#!/bin/bash

set -e

SUPPORTED=/var/lib/locales/supported.d
LOCALES=/usr/share/i18n/locales
STATEDIR=/var/lib/belocs

[ -n "$POSIXLY_CORRECT" ] && unset POSIXLY_CORRECT || true

IS_LANG=no
ARCHIVE=yes
PURGE=no
ALIASES=

#  Handle command-line options
prev=
while true
do
	option="$1"
	if [ -n "$prev" ]; then
                eval "$prev=\$option"
                prev=
                continue
	fi

	case $option in
	-h|--help)
		cat <<EOT
Usage: locale-gen [OPTIONS]
Options:
 -h, --help         display this message and exit
     --purge        remove existing locales before processing
     --archive      store compiled locale data inside a single archive
     --no-archive   do not store compiled locale data inside a single archive
                    (default)
     --aliases=FILE read locale aliases from FILE. (Default: /etc/locale.alias)
     --lang         treat argument as generic language code
EOT
		exit 0
		;;
	--purge)
		PURGE=yes
		;;
	--no-purge)
		PURGE=no
		;;
	--archive)
		ARCHIVE=yes
		;;
	--no-archive)
		ARCHIVE=no
		;;
	--keep-existing)
		#  Cache in belocs-locales-bin makes this flag useless,
		#  but it is provided so that belocs-locales-bin can be
		#  installed on Ubuntu boxes.
		;;
	--aliases)
		prev=ALIASES
		;;
	--aliases=*)
		ALIASES=$(expr "x$option" : 'x[^=]*=\(.*\)')
		;;
	--lang)
		IS_LANG=yes
		;;
	--*)
		echo "locale-gen: invalid option -- $option"
		echo "Try 'locale-gen --help' for more information."
		exit 1
		;;
	*)
		break
		;;
	esac

	shift
done

mkdir -p "$SUPPORTED"

# determine the locales to be generated in $GENERATE
GENERATE=
if [ -z "$1" ]; then
    [ -n "`ls $SUPPORTED`" ] || exit 0
    GENERATE=`cat $SUPPORTED/*`
fi

while [ -n "$1" ]; do
    if [ -f "$SUPPORTED/$1" ]; then
        GENERATE="$GENERATE\n`cat $SUPPORTED/$1`"
    elif [ $IS_LANG = no ] && L=`grep "^${1/%.utf8/.UTF-8} " /usr/share/i18n/SUPPORTED`; then
        GENERATE="$GENERATE\n$L"
        # add to supported.d/local if necessary
        if [ ! -f "$SUPPORTED/local" ] || ! grep -q "^$L" "$SUPPORTED/local"; then
            echo "$L" >> "$SUPPORTED/local"
        fi
    else
        # try to come up with a sensible default
        GENERATE=`grep -E "^${1}( |[._@][^[:space:]]* )UTF-8" /usr/share/i18n/SUPPORTED`
        if [ -z "$GENERATE" ]; then
            echo "Error: '$1' is not a supported language or locale" >&2
            exit 1
        fi
        NL='
'
        IFS_SAVE="$IFS"
        IFS="$NL"
        for L in $GENERATE; do
            IFS="$IFS_SAVE"
            # add to supported.d/local if necessary
            if [ ! -f "$SUPPORTED/local" ] || ! grep -q "^$L" "$SUPPORTED/local"; then
                echo "$L" >> "$SUPPORTED/local"
            fi
            IFS="$NL"
        done
        IFS="$IFS_SAVE"
    fi
    shift
done

[ -d "$STATEDIR" ] || mkdir -p "$STATEDIR" || PURGE=yes

no_archive=
[ "$ARCHIVE" = yes ] || no_archive="--no-archive"
locale_alias=
[ -n "$ALIASES" ] && [ -r "$ALIASES" ] && locale_alias="-A $ALIASES"

umask 022

is_entry_ok() {
	if [ -n "$locale" ] && [ -n "$charset" ]
	then
		true
	else
		echo "Error: Bad entry '$locale $charset'" 1>&2
		false
	fi
}

