t.dump.load

一个C语言写的快速贝叶斯垃圾邮件过滤工具

#!/bin/sh

#Test database maintenance using bogoutil's dump/load capabilities.

. ${srcdir:=.}/t.frame

BASE="dump.load"
OUT="${TMPDIR}/$BASE.out"
DATA="${TMPDIR}/$BASE.${DB_EXT}"
SUM="$BASE.sum"

GREP="grep -v /dump.load-"
SORT="sort"

clear_database() {
    rm -f $DATA
    rm -f ${TMPDIR}/__db.*
    rm -f ${TMPDIR}/log.*
}

clear_database

LEN="--max-token-len 30"

#load database
$BOGOUTIL -C $LEN -l $DATA -y 20020815 < ${srcdir}/inputs/$BASE.inp

#dump database (no restrictions, all dates should by 'yday')
$BOGOUTIL -C $LEN -d $DATA -y 20020815 | $SORT > ${TMPDIR}/dump.load-1.txt

#update some words, including some old dates
$BOGOUTIL -C $LEN -l $DATA -y 20021215 < ${srcdir}/inputs/$BASE.upd
#dump database
$BOGOUTIL -C $LEN -d $DATA | grep -v "^.ENCODING" | $SORT > ${TMPDIR}/dump.load-2.txt

#confirm that updated words have different counts and dates
( diff ${TMPDIR}/dump.load-1.txt ${TMPDIR}/dump.load-2.txt ; echo "" ) | $GREP | tee ${TMPDIR}/diff.1.2.txt >> $OUT

#dump wordlist, excluding oldest entries
$BOGOUTIL -C $LEN -a 20020815 -d $DATA | grep -v "^.ENCODING" | $SORT > ${TMPDIR}/dump.load-3.txt

#confirm exclusion of oldest entries
( diff ${TMPDIR}/dump.load-2.txt ${TMPDIR}/dump.load-3.txt ; echo "" ) | $GREP | tee ${TMPDIR}/diff.2.3.txt >> $OUT

#dump wordlist, excluding oldest, shortest, and longest entries
#     and'ing of command line qualifiers forced changing next line to two lines
#BOGOUTIL -C $LEN -a 20020815 -s4,30 -d $DATA | $SORT > ${TMPDIR}/dump.load-4.txt
$BOGOUTIL -C $LEN -a 20020815 -d $DATA | $SORT > ${TMPDIR}/dump.load-4.txt
clear_database
$BOGOUTIL -C $LEN -l $DATA < ${TMPDIR}/dump.load-4.txt
$BOGOUTIL -C $LEN -s4,30      -d $DATA | $SORT > ${TMPDIR}/dump.load-4.txt

#confirm exclusion of oldest, shortest(<, and longest entries
( diff ${TMPDIR}/dump.load-3.txt ${TMPDIR}/dump.load-4.txt ; echo "" ) | $GREP | tee ${TMPDIR}/diff.3.4.txt >> $OUT

clear_database

#load wordlist, excluding shortest entries
$BOGOUTIL -C $LEN -s5,30 -l $DATA < ${TMPDIR}/dump.load-4.txt
#dump using replace_non-ascii option
$BOGOUTIL -C $LEN -n -d $DATA | $SORT > ${TMPDIR}/dump.load-5.txt

#confirm effect of non-ascii option
( diff ${TMPDIR}/dump.load-4.txt ${TMPDIR}/dump.load-5.txt ; echo "" ) | $GREP | tee ${TMPDIR}/diff.4.5.txt >> $OUT

clear_database

#load wordlist non-ascii list and dump to show value combining
$BOGOUTIL -C $LEN -a 20021010 -l $DATA < ${TMPDIR}/dump.load-5.txt
$BOGOUTIL -C $LEN -a 20021010 -d $DATA | $SORT > ${TMPDIR}/dump.load-6.txt

#confirm token merging
( diff ${TMPDIR}/dump.load-5.txt ${TMPDIR}/dump.load-6.txt ; echo "" ) | $GREP | tee ${TMPDIR}/diff.5.6.txt >> $OUT

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

for i in ${SYSTEST}/outputs/dump.load-?.out ; do
    j=${TMPDIR}/`basename $i .out`.txt
    if  [ $verbose -eq 0 ]; then
	cmp $i $j
    else
	diff $DIFF_BRIEF $i $j
    fi
done