the following options are guessed at after other # options have been checked and given values, so we set them to null here. lisppath="" datadir="" libdir="" lockdir="" window_system="" # Record all the arguments, so we can save them in config.status. arguments="$*" echo "Examining options." while [ $# != 0 ]; do arg="$1" case "${arg}" in -*) # Separate the switch name from the value it's being given. # Also change `-' in the option name to `_'. case "${arg}" in -*=*) opt=`echo ${arg} | sed 's:^-*\([^=]*\)=.*$:\1:' | tr - _` val=`echo ${arg} | sed 's:^-*[^=]*=\(.*\)$:\1:'` valomitted=no ;; -*) # If FOO is a boolean argument, -FOO is equivalent to # -FOO=yes. Otherwise, the value comes from the next # argument - see below. opt=`echo ${arg} | sed 's:^-*\(.*\)$:\1:' | tr - _` val="yes" valomitted=yes ;; esac # Make sure the argument is valid and unambiguous. case ${options} in *:${opt}:* ) # Exact match. optvar=${opt} ;; *:${opt}*:${opt}*:* ) # Ambiguous prefix. echo "\`-${opt}' is an ambiguous switch; it could be any of the following:" # We can't just use tr to translate colons to newlines, since # BSD sed and SYSV sed use different syntaxes for that. spaced_options=`echo ${options} | tr ':' ' '` echo `(for option in ${spaced_options}; do echo $option; done) \ | grep "^${opt}"` echo ${short_usage} exit 1 ;; *:${opt}*:* ) # Unambigous prefix. optvar=`echo ${options} | sed 's/^.*:\('${opt}'[^:]*\):.*$/\1/'` ;; * ) (echo "\`-${opt}' is not a valid option." echo "${short_usage}") | more exit 1 ;; esac case "${optvar}" in usage | help) echo "${usage_message}" | more exit 1 ;; esac # If the variable is supposed to be boolean, make sure the value # given is either "yes" or "no". If not, make sure some value # was given. case "${boolean_opts}" in *:${optvar}:* ) case "${val}" in y | ye | yes ) val=yes ;; n | no ) val=no ;; * ) echo "The \`-${optvar}' option (\`-${opt}') is supposed to have a boolean value - set it to either \`yes' or \`no'." >&2 exit 1 ;; esac ;; *) if [ "${valomitted}" = "yes" ]; then if [ $# = 1 ]; then (echo "${progname}: You must give a value for the \`-${opt}' option, as in \`-${opt}=FOO'." echo "${short_usage}") | more exit 1 fi shift; val="$1" fi ;; esac eval "${optvar}=\"${val}\"" ;; *) machine=${arg} ;; esac shift done if [ "${machine}" = "" ]; then (echo "You must specify a machine name as an argument to ${progname}." echo "${short_usage}") | more exit 1 fi echo "Checking the machine." machfile="m/${machine}.h" if [ ! -r src/${machfile} ]; then echo "${progname}: Emacs has no configuration info for the machine called \`${machine}'. Look at etc/MACHINES for the names of machines that Emacs has been ported to." >&2 exit 1 fi echo "Checking the operating system." if [ "${opsystem}" = "" ]; then echo " No operating system explicitly specified." echo " Guessing, based on machine." # Get the default operating system to go with the specified machine. opsystem=`grep 'USUAL-OPSYS="' src/${machfile} \ | sed 's/^.*USUAL-OPSYS="\([^"]*\)".