|
1
|
|
|
<?php |
|
2
|
|
|
|
|
3
|
|
|
/** |
|
4
|
|
|
* File holding the Extensions\Lingo\LingoTree class |
|
5
|
|
|
* |
|
6
|
|
|
* This file is part of the MediaWiki extension Lingo. |
|
7
|
|
|
* |
|
8
|
|
|
* @copyright 2011 - 2016, Stephan Gambke |
|
9
|
|
|
* @license GNU General Public License, version 2 (or any later version) |
|
10
|
|
|
* |
|
11
|
|
|
* The Lingo extension is free software: you can redistribute it and/or modify |
|
12
|
|
|
* it under the terms of the GNU General Public License as published by the Free |
|
13
|
|
|
* Software Foundation; either version 2 of the License, or (at your option) any |
|
14
|
|
|
* later version. |
|
15
|
|
|
* |
|
16
|
|
|
* The Lingo extension is distributed in the hope that it will be useful, but |
|
17
|
|
|
* WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or |
|
18
|
|
|
* FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more |
|
19
|
|
|
* details. |
|
20
|
|
|
* |
|
21
|
|
|
* You should have received a copy of the GNU General Public License along |
|
22
|
|
|
* with this program. If not, see <http://www.gnu.org/licenses/>. |
|
23
|
|
|
* |
|
24
|
|
|
* @author Stephan Gambke |
|
25
|
|
|
* |
|
26
|
|
|
* @file |
|
27
|
|
|
* @ingroup Lingo |
|
28
|
|
|
*/ |
|
29
|
|
|
namespace Extensions\Lingo; |
|
30
|
|
|
|
|
31
|
|
|
/** |
|
32
|
|
|
* The Extensions\Lingo\LingoTree class. |
|
33
|
|
|
* |
|
34
|
|
|
* Vocabulary: |
|
35
|
|
|
* Term - The term as a normal string |
|
36
|
|
|
* Definition - Its definition (any object) |
|
37
|
|
|
* Element - An element (leaf) in the glossary tree |
|
38
|
|
|
* Path - The path in the tree to the leaf representing a term |
|
39
|
|
|
* |
|
40
|
|
|
* The glossary is organized as a tree (nested arrays) where the path to the |
|
41
|
|
|
* definition of a term is the lexemes of the term followed by -1 as the end |
|
42
|
|
|
* marker. |
|
43
|
|
|
* |
|
44
|
|
|
* Example: |
|
45
|
|
|
* The path to the definition of the term "foo bar baz" would be |
|
46
|
|
|
* 'foo'.' '.'bar'.' '.'baz'.'-1'. It could thus be accessed as |
|
47
|
|
|
* $mTree['foo'][' ']['bar'][' ']['baz'][-1] |
|
48
|
|
|
* |
|
49
|
|
|
* @ingroup Lingo |
|
50
|
|
|
*/ |
|
51
|
|
|
class LingoTree { |
|
52
|
|
|
|
|
53
|
|
|
private $mTree = array(); |
|
54
|
|
|
private $mList = array(); |
|
55
|
|
|
private $mMinLength = 1000; |
|
56
|
|
|
|
|
57
|
|
|
/** |
|
58
|
|
|
* Adds a string to the Lingo Tree |
|
59
|
|
|
* @param String $term |
|
60
|
|
|
*/ |
|
61
|
|
|
public function addTerm( &$term, $definition ) { |
|
62
|
|
|
if ( !$term ) { |
|
63
|
|
|
return; |
|
64
|
|
|
} |
|
65
|
|
|
|
|
66
|
|
|
if ( isset( $this->mList[ $term ] ) ) { // term exists, store 2nd definition |
|
67
|
|
|
$this->mList[ $term ]->addDefinition( $definition ); |
|
68
|
|
|
} else { |
|
69
|
|
|
|
|
70
|
|
|
$matches = array(); |
|
71
|
|
|
preg_match_all( LingoParser::$regex, $term, $matches ); |
|
72
|
|
|
|
|
73
|
|
|
$elt = $this->addElement( $matches[ 0 ], $term, $definition ); |
|
74
|
|
|
$this->mList[ $term ] = &$elt[ -1 ]; |
|
75
|
|
|
|
|
76
|
|
|
