Commit | Line | Data |
---|---|---|
173ee337 KW |
1 | # Tools to aid testing across platforms with different character sets. |
2 | ||
3 | $::IS_ASCII = ord 'A' == 65; | |
4 | $::IS_EBCDIC = ord 'A' == 193; | |
5 | ||
6 | # The following functions allow tests to work on both EBCDIC and ASCII-ish | |
7 | # platforms. They convert string scalars between the native character set and | |
8 | # the set of 256 characters which is usually called Latin1. However, they | |
9 | # will work properly with any character input, not just Latin1. | |
10 | ||
d1cef54a | 11 | sub native_to_uni($) { |
173ee337 KW |
12 | my $string = shift; |
13 | ||
14 | return $string if $::IS_ASCII; | |
15 | my $output = ""; | |
16 | for my $i (0 .. length($string) - 1) { | |
17 | $output .= chr(utf8::native_to_unicode(ord(substr($string, $i, 1)))); | |
18 | } | |
19 | # Preserve utf8ness of input onto the output, even if it didn't need to be | |
20 | # utf8 | |
21 | utf8::upgrade($output) if utf8::is_utf8($string); | |
22 | ||
23 | return $output; | |
24 | } | |
25 | ||
d1cef54a | 26 | sub uni_to_native($) { |
173ee337 KW |
27 | my $string = shift; |
28 | ||
29 | return $string if $::IS_ASCII; | |
30 | my $output = ""; | |
31 | for my $i (0 .. length($string) - 1) { | |
98c62be8 | 32 | $output .= chr(utf8::unicode_to_native(ord(substr($string, $i, 1)))); |
173ee337 KW |
33 | } |
34 | # Preserve utf8ness of input onto the output, even if it didn't need to be | |
35 | # utf8 | |
36 | utf8::upgrade($output) if utf8::is_utf8($string); | |
37 | ||
38 | return $output; | |
39 | } | |
40 | ||
41 | sub byte_utf8a_to_utf8n { | |
42 | # Convert a UTF-8 byte sequence into the platform's native UTF-8 | |
43 | # equivalent, currently only UTF-8 and UTF-EBCDIC. | |
44 | ||
45 | my @utf8_skip = ( | |
46 | # This translates a utf-8-encoded byte into how many bytes the full utf8 | |
47 | # character occupies. | |
48 | ||
49 | # 0 1 2 3 4 5 6 7 8 9 A B C D E F | |
50 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, # 0 | |
51 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, # 1 | |
52 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, # 2 | |
53 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, # 3 | |
54 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, # 4 | |
55 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, # 5 | |
56 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, # 6 | |
57 | 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, # 7 | |
58 | -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, # 8 | |
59 | -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, # 9 | |
60 | -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, # A | |
61 | -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, # B | |
62 | -1,-1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, # C | |
63 | 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, # D | |
64 | 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, # E | |
65 | 4, 4, 4, 4, 4, 4, 4, 4, 5, 5, 5, 5, 6, 6, 7,13, # F | |
66 | ); | |
67 | ||
68 | my $string = shift; | |
69 | die "Input to byte_utf8a-to_utf8n() must not be flagged UTF-8" | |
70 | if utf8::is_utf8($string); | |
71 | return $string if $::IS_ASCII; | |
72 | die "Expecting ASCII or EBCDIC" unless $::IS_EBCDIC; | |
73 | ||
74 | my $length = length($string); | |
75 | #diag($string); | |
76 | #diag($length); | |
77 | my $out = ""; | |
78 | for ($i = 0; $i < $length; $i++) { | |
79 | my $byte = ord substr($string, $i, 1); | |
80 | my $byte_count = $utf8_skip[$byte]; | |
81 | #diag($byte); | |
82 | #diag($byte_count); | |
83 | ||
84 | die "Illegal start byte" if $byte_count < 0; | |
85 | if ($i + $byte_count > $length) { | |
86 | die "Attempt to read " . $i + $byte_count - $length . " beyond end-of-string"; | |
87 | } | |
88 | ||
89 | # Just translate UTF-8 invariants directly. | |
90 | if ($byte_count == 1) { | |
91 | $out .= chr utf8::unicode_to_native($byte); | |
92 | next; | |
93 | } | |
94 | ||
95 | # Otherwise calculate the code point ordinal represented by the | |
96 | # sequence beginning with this byte, using the algorithm adapted from | |
