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
|
# The encoding detection heuristic will choose UTF8 or Latin-1. The current
# implementation will usually treat CP1252 (aka "Win-Latin-1") as Latin-1 but
# can be fooled into seeing it as UTF8.
#
# Note 1: Neither guess is 'correct' since even if we choose Latin-1, all the
# smart quote symbols will be rendered as control characters
#
# Note 2: the guess is only applied if the source POD omits =encoding, so
# CP1252 source will render correctly if properly declared
#
BEGIN {
if($ENV{PERL_CORE}) {
chdir 't';
@INC = '../lib';
}
}
use strict;
use Test;
BEGIN { plan tests => 5 };
ok 1;
use Pod::Simple::DumpAsXML;
use Pod::Simple::XMLOutStream;
# Initial, isolated, non-ASCII byte triggers Latin-1 guess and later
# multi-byte sequence is not considered by heuristic.
my @output_lines = split m/[\cm\cj]+/, Pod::Simple::XMLOutStream->_out( qq{
=head1 NAME
Em::Dash \x97 \x91CAF\xC9\x92
=cut
} );
my($guess) = "@output_lines" =~ m{Non-ASCII.*?Assuming ([\w-]+)};
if( $guess ) {
if( $guess eq 'ISO8859-1' ) {
if( grep m{Dash (\x97|—|—)}, @output_lines ) {
ok 1;
} else {
ok 0;
print "# failed to find expected control character in output\n"
}
} else {
ok 0;
print "# parser guessed wrong encoding expected 'ISO8859-1' got '$guess'\n";
}
} else {
ok 0;
print "# parser failed to detect non-ASCII bytes in input\n";
}
# Initial smart-quote character triggers Latin-1 guess as expected
@output_lines = split m/[\cm\cj]+/, Pod::Simple::XMLOutStream->_out( qq{
=head1 NAME
Smart::Quote - \x91FUT\xC9\x92
=cut
} );
($guess) = "@output_lines" =~ m{Non-ASCII.*?Assuming ([\w-]+)};
if( $guess ) {
if( $guess eq 'ISO8859-1' ) {
ok 1;
} else {
ok 0;
print "# parser guessed wrong encoding expected 'ISO8859-1' got '$guess'\n";
}
} else {
ok 0;
print "# parser failed to detect non-ASCII bytes in input\n";
}
# Initial accented character followed by 'smart' apostrophe causes heuristic
# to choose UTF8 (a rather contrived example)
@output_lines = split m/[\cm\cj]+/, Pod::Simple::XMLOutStream->_out( qq{
=head1 NAME
Smart::Apostrophe::Fail - L\xC9\x92STRANGE
=cut
} );
($guess) = "@output_lines" =~ m{Non-ASCII.*?Assuming ([\w-]+)};
if( $guess ) {
if( $guess eq 'UTF-8' ) {
ok 1;
} else {
ok 0;
print "# parser guessed wrong encoding expected 'UTF-8' got '$guess'\n";
}
} else {
ok 0;
print "# parser failed to detect non-ASCII bytes in input\n";
}
# The previous example used a CP1252 byte sequence that also happened to be a
# valid UTF8 byte sequence. In this example the heuristic also guesses 'wrong'
# despite the byte sequence not being valid UTF8 (it's too short). This could
# arguably be 'fixed' by using a less naive regex.
@output_lines = split m/[\cm\cj]+/, Pod::Simple::XMLOutStream->_out( qq{
=head1 NAME
Smart::Apostrophe::Fail - L\xE9\x92Strange
=cut
} );
($guess) = "@output_lines" =~ m{Non-ASCII.*?Assuming ([\w-]+)};
if( $guess ) {
if( $guess eq 'UTF-8' ) {
ok 1;
} else {
ok 0;
print "# parser guessed wrong encoding expected 'UTF-8' got '$guess'\n";
}
} else {
ok 0;
print "# parser failed to detect non-ASCII bytes in input\n";
}
exit;
|
