# .Sanitize for gdb/doc. # Each directory to survive its way into a release will need a file # like this one called "./.Sanitize". All keyword lines must exist, # and must exist in the order specified by this file. Each directory # in the tree will be processed, top down, in the following order. # Hash started lines like this one are comments and will be deleted # before anything else is done. Blank lines will also be squashed # out. # The lines between the "Do-first:" line and the "Things-to-keep:" # line are executed as a /bin/sh shell script before anything else is # done in this directory. Do-first: # Note that gdbgui.texinfo is actually a generic document, but right # now it only describes gdbtk, so we keep/lose as a gdbtk file. gdbtk_files="gdbgui.texinfo" if ( echo $* | grep lose\-gdbtk > /dev/null ) ; then lose_these_too="${gdbtk_files} ${lose_these_too}" if [ -n "${verbose}" ] ; then echo Deleting ${gdbtk_files} fi else keep_these_too="${gdbtk_files} ${keep_these_too}" if [ -n "${verbose}" ] ; then echo Keeping ${gdbtk_files} fi fi # All files listed between the "Things-to-keep:" line and the # "Files-to-sed:" line will be kept. All other files will be removed. # Directories listed in this section will have their own Sanitize # called. Directories not listed will be removed in their entirety # with rm -rf. # Note that we don't even keep the "config" directory, since it is # not currently used (since we abolished use of M4 in the docs). Things-to-keep: ChangeLog HPPA-cfg.texi LRS Makefile.in a4rc.sed agentexpr.texi all-cfg.texi annotate.texi configure configure.in libgdb.texinfo gdb.texinfo gdbint.texinfo h8-cfg.texi lpsrc.sed psrc.sed refcard.tex remote.texi stabs.texinfo Things-to-lose: # Cygnus logo bitmap, huge file, no point in bundling all over the place. cyglogo.eps # Includes cyglogo.eps. cygref.tex # The README file for gdb testers using snapshots. snapshots.readme Do-last: # Don't try to clean directories here, as the 'mv' command will fail. # Also, grep fails on NFS mounted directories. if ( echo $* | grep lose\-gdbtk > /dev/null ) ; then echo Catering to RMS by removing traces of \"gdbtk\"... for i in * ; do if test ! -d $i && (grep sanitize-gdbtk $i > /dev/null) ; then echo Removing traces of \"gdbtk\" out of $i... cp $i new sed '/start\-sanitize\-gdbtk/,/end-\sanitize\-gdbtk/d' < $i > new if [ -n "${safe}" -a ! -f .Recover/$i ] ; then echo Caching $i in .Recover... mv $i .Recover fi mv new $i fi done else echo Leaving \"gdbtk\" in the sources... for i in * ; do if test ! -d $i && (grep sanitize-gdbtk $i > /dev/null) ; then echo Keeping \"gdbtk\" stuff in $i, but editing out sanitize lines... cp $i new sed -e '/start\-sanitize\-gdbtk/d' -e '/end\-sanitize\-gdbtk/d' < $i > new if [ -n "${safe}" -a ! -f .Recover/$i ] ; then echo Caching $i in .Recover... mv $i .Recover fi mv new $i fi done fi # End of file.