*$/\1/'` if [ "${opsystem}" = "" ]; then echo "${progname}: Emacs's configuration files don't suggest what operating system a \`${machine}' machine might run. Try specifying the operating system explicitly by passing ${progname} an \`-opsystem=SYSTEM-NAME' flag. Look at etc/MACHINES for the names of operating systems that Emacs has been ported to." >&2 exit 1 fi if [ "${opsystem}" = "note" ]; then echo "The \`${machine}' machine can be used with more than one operating system, and Emacs's configuration info isn't clever enough to figure out which one you're running. Run ${progname} with -machine and -opsystem arguments as specified below for the appropriate system. (This information comes from the file \`etc/MACHINES' - see that file for more detail.) " >&2 sed < src/${machfile} -e '1,/NOTE-START/d' -e '/NOTE-END/,$d' | more echo exit 1 fi opsysfile="s/${opsystem}.h" if [ ! -r src/${opsysfile} ]; then echo "${progname}: Emacs's configuration files say that the default operating system for the machine \`${machine}' is \`${opsystem}', but there is no configuration file for \`${opsystem}', so Emacs's default info is screwed up. Try specifying the operating system explicitly by passing ${progname} an \`-opsystem=SYSTEM-NAME' flag." >&2 exit 1 fi else opsysfile="s/${opsystem}.h" if [ ! -r src/${opsysfile} ]; then echo "${progname}: Emacs has no configuration info for the operating system \`${opsystem}'. Look at etc/MACHINES for the names of operating systems that Emacs has been ported to." >&2 exit 1 fi fi if [ ! "${prefix}" ]; then prefix="/usr/local" fi if [ ! "${emacsdir}" ]; then emacsdir="${prefix}/emacs-19.0" fi if [ ! "${datadir}" ]; then datadir="${emacsdir}/etc" fi if [ ! "${lispdir}" ]; then lispdir="${emacsdir}/lisp" fi if [ ! "${locallisppath}" ]; then locallisppath="${emacsdir}/local-lisp" fi if [ ! "${lisppath}" ]; then lisppath="${locallisppath}:${lispdir}" fi if [ ! "${buildlisppath}" ]; then buildlisppath=../lisp fi if [ ! "${statedir}" ]; then statedir="${emacsdir}" fi if [ ! "${lockdir}" ]; then lockdir="${statedir}/lock" fi if [ "${libdir}" = "" ]; then libdir="${emacsdir}/arch-lib" fi if [ ! "${mandir}" ]; then mandir="/usr/man/man1" fi if [ ! "${infodir}" ]; then infodir="${prefix}/info" fi echo "Checking window system." case "${with_x11}" in yes ) window_system=${window_system}x11 ;; esac case "${with_x}" in yes ) window_system=${window_system}x11 ;; esac case "${with_x10}" in yes ) window_system=${window_system}x10 ;; esac # Note that SYSV `tr' doesn't handle character ranges. window_system="`echo ${window_system} \ | tr ABCDEFGHIJKLMNOPQRSTUVWXYZ abcdefghijklmnopqrstuvwxyz`" case "${window_system}" in "none" | "x11" | "x10" ) ;; "x" ) window_system=x11 ;; "" ) echo " No window system specifed. Looking for X Windows." window_system=none if [ -r /usr/lib/libX11.a -o -d /usr/include/X11 ]; then window_system=x11 fi ;; * ) echo "Don\'t specify the window system more than once." >&2 exit 1 ;; esac case "${window_system}" in x11 ) have_x_windows=yes have_x11=yes ;; x10 ) have_x_windows=yes have_x11=no ;; none ) have_x_windows=no have_x11=no ;; esac echo "Checking for GCC." case "${cc}" in "" ) temppath=`echo $PATH | sed 's/^:/.:/ s/::/:.:/g s/:$/:./ s/:/ /g'` cc=`( for dir in ${temppath}; do if [ -f ${dir}/gcc ]; then echo gcc; exit 0; fi done echo cc )` ;; esac case "${cc}" in "gcc" ) # With GCC, both O and g should default to yes, no matter what # the other is. case "${O},${g}" in , ) O=yes; g=yes ;; ,* ) O=yes; ;; *, ) g=yes ;; esac ;; "*" ) # With other compilers, treat them as mutually exclusive, # defaulting to debug. case "${O},${g}" in , ) O=no ; g=yes ;; ,no ) O=yes; ;; ,yes ) O=no ; ;; no, ) g=yes ;; yes, ) g=no ;; esac ;; esac # What is the return type of a signal handler? We run # /usr/include/signal.h through cpp and grep for the declaration of # the signal function. Yuck. echo "Looking for return type of signal handler functions." signal_h_file='' if [ -r /usr/include/signal.h ]; then signal_h_file=/usr/include/signal.h elif [ -r /usr/include/sys/signal.h ]; then signal_h_file=/usr/include/sys/signal.h fi sigtype=void if [ "${signal_h_file}" ]; then sigpattern='[ ]*([ ]*\*[ ]*signal[ ]*(' # We make a copy whose name ends in .c, so the compiler # won't complain about having only been given a .h file. tempcname="configure.tmp.