1
|
|
|
<?php |
2
|
|
|
|
3
|
|
|
namespace CHMLib\TOCIndex; |
4
|
|
|
|
5
|
|
|
use CHMLib\CHM; |
6
|
|
|
use CHMLib\Map; |
7
|
|
|
use DOMDocument; |
8
|
|
|
use DOMElement; |
9
|
|
|
use DOMXpath; |
10
|
|
|
use Exception; |
11
|
|
|
|
12
|
|
|
/** |
13
|
|
|
* A list of items in the TOC or in the Index of an CHM file. |
14
|
|
|
*/ |
15
|
|
|
class Tree |
16
|
|
|
{ |
17
|
|
|
/** |
18
|
|
|
* List of Item instances children of this tree. |
19
|
|
|
* |
20
|
|
|
* @var Item[] |
21
|
|
|
*/ |
22
|
|
|
protected $items; |
23
|
|
|
|
24
|
|
|
/** |
25
|
|
|
* Initializes the instance. |
26
|
|
|
*/ |
27
|
|
|
public function __construct() |
28
|
|
|
{ |
29
|
|
|
$this->items = array(); |
30
|
|
|
} |
31
|
|
|
|
32
|
|
|
/** |
33
|
|
|
* Get the items contained in this tree. |
34
|
|
|
* |
35
|
|
|
* @return Item[] |
36
|
|
|
*/ |
37
|
|
|
public function getItems() |
38
|
|
|
{ |
39
|
|
|
return $this->items; |
40
|
|
|
} |
41
|
|
|
|
42
|
|
|
/** |
43
|
|
|
* Append to the children of this tree the children of another tree. |
44
|
|
|
* |
45
|
|
|
* @param Tree $tree |
46
|
|
|
*/ |
47
|
|
|
public function mergeItems(Tree $tree) |
48
|
|
|
{ |
49
|
|
|
$this->items = array_merge($this->items, $tree->items); |
50
|
|
|
} |
51
|
|
|
|
52
|
|
|
/** |
53
|
|
|
* Resolve the items contained in other CHM files. |
54
|
|
|
* |
55
|
|
|
* @param Map $map |
56
|
|
|
* |
57
|
|
|
* @throws Exception Throw an Exception in case of errors. |
58
|
|
|
*/ |
59
|
|
|
public function resolve(Map $map) |
60
|
|
|
{ |
61
|
|
|
$result = array(); |
62
|
|
|
foreach ($this->items as $item) { |
63
|
|
|
$merge = $item->getMerge(); |
64
|
|
|
if ($merge === null) { |
65
|
|
|
$result[] = $item; |
66
|
|
|
} else { |
67
|
|
|
$result = array_merge($result, $item->resolve($map)); |
68
|
|
|
} |
69
|
|
|
} |
70
|
|
|
|
71
|
|
|
$this->items = $result; |
72
|
|
|
} |
73
|
|
|
|
74
|
|
|
/** |
75
|
|
|
* Create a new instance starting from the whole TOC/Index source 'HTML'. |
76
|
|
|
* |
77
|
|
|
* @param CHM $chm The parent CHM instance. |
78
|
|
|
* @param string $data The contents of the .hhc/.hhk file. |
79
|
|
|
* |
80
|
|
|
* @throws Exception Throw an Exception in case of errors. |
81
|
|
|
* |
82
|
|
|
* @return static |
83
|
|
|
*/ |
84
|
|
|
public static function fromString(CHM $chm, $data) |
85
|
|
|
{ |
86
|
|
|
if (!class_exists('DOMDocument', false) || !class_exists('DOMXpath', false)) { |
87
|
|
|
throw new Exception('Missing PHP extension: php-xml'); |
88
|
|
|
} |
89
|
|
|
$result = new static(); |
90
|
|
|
$data = trim((string) $data); |
91
|
|
|
if (stripos($data, '<object') !== false) { |
92
|
|
|
$doc = new DOMDocument(); |
93
|
|
|
$charset = 'UTF-8'; |
94
|
|
|
if (preg_match('%^<\?xml\s+encoding\s*=\s*"([^"]+)"%i', $data, $m)) { |
95
|
|
|
$charset = $m[1]; |
|
|
|
|
96
|
|
|
} else { |
97
|
|
|
if (preg_match('%<meta\s+http-equiv\s*=\s*"Content-Type"\s+content\s*=\s*"text/html;\s*charset=([^"]+)">%i', $data, $m)) { |
98
|
|
|
$charset = $m[1]; |
99
|
|
|
} |
100
|
|
|
$data = '<?xml encoding="'.$charset.'">'.$data; |
101
|
|
|
} |
102
|
|
|
// LI elements are very often malformed/misplaced: let's remove them |
103
|
|
|
$data = preg_replace('$</?li((\\s+[^>]*)|\\s*)>$i', '', $data); |
104
|
|
|
if (@$doc->loadHTML($data) !== true) { |
105
|
|
|
throw new Exception('Failed to parse the .hhc/.hhk file contents'); |
106
|
|
|
} |
107
|
|
|
$result->parseParentElement($chm, $doc->documentElement, 0); |
108
|
|
|
} |
109
|
|
|
|
110
|
|
|
return $result; |
111
|
|
|
} |
112
|
|
|
|
113
|
|
|
/** |
114
|
|
|
* Depth of the found child items. |
115
|
|
|
* |
116
|
|
|
* @var int |
117
|
|
|
*/ |
118
|
|
|
protected $depth; |
119
|
|
|
|
120
|
|
|
/** |
121
|
|
|
* Parse a DOMElement and read the items/sub trees. |
122
|
|
|
* |
123
|
|
|
* @param CHM $chm |
124
|
|
|
* @param DOMElement $parentElement |
125
|
|
|
* @param int $depth |
126
|
|
|
*/ |
127
|
|
|
protected function parseParentElement(CHM $chm, DOMElement $parentElement, $depth) |
128
|
|
|
{ |
129
|
|
|
foreach ($parentElement->childNodes as $node) { |
130
|
|
|
if ($node->nodeType === XML_ELEMENT_NODE) { |
131
|
|
|
switch (strtolower($node->tagName)) { |
132
|
|
|
case 'object': |
133
|
|
|
if (strtolower($node->getAttribute('type')) === 'text/sitemap') { |
134
|
|
|
$this->depth = $depth; |
135
|
|
|
$this->items[] = new Item($chm, $node); |
136
|
|
|
} |
137
|
|
|
break; |
138
|
|
|
case 'ul': |
139
|
|
|
case 'ol': |
140
|
|
|
$n = count($this->items); |
141
|
|
|
if ($n > 0 && $depth >= $this->depth) { |
142
|
|
|
$this->items[$n - 1]->getChildren()->parseParentElement($chm, $node, $depth + 1); |
143
|
|
|
} else { |
144
|
|
|
$this->parseParentElement($chm, $node, $depth + 1); |
145
|
|
|
} |
146
|
|
|
break; |
147
|
|
|
default: |
148
|
|
|
$this->parseParentElement($chm, $node, $depth + 1); |
149
|
|
|
break; |
150
|
|
|
} |
151
|
|
|
} |
152
|
|
|
} |
153
|
|
|
} |
154
|
|
|
} |
155
|
|
|
|
This check looks for variable assignements that are either overwritten by other assignments or where the variable is not used subsequently.
Both the
$myVar
assignment in line 1 and the$higher
assignment in line 2 are dead. The first because$myVar
is never used and the second because$higher
is always overwritten for every possible time line.