istext

#!/usr/bin/env bash
#
#  Synopsis:
#	List UTF-8 text compliant files path in either argv or standard input.
#  Usage:
#	istext [-v] /etc/*
#	find / -xdev -print0 | xargs -0 istext | xargs fgrep -l PATTERN
#  Note:
#	Need option "--reverse-match".
#
#	Investigate why bash version of istext is 100 times slower then
#	old c version.  probably because only a fraction of the file
#	is searched in old version.
#
#	Consider an option "--frisk" that uses program jmscott/slice-file
#	to extract just a portion of the file instead of the whole file
#	(as legacy istext.c does).
#

REVERSE=no

_istext()
{
	local F;
	F="$1"

	if [ ! -f "$F" ];  then
		if [ ! -e "$F" ];  then
			echo "file does not exist: $F" >&2
		fi
		return
	fi
	is-utf8wf <"$F"
	STATUS=$?
	case $STATUS in
	0)
		test $REVERSE = no && echo "$F"
		;;
	1|2)
		test $REVERSE = yes && echo "$F"
		;;
	*)
		echo "$PROG: ERROR: is-utf8wf failed: exit status=$STATUS" >&2
		exit 2
		;;
	esac
	return 0
}

if [ "$1" = -v ];  then
	REVERSE=yes
	shift
fi

if [ $# = 0 ];  then
	while read -r F;  do
		_istext "$F"
	done
else
	for F in $@;  do
		_istext "$F"
	done
fi