xref: /dokuwiki/inc/Utf8/Asian.php (revision dbc189b29a5c7d5fe8dffbe81de5ce8d624e519d)
1*dbc189b2SAndreas Gohr<?php
2*dbc189b2SAndreas Gohr
3*dbc189b2SAndreas Gohrnamespace dokuwiki\Utf8;
4*dbc189b2SAndreas Gohr
5*dbc189b2SAndreas Gohr/**
6*dbc189b2SAndreas Gohr * Methods and constants to handle Asian "words"
7*dbc189b2SAndreas Gohr *
8*dbc189b2SAndreas Gohr * This uses a crude regexp to determine which parts of an Asian string should be treated as words.
9*dbc189b2SAndreas Gohr * This is necessary because in some Asian languages a single unicode char represents a whole idea
10*dbc189b2SAndreas Gohr * without spaces separating them.
11*dbc189b2SAndreas Gohr */
12*dbc189b2SAndreas Gohrclass Asian
13*dbc189b2SAndreas Gohr{
14*dbc189b2SAndreas Gohr
15*dbc189b2SAndreas Gohr    /**
16*dbc189b2SAndreas Gohr     * This defines a non-capturing group for the use in regular expressions to match any asian character that
17*dbc189b2SAndreas Gohr     * needs to be treated as a word. Uses the Unicode-Ranges for Asian characters taken from
18*dbc189b2SAndreas Gohr     * http://en.wikipedia.org/wiki/Unicode_block
19*dbc189b2SAndreas Gohr     */
20*dbc189b2SAndreas Gohr    const REGEXP =
21*dbc189b2SAndreas Gohr        '(?:' .
22*dbc189b2SAndreas Gohr
23*dbc189b2SAndreas Gohr        '[\x{0E00}-\x{0E7F}]' . // Thai
24*dbc189b2SAndreas Gohr
25*dbc189b2SAndreas Gohr        '|' .
26*dbc189b2SAndreas Gohr
27*dbc189b2SAndreas Gohr        '[' .
28*dbc189b2SAndreas Gohr        '\x{2E80}-\x{3040}' .  // CJK -> Hangul
29*dbc189b2SAndreas Gohr        '\x{309D}-\x{30A0}' .
30*dbc189b2SAndreas Gohr        '\x{30FD}-\x{31EF}\x{3200}-\x{D7AF}' .
31*dbc189b2SAndreas Gohr        '\x{F900}-\x{FAFF}' .  // CJK Compatibility Ideographs
32*dbc189b2SAndreas Gohr        '\x{FE30}-\x{FE4F}' .  // CJK Compatibility Forms
33*dbc189b2SAndreas Gohr        "\xF0\xA0\x80\x80-\xF0\xAA\x9B\x9F" . // CJK Extension B
34*dbc189b2SAndreas Gohr        "\xF0\xAA\x9C\x80-\xF0\xAB\x9C\xBF" . // CJK Extension C
35*dbc189b2SAndreas Gohr        "\xF0\xAB\x9D\x80-\xF0\xAB\xA0\x9F" . // CJK Extension D
36*dbc189b2SAndreas Gohr        "\xF0\xAF\xA0\x80-\xF0\xAF\xAB\xBF" . // CJK Compatibility Supplement
37*dbc189b2SAndreas Gohr        ']' .
38*dbc189b2SAndreas Gohr
39*dbc189b2SAndreas Gohr        '|' .
40*dbc189b2SAndreas Gohr
41*dbc189b2SAndreas Gohr        '[' .                // Hiragana/Katakana (can be two characters)
42*dbc189b2SAndreas Gohr        '\x{3042}\x{3044}\x{3046}\x{3048}' .
43*dbc189b2SAndreas Gohr        '\x{304A}-\x{3062}\x{3064}-\x{3082}' .
44*dbc189b2SAndreas Gohr        '\x{3084}\x{3086}\x{3088}-\x{308D}' .
45*dbc189b2SAndreas Gohr        '\x{308F}-\x{3094}' .
46*dbc189b2SAndreas Gohr        '\x{30A2}\x{30A4}\x{30A6}\x{30A8}' .
47*dbc189b2SAndreas Gohr        '\x{30AA}-\x{30C2}\x{30C4}-\x{30E2}' .
