#!/bin/bash # This script is used for translations using .po files. # It merges .xml files per chapter (or appendix). # The reason files are merged is that individual .xml files # are often not well-formed xml. if [ "$1" = "--help" ]; then echo "Usage: $0 lang" exit 0 fi language=${1:-pl} SCRIPTDIR="./scripts" BUILDDIR="./build" if [ -z "$PO_USEBUILD" ] ; then WORKDIR="./integrated" else WORKDIR="$BUILDDIR/build.po" fi LANGDIR="./$language" [ -d $LANGDIR ] || exit 1 TEMPDIR=/tmp/merge_xml.$$ LOG=$TEMPDIR/merge_xml.$language.log [ -d $TEMPDIR ] || mkdir $TEMPDIR # OLD_IFS=$IFS IFS=" " :>$TEMPDIR/entlist echo "Building list of entities..." for ENT in `grep ">$TEMPDIR/entlist echo -n ":" >>$TEMPDIR/entlist echo "$(echo $ENT | sed "s/.*SYSTEM\ *\"##SRCPATH##\///" | sed "s/\">//")" >>$TEMPDIR/entlist done IFS=$OLD_IFS # Make sure that all files are in UTF-8 first echo "Converting XML files to UTF-8..." echo "Converting XML files to UTF-8..." >>$LOG for FILE in `find $LANGDIR -name "*.xml"` ; do SUBDIR=$(dirname $FILE | sed "s:$LANGDIR::"); XML=$(basename $FILE) mkdir -p $TEMPDIR/in/$SUBDIR REGEXP="^>$LOG ENC=$(egrep $REGEXP $FILE | sed "s/.*xml.*encoding=\"//" | sed "s/\"?>//") iconv -f $ENC -t utf-8 $FILE | egrep -v $REGEXP >$TEMPDIR/in/$SUBDIR/$XML else echo "Not encoded: $FILE" >>$LOG cp $FILE $TEMPDIR/in/$SUBDIR/$XML fi done echo "" >>$LOG # Include lower level xml-files for all the main level xml-files echo "Merging XML files per 'chapter'..." echo "Merging XML files per 'chapter'..." >>$LOG gawk -v WORKDIR="$TEMPDIR" -v LOG=$LOG -v ENTLIST="$TEMPDIR/entlist" \ -f $SCRIPTDIR/merge_xml.awk $BUILDDIR/templates/install.xml.template # Copy the results to their proper location TARGET="$WORKDIR/$language" if [ -d $TARGET ]; then rm -r $TARGET fi mkdir -p $TARGET cp -r $TEMPDIR/out/* $TARGET cp $LOG $TARGET #rm -r $TEMPDIR exit 0