normalize_locale() {
	#  Insert a leading x in case $1 begins with a dash
	this_locale=x$1
	charset=
	if echo $this_locale | LC_ALL=C grep '\.' > /dev/null 2>&1; then
		charset=$(echo $this_locale | sed -e 's/^x//' -e 's/.*\(\.[^@]*\).*/\1/' | LC_ALL=C tr '[A-Z]' '[a-z]' | LC_ALL=C sed -e 's/[^a-z0-9.]//g')
	fi
	modifier=
	if echo $this_locale | LC_ALL=C grep '@' > /dev/null 2>&1; then
		modifier=$(echo $this_locale | sed -e 's/^x//' -e 's/.*\(@[^.]*\).*/\1/')
	fi
	main=$(echo $this_locale | sed -e 's/^x//' -e 's/[.@].*//')
	echo $main$charset$modifier
}

gen_dep() {
	[ -d "$STATEDIR" ] || return 1
	#  Generate the dependency list.  It is needed by update_md5sum()
	( LC_ALL=C grep '^\(include\|copy\)' "$LOCALES"/* |\
	  sed -e "s,^$LOCALES/,," -e 's/\(include\|copy\)[ 	]*"/ /' -e 's/".*//'
	) | LC_ALL=C sort -u > "$STATEDIR/locales.dep" 2>/dev/null
}

gen_md5sum() {
	[ -d "$STATEDIR" ] || return 1
	#  If old md5sums are not present, regenerate all locales.
	[ -f "$STATEDIR/hashfile" ] || return 1
	mv -f "$STATEDIR/hashfile" "$STATEDIR/hashfile.old" || return 1

	md5sum "$LOCALES"/* > "$STATEDIR/hashfile.new" || return 1
	locale -a > "$STATEDIR/list" 2>/dev/null || return 1
}

#  Check whether a locale was already generated and all its dependencies
#  have not been modified.
already_generated() {
	[ -f "$STATEDIR/locales.dep" ] || return 1
	[ -f "$STATEDIR/hashfile.old" ] || return 1
	[ -f "$STATEDIR/hashfile.new" ] || return 1
	[ -f "$STATEDIR/list" ] || return 1
	this_locale=$(normalize_locale $1)
	this_input=$2
	#  Check whether the locale is already present.
	LC_ALL=C grep "^$this_locale\$" "$STATEDIR/list" > /dev/null 2>&1 || return 1
	#  If yes, check if this file and its dependencies have changed
	for file in "$this_input" $(LC_ALL=C grep "^$this_input:" "$STATEDIR/locales.dep" | awk '{printf "%s ", $2}')
	do
		md5old=$(LC_ALL=C grep "/$file\$" "$STATEDIR/hashfile.old" | sed -e 1q)
		md5new=$(LC_ALL=C grep "/$file\$" "$STATEDIR/hashfile.new" | sed -e 1q)
		[ -n "$md5old" ] || return 1
		[ -n "$md5new" ] || return 1
		[ "$md5new" = "$md5old" ] || return 1
	done
	return 0
}

update_md5sum() {
	[ -f "$STATEDIR/locales.dep" ] || return 0
	this_locale=$1
	for file in "$this_locale" $(LC_ALL=C grep "^$this_locale:" "$STATEDIR/locales.dep" | awk '{printf "%s ", $2}')
	do
		md5sum "$LOCALES/$file" >> "$STATEDIR/hashfile" || return 0
	done
}

gen_dep 2>/dev/null || true
[ "$PURGE" = yes ] || gen_md5sum 2>/dev/null || PURGE=yes
if [ "$PURGE" = yes ]; then
	rm -f "$STATEDIR/hashfile.old" "$STATEDIR/hashfile.new" "$STATEDIR/list" 2>/dev/null || true
	for dir in /usr/lib/locale/*; do
		[ -e "$dir" ] || continue
		if [ "${dir#/usr/lib/locale/}" = C.UTF-8 ]; then
			# owned by libc-bin
			continue
		fi
		rm -rf "$dir" 2>/dev/null || true
	done
fi

echo "Generating locales..."
/bin/echo -e "$GENERATE" | sort -u | while read locale charset
do
	case $locale in
		\#*) continue;;
		"") continue;;
	esac
	is_entry_ok || continue
	#  Remove charset if present
	input=$(echo $locale | sed -e 's/\.[^@]*//')
	#  Insert charset before modifier
	printf "%s" $input | sed -e "s/^\\([^@]*\\)\\(.*\\)/  \\1.${charset}\\2... /"
	if already_generated $locale $input; then
		echo 'up-to-date'
	else
		if localedef $no_archive -i $input -c -f $charset $locale_alias $locale; then	
		    echo 'done'
		else
		    echo 'failed'
		fi
	fi
	update_md5sum $input 2>/dev/null
done
echo "Generation complete."