$$.c" cp ${signal_h_file} ${tempcname} if ${cc} -E ${tempcname} | grep "int${sigpattern}" > /dev/null; then sigtype=int fi rm -f ${tempcname} fi echo "Examining the machine- and system-dependent files to find out" echo " - which libraries the lib-src programs will want, and" echo " - whether the GNU malloc routines are usable." tempcname="configure.tmp.$$.c" echo '#include "src/'${opsysfile}'" #include "src/'${machfile}'" #ifndef LIBS_MACHINE #define LIBS_MACHINE #endif #ifndef LIBS_SYSTEM #define LIBS_SYSTEM #endif @configure@ libsrc_libs=LIBS_MACHINE LIBS_SYSTEM #ifdef SYSTEM_MALLOC @configure@ system_malloc=yes #else @configure@ system_malloc=no #endif ' > ${tempcname} eval `${cc} -E ${tempcname} \ | grep '@configure@' \ | sed -e 's/^@configure@//'` rm ${tempcname} # Do the opsystem or machine files prohibit the use of the GNU malloc? if [ "${system_malloc}" = "yes" ]; then gnu_malloc=no gnu_malloc_reason=" (The GNU allocators don't work with this machine and/or operating system.)" fi if [ ! "${rel_alloc}" ]; then rel_alloc=${gnu_malloc} fi rm -f config.status set -e # Make the proper settings in the config file. echo "Making src/config.h from src/config.h.in" case "${g}" in "yes" ) c_switch_site="${c_switch_site} -g" ;; esac case "${O}" in "yes" ) c_switch_site="${c_switch_site} -O" ;; esac sed_flags="-e 's:@machine@:${machfile}:'" sed_flags="${sed_flags} -e 's:@opsystem@:${opsysfile}:'" for flag in `echo ${config_h_opts} | tr ':' ' '`; do # Note that SYSV `tr' doesn't handle character ranges. cflagname=`echo ${flag} \ | tr abcdefghijklmnopqrstuvwxyz ABCDEFGHIJKLMNOPQRSTUVWXYZ` val=`eval echo '$'${flag}` case ${val} in no | "") f="-e 's:.*#define ${cflagname}.*:/\\* #define ${cflagname} \\*/:'" ;; yes) f="-e 's:.*#define ${cflagname}.*:#define ${cflagname}:'" ;; *) f="-e 's:.*#define ${cflagname}.*:#define ${cflagname} ${val}:'" ;; esac sed_flags="${sed_flags} ${f}" done rm -f src/config.h.tmp eval '/bin/sed '${sed_flags}' < src/config.h.in > src/config.h.tmp' ./move-if-change src/config.h.tmp src/config.h # Remind people not to edit this. chmod -w src/config.h # Modify the parameters in the top makefile. echo "Producing ./Makefile from ./Makefile.in." rm -f Makefile.tmp (echo "# This file is generated by \`${progname}' from \`./Makefile.in'. # If you are thinking about editing it, you should seriously consider # editing \`./Makefile.in' itself, or running \`${progname}' instead." /bin/sed < Makefile.in \ -e '/^# DIST: /d' \ -e 's;^\(prefix=\).*$;\1'"${prefix};" \ -e 's;^\(bindir=\).*$;\1'"${bindir};" \ -e 's;^\(emacsdir=\).*$;\1'"${emacsdir};" \ -e 's;^\(datadir=\).*$;\1'"${datadir};" \ -e 's;^\(lispdir=\).*$;\1'"${lispdir};" \ -e 's;^\(locallisppath=\).*$;\1'"${locallisppath};" \ -e 's;^\(lisppath=\).*$;\1'"${lisppath};" \ -e 's;^\(buildlisppath=\).*$;\1'"${buildlisppath};" \ -e 's;^\(statedir=\).*$;\1'"${statedir};" \ -e 's;^\(lockdir=\).*$;\1'"${lockdir};" \ -e 's;^\(libdir=\).*$;\1'"${libdir};" \ -e 's;^\(mandir=\).*$;\1'"${mandir};" \ -e 's;^\(infodir=\).