97 | # utf8.c. We absorb each byte in the sequence as we go along | |
98 | my $ord = $byte & (0x1F >> ($byte_count - 2)); | |
99 | my $bytes_remaining = $byte_count - 1; | |
100 | while ($bytes_remaining > 0) { | |
101 | $byte = ord substr($string, ++$i, 1); | |
102 | unless (($byte & 0xC0) == 0x80) { | |
103 | die sprintf "byte '%X' is not a valid continuation", $byte; | |
104 | } | |
105 | $ord = $ord << 6 | ($byte & 0x3f); | |
106 | $bytes_remaining--; | |
107 | } | |
108 | #diag($byte); | |
109 | #diag($ord); | |
110 | ||
111 | my $expected_bytes = $ord < 0x80 | |
112 | ? 1 | |
113 | : $ord < 0x800 | |
114 | ? 2 | |
115 | : $ord < 0x10000 | |
116 | ? 3 | |
117 | : $ord < 0x200000 | |
118 | ? 4 | |
119 | : $ord < 0x4000000 | |
120 | ? 5 | |
121 | : $ord < 0x80000000 | |
122 | ? 6 | |
123 | : 7; | |
124 | #: (uv) < UTF8_QUAD_MAX ? 7 : 13 ) | |
125 | ||
126 | # Make sure is not an overlong sequence | |
127 | if ($byte_count != $expected_bytes) { | |
128 | die sprintf "character U+%X should occupy %d bytes, not %d", | |
129 | $ord, $expected_bytes, $byte_count; | |
130 | } | |
131 | ||
132 | # Now that we have found the code point the original UTF-8 meant, we | |
133 | # use the native chr function to get its native string equivalent. | |
134 | $out .= chr utf8::unicode_to_native($ord); | |
135 | } | |
136 | ||
137 | utf8::encode($out); # Turn off utf8 flag. | |
138 | #diag($out); | |
139 | return $out; | |
140 | } | |
141 | ||
5d4414ef KW |
142 | my @i8_to_native = ( # Only code page 1047 so far. |
143 | # _0 _1 _2 _3 _4 _5 _6 _7 _8 _9 _A _B _C _D _E _F | |
144 | 0x00,0x01,0x02,0x03,0x37,0x2D,0x2E,0x2F,0x16,0x05,0x15,0x0B,0x0C,0x0D,0x0E,0x0F, | |
145 | 0x10,0x11,0x12,0x13,0x3C,0x3D,0x32,0x26,0x18,0x19,0x3F,0x27,0x1C,0x1D,0x1E,0x1F, | |
146 | 0x40,0x5A,0x7F,0x7B,0x5B,0x6C,0x50,0x7D,0x4D,0x5D,0x5C,0x4E,0x6B,0x60,0x4B,0x61, | |
147 | 0xF0,0xF1,0xF2,0xF3,0xF4,0xF5,0xF6,0xF7,0xF8,0xF9,0x7A,0x5E,0x4C,0x7E,0x6E,0x6F, | |
148 | 0x7C,0xC1,0xC2,0xC3,0xC4,0xC5,0xC6,0xC7,0xC8,0xC9,0xD1,0xD2,0xD3,0xD4,0xD5,0xD6, | |
149 | 0xD7,0xD8,0xD9,0xE2,0xE3,0xE4,0xE5,0xE6,0xE7,0xE8,0xE9,0xAD,0xE0,0xBD,0x5F,0x6D, | |
150 | 0x79,0x81,0x82,0x83,0x84,0x85,0x86,0x87,0x88,0x89,0x91,0x92,0x93,0x94,0x95,0x96, | |
151 | 0x97,0x98,0x99,0xA2,0xA3,0xA4,0xA5,0xA6,0xA7,0xA8,0xA9,0xC0,0x4F,0xD0,0xA1,0x07, | |
152 | 0x20,0x21,0x22,0x23,0x24,0x25,0x06,0x17,0x28,0x29,0x2A,0x2B,0x2C,0x09,0x0A,0x1B, | |
153 | 0x30,0x31,0x1A,0x33,0x34,0x35,0x36,0x08,0x38,0x39,0x3A,0x3B,0x04,0x14,0x3E,0xFF, | |
154 | 0x41,0x42,0x43,0x44,0x45,0x46,0x47,0x48,0x49,0x4A,0x51,0x52,0x53,0x54,0x55,0x56, | |
155 | 0x57,0x58,0x59,0x62,0x63,0x64,0x65,0x66,0x67,0x68,0x69,0x6A,0x70,0x71,0x72,0x73, | |
156 | 0x74,0x75,0x76,0x77,0x78,0x80,0x8A,0x8B,0x8C,0x8D,0x8E,0x8F,0x90,0x9A,0x9B,0x9C, | |
157 | 0x9D,0x9E,0x9F,0xA0,0xAA,0xAB,0xAC,0xAE,0xAF,0xB0,0xB1,0xB2,0xB3,0xB4,0xB5,0xB6, | |
158 | 0xB7,0xB8,0xB9,0xBA,0xBB,0xBC,0xBE,0xBF,0xCA,0xCB,0xCC,0xCD,0xCE,0xCF,0xDA,0xDB, | |
159 | 0xDC,0xDD,0xDE,0xDF,0xE1,0xEA,0xEB,0xEC,0xED,0xEE,0xEF,0xFA,0xFB,0xFC,0xFD,0xFE, | |
160 | ); | |
161 | ||
162 | my @native_to_i8; | |
163 | for (my $i = 0; $i < 256; $i++) { | |
164 | $native_to_i8[$i8_to_native[$i]] = $i; | |
165 | } | |
166 | ||
05d49a9a KW |
167 | # Use these to convert to/from UTF-8 bytes. I8 is the encoding that |
168 | # corresponds to UTF-8 with start bytes, continuation bytes, and invariant | |
169 | # bytes. UTF-EBCDIC is derived from this by a mapping which causes things | |
170 | # like the start byte C5 to map to something else, as C5 is actually an 'E' in | |
171 | # EBCDIC so can't be a real start byte, as it must be an invariant; and it | |
172 | # maps 0x45 (an ASCII 'E') to C5. | |
5d4414ef KW |
173 | *I8_to_native = ($::IS_ASCII) |
174 | ? sub { return shift } | |
175 | : sub { return join "", map { chr $i8_to_native[ord $_] } | |
176 | split "", shift }; | |
177 | *native_to_I8 = ($::IS_ASCII) | |
178 | ? sub { return shift } | |
179 | : sub { return join "", map { chr $native_to_i8[ord $_] } | |
180 | split "", shift }; | |
181 | ||
173ee337 | 182 | 1 |