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
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
|
#!/bin/sh
############################################################
#
# Test scan to see if multibyte support (UTF-8 locale) works
#
# Other tests will get the normal ASCII case, so all we care
# about here is UTF-8 encoded headers (RFC 2047).
#
# Note that this file should be edited via a UTF-8 aware
# editor, since UTF-8 characters are in it.
#
############################################################
set -e
if test -z "${MH_OBJ_DIR}"; then
srcdir=`dirname "$0"`/../..
MH_OBJ_DIR=`cd "$srcdir" && pwd`; export MH_OBJ_DIR
fi
. "$MH_OBJ_DIR/test/common.sh"
setup_test
if test "${MULTIBYTE_ENABLED}" -ne 1; then
test_skip "configure did not detect multibyte support"
fi
require_locale $en_locales
#
# Create a test message with RFC 2047 headers we can scan
#
# In this Subject header in this message is a "n" with a Combining Diaeresis
# (U+0308). There is different interpretation of this character with respect
# to wcwidth() (which is supposed to return the column width of a character).
# We use a test program to determine what the output width of U+0308 is
# and adjust our test output appropriately.
#
# True Spın̈al Tap fans will note that David st Hubbins was born in Squatney,
# London, England, and thus having his name language-tagged with "cy" is almost
# certainly incorrect. But in his own words: "Here lies David st Hubbins,
# and why not?".
#
# The second "* in the To line is just to exercise the parser a bit.
#
cat > "${MH_TEST_DIR}/Mail/inbox/11" <<EOF
From: David =?utf-8*cy?q?=EF=AC=86?= Hubbins <hubbins@example.com>
To: Sir Denis =?utf-8*?q?Eton=E2=80=93Hogg? <sirdenis@example.com>
Date: Friday, 2 Mar 1984 00:00:00
Subject: =?utf-8?q?Sp=C4=B1n=CC=88al_Tap_=E2=86=92_Tap_into_America!?=
Things are looking great!
EOF
width=`${MH_OBJ_DIR}/test/getcwidth "→n̈"`
if test $? -ne 0; then
echo "getcwidth failed to run"
exit 1
fi
expected="$MH_TEST_DIR/$$.expected"
actual="$MH_TEST_DIR/$$.actual"
start_test 'RFC 2047 headers'
if test "$width" -eq 3; then
cat > "$expected" <<EOF
11 03/02 David st Hubbins Spın̈al Tap → Tap into America!<<Things are looki
EOF
elif test "$width" -eq 2; then
cat > "$expected" <<EOF
11 03/02 David st Hubbins Spın̈al Tap → Tap into America!<<Things are lookin
EOF
else
echo "Unsupported width for UTF-8 test string: $width"
exit 1
fi
run_prog scan -width 80 +inbox 11 > $actual || exit 1
check "$expected" "$actual"
#
# Check decoding with an invalid multibyte sequence. We skip this test
# if we don't have iconv support, since it requires converting from one
# character set to another. Be sure we created the test file, though, because
# it's required for the test right after it.
#
start_test 'invalid multibyte sequence'
cat >`mhpath new` <<EOF
From: Test12 <test12@example.com>
To: Some User <user@example.com>
Date: Mon, 31 Dec 2012 00:00:00
Message-Id: 12@test.nmh
Subject: =?UTF-8?B?MjAxMyBOZXcgWWVhcuKAmXMgRGVhbHMhIFN0YXJ0IHRoZSB5ZWFy?=
=?UTF-8?B?IHJpZ2h0IHdpdGggYmlnIHNhdmluZ3M=?=
This message has an encoded Subject with an invalid character for
single-byte character sets, but it (U+2019) is valid UTF-8.
EOF
if test "$ICONV_ENABLED" -eq 1; then
cat >"$expected" <<EOF
12 12/31 Test12 2013 New Year?s Deals! Start the year right
EOF
# Don't use run_prog here because it loses the environment setting.
LC_ALL=C scan -width 74 last >"$actual"
check "$expected" "$actual"
fi
#
# Find out the width of our Unicode apostrophe (U+2019). Some implementations
# say it has a width of 2, but that seems totally bizarre to me.
#
width=`${MH_OBJ_DIR}/test/getcwidth U+2019`
if test $? -ne 0; then
echo "getcwidth failed to run"
exit 1
fi
# check scan width with a valid multibyte sequence
start_test 'scan width with a valid multibyte sequence'
if test "$width" -eq 1; then
cat >"$expected" <<EOF
12 12/31 Test12 2013 New Year’s Deals! Start the year right
EOF
elif test "$width" -eq 2; then
cat >"$expected" <<EOF
12 12/31 Test12 2013 New Year’s Deals! Start the year righ
EOF
else
echo "Unsupported width for U+2019: $width"
fi
run_prog scan -width 74 last >"$actual"
check "$expected" "$actual"
if test "$ICONV_ENABLED" -eq 1; then
start_test 'encoded single quote'
cat >"$expected" <<EOF
13 01/13 sender@example.co <<The Subject: is an encoded single quote, 0x92.
EOF
cat >"${MH_TEST_DIR}/Mail/inbox/13" <<EOF
From: <sender@example.com>
Subject: =?iso-8859-1?B?kgo=?=
Date: Mon, 13 Jan 2014 14:18:33 -0600
The Subject: is an encoded single quote, 0x92. cpstripped() didn't
properly count it when decoding, which could be seen with:
scan -format '%(decode{subject})%{body}'
The scan listing was two characters too long.
EOF
run_prog scan -width 80 last >"$actual"
check "$expected" "$actual"
fi
start_test 'insufficient room for multicolumn character'
#### This multibyte character requires 2 columns for display, but
#### only 1 is availble. cpstripped() used to get this wrong.
cat >"$expected" <<EOF
EOF
cat >`mhpath new` <<EOF
Mime-Version: 1.0
在 Should not see any of this text
EOF
run_prog scan -format '%{body}' -width 1 last >"$actual"
check "$expected" "$actual"
finish_test
exit $failed
|
