t.split

一个C语言写的快速贝叶斯垃圾邮件过滤工具

#! /bin/sh

# Note:  When run via "make check", test output files are automatically deleted.
#	 When run from the command line, output files are left in directory
#	 split.YYYYMMDD.  This is useful when something is different.
#
#	 ./inputs  - test inputs
#	 ./outputs - known correct outputs
#
#	 split.YYYYMMDD:
#		directory containing individual output files

NODB=1 . ${srcdir=.}/t.frame

map_rc()
{
    (
	set +e
	eval "$@"
	a=$?
	[ $a -eq 0 ] && exit 0
	[ $a -eq 1 ] && exit 0
	[ $a -eq 2 ] && exit 0
	exit $a
    )
}

$BOGOFILTER -C -s < ${srcdir}/inputs/spam.mbx
$BOGOFILTER -C -n < ${srcdir}/inputs/good.mbx

OUT="split.out"

files=`ls ${srcdir}/inputs/msg.split.*`
for msg in $files ; do
    nam=`basename $msg`
    out="$TMPDIR/$nam"
    map_rc $BOGOLEXER -C -p $v -I $msg > $out.l
    map_rc $BOGOFILTER -C -vvv -I $msg > $out.f
    printf "%-24s %s\n" $nam `head -1 $out.f | sed s@spamicity=@@ \
    | tr -d "," | $AWK '{print $4}'` >> ${TMPDIR}/$OUT
done

if [ "$BF_SAVEDIR" ] ; then . ${srcdir}/t.save ; fi

test $verbose -eq 0 && cmp ${OUTPUTS}/$OUT ${TMPDIR}/$OUT \
|| diff $DIFF_BRIEF ${OUTPUTS}/$OUT ${TMPDIR}/$OUT