$this->mMinLength = min( array( $this->mMinLength, strlen( $term ) ) ); |
|
77
|
|
|
} |
|
78
|
|
|
} |
|
79
|
|
|
|
|
80
|
|
|
/** |
|
81
|
|
|
* Adds an element to the Lingo Tree |
|
82
|
|
|
* |
|
83
|
|
|
* @param array $path An array containing the constituing lexemes of the term |
|
84
|
|
|
* @param String $term |
|
85
|
|
|
* @param String $definition |
|
86
|
|
|
* @return Array the tree node the element was stored in |
|
87
|
|
|
*/ |
|
88
|
|
|
protected function &addElement( Array &$path, &$term, &$definition ) { |
|
89
|
|
|
|
|
90
|
|
|
$tree = &$this->mTree; |
|
91
|
|
|
|
|
92
|
|
|
// end of path, store description; end of recursion |
|
93
|
|
|
while ( ( $step = array_shift( $path ) ) !== null ) { |
|
94
|
|
|
|
|
95
|
|
|
if ( !isset( $tree[ $step ] ) ) { |
|
96
|
|
|
$tree[ $step ] = array(); |
|
97
|
|
|
} |
|
98
|
|
|
|
|
99
|
|
|
$tree = &$tree[ $step ]; |
|
100
|
|
|
} |
|
101
|
|
|
|
|
102
|
|
|
if ( isset( $tree[ -1 ] ) ) { |
|
103
|
|
|
$tree[ -1 ]->addDefinition( $definition ); |
|
104
|
|
|
} else { |
|
105
|
|
|
$tree[ -1 ] = new LingoElement( $term, $definition ); |
|
106
|
|
|
} |
|
107
|
|
|
|
|
108
|
|
|
return $tree; |
|
109
|
|
|
} |
|
110
|
|
|
|
|
111
|
|
|
public function getMinTermLength() { |
|
112
|
|
|
return $this->mMinLength; |
|
113
|
|
|
} |
|
114
|
|
|
|
|
115
|
|
|
public function getTermList() { |
|
116
|
|
|
return $this->mList; |
|
117
|
|
|
} |
|
118
|
|
|
|
|
119
|
|
|
public function findNextTerm( &$lexemes, $index, $countLexemes ) { |
|
120
|
|
|
|
|
121
|
|
|
$start = $lastindex = $index; |
|
122
|
|
|
$definition = null; |
|
123
|
|
|
|
|
124
|
|
|
// skip until the start of a term is found |
|
125
|
|
|
while ( $index < $countLexemes && !$definition ) { |
|
126
|
|
|
$currLex = &$lexemes[ $index ][ 0 ]; |
|
127
|
|
|
|
|
128
|
|
|
// Did we find the start of a term? |
|
129
|
|
|
if ( array_key_exists( $currLex, $this->mTree ) ) { |
|
130
|
|
|
list( $lastindex, $definition ) = $this->findNextTermNoSkip( $this->mTree[ $currLex ], $lexemes, $index, $countLexemes ); |
|
131
|
|
|
} |
|
132
|
|
|
|
|
133
|
|
|
// this will increase the index even if we found something; |
|
134
|
|
|
// will be corrected after the loop |
|
135
|
|
|
$index++; |
|
136
|
|
|
} |
|
137
|
|
|
|
|
138
|
|
|
if ( $definition ) { |
|
139
|
|
|
return array( $index - $start - 1, $lastindex - $index + 2, $definition ); |
|
140
|
|
|
} else { |
|
141
|
|
|
return array( $index - $start, 0, null ); |
|
142
|
|
|
} |
|
143
|
|
|
} |
|
144
|
|
|
|
|
145
|
|
|
public function findNextTermNoSkip( Array &$tree, &$lexemes, $index, $countLexemes ) { |
|
146
|
|
|
|
|
147
|
|
|
if ( $index + 1 < $countLexemes && array_key_exists( $currLex = $lexemes[ $index + 1 ][ 0 ], $tree ) ) { |
|
148
|
|
|
$ret = $this->findNextTermNoSkip( $tree[ $currLex ], $lexemes, $index + 1, $countLexemes ); |
|
149
|
|
|
} else { |
|
150
|
|
|
$ret = array( $index, &$tree[ -1 ] ); |
|
151
|
|
|
} |
|
152
|
|
|
|
|
153
|
|
|
return $ret; |
|
154
|
|
|
} |
|
155
|
|
|
|
|
156
|
|
|
} |
|
157
|
|
|
|
wikimedia /
mediawiki-extensions-Lingo