48*dbc189b2SAndreas Gohr        '\x{30E4}\x{30E6}\x{30E8}-\x{30ED}' .
49*dbc189b2SAndreas Gohr        '\x{30EF}-\x{30F4}\x{30F7}-\x{30FA}' .
50*dbc189b2SAndreas Gohr        '][' .
51*dbc189b2SAndreas Gohr        '\x{3041}\x{3043}\x{3045}\x{3047}\x{3049}' .
52*dbc189b2SAndreas Gohr        '\x{3063}\x{3083}\x{3085}\x{3087}\x{308E}\x{3095}-\x{309C}' .
53*dbc189b2SAndreas Gohr        '\x{30A1}\x{30A3}\x{30A5}\x{30A7}\x{30A9}' .
54*dbc189b2SAndreas Gohr        '\x{30C3}\x{30E3}\x{30E5}\x{30E7}\x{30EE}\x{30F5}\x{30F6}\x{30FB}\x{30FC}' .
55*dbc189b2SAndreas Gohr        '\x{31F0}-\x{31FF}' .
56*dbc189b2SAndreas Gohr        ']?' .
57*dbc189b2SAndreas Gohr        ')';
58*dbc189b2SAndreas Gohr
59*dbc189b2SAndreas Gohr
60*dbc189b2SAndreas Gohr    /**
61*dbc189b2SAndreas Gohr     * Check if the given term contains Asian word characters
62*dbc189b2SAndreas Gohr     *
63*dbc189b2SAndreas Gohr     * @param string $term
64*dbc189b2SAndreas Gohr     * @return bool
65*dbc189b2SAndreas Gohr     */
66*dbc189b2SAndreas Gohr    public static function isAsianWords($term)
67*dbc189b2SAndreas Gohr    {
68*dbc189b2SAndreas Gohr        return (bool)preg_match('/' . self::REGEXP . '/u', $term);
69*dbc189b2SAndreas Gohr    }
70*dbc189b2SAndreas Gohr
71*dbc189b2SAndreas Gohr    /**
72*dbc189b2SAndreas Gohr     * Surround all Asian words in the given text with the given separator
73*dbc189b2SAndreas Gohr     *
74*dbc189b2SAndreas Gohr     * @param string $text Original text containing asian words
75*dbc189b2SAndreas Gohr     * @param string $sep the separator to use
76*dbc189b2SAndreas Gohr     * @return string Text with separated asian words
77*dbc189b2SAndreas Gohr     */
78*dbc189b2SAndreas Gohr    public static function separateAsianWords($text, $sep = ' ')
79*dbc189b2SAndreas Gohr    {
80*dbc189b2SAndreas Gohr        // handle asian chars as single words (may fail on older PHP version)
81*dbc189b2SAndreas Gohr        $asia = @preg_replace('/(' . self::REGEXP . ')/u', $sep . '\1' . $sep, $text);
82*dbc189b2SAndreas Gohr        if (!is_null($asia)) $text = $asia; // recover from regexp falure
83*dbc189b2SAndreas Gohr
84*dbc189b2SAndreas Gohr        return $text;
85*dbc189b2SAndreas Gohr    }
86*dbc189b2SAndreas Gohr
87*dbc189b2SAndreas Gohr    /**
88*dbc189b2SAndreas Gohr     * Split the given text into separate parts
89*dbc189b2SAndreas Gohr     *
90*dbc189b2SAndreas Gohr     * Each part is either a non-asian string, or a single asian word
91*dbc189b2SAndreas Gohr     *
92*dbc189b2SAndreas Gohr     * @param string $term
93*dbc189b2SAndreas Gohr     * @return string[]
94*dbc189b2SAndreas Gohr     */
95*dbc189b2SAndreas Gohr    public static function splitAsianWords($term)
96*dbc189b2SAndreas Gohr    {
97*dbc189b2SAndreas Gohr        return preg_split('/(' . self::REGEXP . '+)/u', $term, -1, PREG_SPLIT_DELIM_CAPTURE | PREG_SPLIT_NO_EMPTY);
98*dbc189b2SAndreas Gohr    }
99*dbc189b2SAndreas Gohr}
100