blob: a1a0de65585820f5734ea5cc5108f366e78bac30 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
|
#! /bin/sh
# @(#) $Id: test-recode.sh,v 1.2 2003/11/17 12:27:40 yeti Exp $
# Purpose: test whether librecode interface works.
# FIXME: this may fail when the interface works but librecode is broken!
. $srcdir/setup.sh
if $ENCA --list converters | grep '^librecode$' >/dev/null; then
TEST_TEXT=$srcdir/cs-s.iso88592
OPTS="-L cs -C librecode"
# File
cp $TEST_TEXT $TESTNAME.actual
$ENCA $OPTS -x UTF-8 $TESTNAME.actual || DIE=1
$ENCA -L none $TESTNAME.actual | grep UTF-8 >/dev/null || DIE=1
$ENCA $OPTS -x ISO-8859-2 $TESTNAME.actual || DIE=1
diff $TEST_TEXT $TESTNAME.actual || DIE=1
# Pipe
cp $TEST_TEXT $TESTNAME.actual
$ENCA $OPTS -x UTF-8 <$TESTNAME.actual >$TESTNAME.tmp || DIE=1
$ENCA -L none $TESTNAME.tmp | grep UTF-8 >/dev/null || DIE=1
$ENCA $OPTS -x ISO-8859-2 <$TESTNAME.tmp >$TESTNAME.actual || DIE=1
diff $TEST_TEXT $TESTNAME.actual || DIE=1
# Failures
cp $TEST_TEXT $TESTNAME.actual
$ENCA $OPTS -x solzenicyn $TESTNAME.actual 2>/dev/null && DIE=1
diff $TEST_TEXT $TESTNAME.actual || DIE=1
# TODO: this is not enough for recode to fail, it just wipes out the
# offending characters
# One copy doesn't contain enough characters to overweight the noise
#cat $TESTNAME.tmp $TESTNAME.tmp $TESTNAME.tmp >$TESTNAME.actual
#echo '�' >>$TESTNAME.actual
#cat $TESTNAME.actual >$TESTNAME.tmp
#$ENCA $OPTS -x ISO-8859-2 $TESTNAME.tmp && DIE=1
#diff $TESTNAME.tmp $TESTNAME.actual || DIE=1
else
E77=1
fi
. $srcdir/finish.sh
|