*$;\1'"${infodir};" \ ) > ./Makefile.tmp ./move-if-change Makefile.tmp Makefile # Remind people not to edit this. chmod -w ./Makefile # Modify the parameters in the `build-install' script. echo "Producing ./build-install from ./build-install.in." rm -f ./build-install.tmp (echo "# This file is generated by \`${progname}' from \`./build-install.in'. # If you are thinking about editing it, you should seriously consider # editing \`./build-install.in' itself, or running \`${progname}' instead." /bin/sed < build-install.in \ -e 's;^\(prefix=\).*$;\1'"${prefix};" \ -e 's;^\(bindir=\).*$;\1'"${bindir};" \ -e 's;^\(lisppath=\).*$;\1'"${lisppath};" \ -e 's;^\(datadir=\).*$;\1'"${datadir};" \ -e 's;^\(lockdir=\).*$;\1'"${lockdir};" \ -e 's;^\(libdir=\).*$;\1'"${libdir};") > ./build-install.tmp ./move-if-change build-install.tmp build-install # Remind people not to edit this. chmod -w build-install chmod +x build-install # Modify the parameters in the src makefile. echo "Producing src/Makefile from src/Makefile.in." rm -f src/Makefile.tmp (echo "# This file is generated by \`${progname}' from \`Makefile.in'. # If you are thinking about editing it, you should seriously consider # editing \`Makefile.in' itself, or running \`${progname}' instead." /bin/sed < src/Makefile.in \ -e '/^# DIST: /d' \ -e 's;^\(CC[ ]*=\).*$;\1'"${cc};") > src/Makefile.tmp ./move-if-change src/Makefile.tmp src/Makefile # Remind people not to edit this. chmod -w src/Makefile # Modify the parameters in the lib-src makefile. echo "Producing lib-src/Makefile from lib-src/Makefile.in." rm -f lib-src/Makefile.tmp (echo "# This file is generated by \`${progname}' from \`Makefile.in'. # If you are thinking about editing it, you should seriously consider # editing \`Makefile.in' itself, or running \`${progname}' instead." /bin/sed < lib-src/Makefile.in \ -e '/^# DIST: /d' \ -e 's;^\(CFLAGS=\).*$;\1'"${c_switch_site};" \ -e 's;^\(LOADLIBES=\).*$;\1'"${libsrc_libs};" \ -e 's;^\(CC=\).*$;\1'"${cc};") > lib-src/Makefile.tmp ./move-if-change lib-src/Makefile.tmp lib-src/Makefile # Remind people not to edit this. chmod -w lib-src/Makefile # Create a verbal description of what we have done. message="Configured for machine \`${machine}' running \`${opsystem}'. The following values have been set in ./Makefile and ./build-install: \`make install' or \`build-install' will placed executables in ${bindir}. Emacs's lisp search path will be \`${lisppath}'. Emacs will look for its architecture-independent data in ${datadir}. Emacs will look for its utility programs and other architecture- dependent data in ${libdir}. Emacs will keep track of file-locking in ${lockdir}. The following values have been set in src/config.h: Should Emacs use the GNU version of malloc? ${gnu_malloc}${gnu_malloc_reason} Should Emacs use the relocating allocator for buffers? ${rel_alloc} Should Emacs support a floating point data type? ${lisp_float_type} What window system should Emacs use? ${window_system} What compiler should emacs be built with? ${cc} Should the compilation use \`-g' and/or \`-O'? ${c_switch_site- neither}" # Document the damage we have done by writing config.status. echo '#!/bin/sh' > config.status echo "# This file is generated by \`${progname}.' # If you are thinking about editing it, you should seriously consider # running \`${progname}' instead. " >> config.status echo "${message}" | sed -e 's/^/# /' >> config.status echo "'${progname}' ${arguments} "'$@' >> config.status # Remind people not to edit this. chmod -w config.status chmod +x config.status # Print the description. echo echo "${message}" exit 0