1--TEST--
2Exhaustive test of verification and conversion of CP950 text
3--EXTENSIONS--
4mbstring
5--SKIPIF--
6<?php
7if (getenv("SKIP_SLOW_TESTS")) die("skip slow test");
8?>
9--FILE--
10<?php
11include('encoding_tests.inc');
12srand(1000); // Make results consistent
13mb_substitute_character(0x25); // '%'
14readConversionTable(__DIR__ . '/data/CP950.txt', $toUnicode, $fromUnicode);
15
16// Non-reversible mappings (two different CP950 codepoints map to same Unicode codepoint)
17$fromUnicode["\x25\x50"] = "\xA2\xA4";
18$fromUnicode["\x25\x5E"] = "\xA2\xA5";
19$fromUnicode["\x25\x61"] = "\xA2\xA7";
20$fromUnicode["\x25\x6A"] = "\xA2\xA6";
21$fromUnicode["\x25\x6D"] = "\xA2\x7E";
22$fromUnicode["\x25\x6E"] = "\xA2\xA1";
23$fromUnicode["\x25\x6F"] = "\xA2\xA3";
24$fromUnicode["\x25\x70"] = "\xA2\xA2";
25
26// CP950 explicitly maps some unused Big5 codepoint ranges to Unicode "Private User Area" codepoints
27
28function addPUAMapping($hi, $lo, $unicode) {
29  global $fromUnicode, $toUnicode;
30  $cp950 = pack('n', ($hi << 8) + $lo);
31  $unicode = pack('n', $unicode);
32  $fromUnicode[$unicode] = $cp950;
33  $toUnicode[$cp950] = $unicode;
34}
35
36for ($hi = 0x81; $hi <= 0x8D; $hi++) {
37  for ($lo = 0x40; $lo <= 0x7E; $lo++)
38    addPUAMapping($hi, $lo, 0xEEB8 + (157 * ($hi - 0x81)) + ($lo - 0x40));
39  for ($lo = 0xA1; $lo <= 0xFE; $lo++)
40    addPUAMapping($hi, $lo, 0xEEB8 + (157 * ($hi - 0x81)) + ($lo - 0x62));
41}
42
43for ($hi = 0x8E; $hi <= 0xA0; $hi++) {
44  for ($lo = 0x40; $lo <= 0x7E; $lo++)
45    addPUAMapping($hi, $lo, 0xE311 + (157 * ($hi - 0x8E)) + ($lo - 0x40));
46  for ($lo = 0xA1; $lo <= 0xFE; $lo++)
47    addPUAMapping($hi, $lo, 0xE311 + (157 * ($hi - 0x8E)) + ($lo - 0x62));
48}
49
50$hi = 0xC6;
51for ($lo = 0xA1; $lo <= 0xFE; $lo++)
52  addPUAMapping($hi, $lo, 0xF672 + (157 * ($hi - 0xC6)) + ($lo - 0x62));
53
54for ($hi = 0xC7; $hi <= 0xC8; $hi++) {
55  for ($lo = 0x40; $lo <= 0x7E; $lo++)
56    addPUAMapping($hi, $lo, 0xF672 + (157 * ($hi - 0xC6)) + ($lo - 0x40));
57  for ($lo = 0xA1; $lo <= 0xFE; $lo++)
58    addPUAMapping($hi, $lo, 0xF672 + (157 * ($hi - 0xC6)) + ($lo - 0x62));
59}
60
61for ($hi = 0xFA; $hi <= 0xFE; $hi++) {
62  for ($lo = 0x40; $lo <= 0x7E; $lo++)
63    addPUAMapping($hi, $lo, 0xE000 + (157 * ($hi - 0xFA)) + ($lo - 0x40));
64  for ($lo = 0xA1; $lo <= 0xFE; $lo++)
65    addPUAMapping($hi, $lo, 0xE000 + (157 * ($hi - 0xFA)) + ($lo - 0x62));
66}
67
68findInvalidChars($toUnicode, $invalid, $truncated);
69testAllValidChars($toUnicode, 'CP950', 'UTF-16BE', false);
70testAllInvalidChars($invalid, $toUnicode, 'CP950', 'UTF-16BE', "\x00%");
71testTruncatedChars($truncated, 'CP950', 'UTF-16BE', "\x00%");
72echo "Tested CP950 -> UTF-16BE\n";
73
74testAllValidChars($fromUnicode, 'UTF-16BE', 'CP950', false);
75findInvalidChars($fromUnicode, $invalid, $unused, array_fill_keys(range(0,0xFF), 2));
76convertAllInvalidChars($invalid, $fromUnicode, 'UTF-16BE', 'CP950', '%');
77echo "Tested UTF-16BE -> CP950\n";
78
79// Test "long" illegal character markers
80mb_substitute_character("long");
81convertInvalidString("\x80", "%", "CP950", "UTF-8");
82convertInvalidString("\x26\x09", "U+2609", "UTF-16BE", "CP950");
83
84echo "Done!\n";
85?>
86--EXPECT--
87Tested CP950 -> UTF-16BE
88Tested UTF-16BE -> CP950
89Done!
90