|
1
|
|
|
<?php |
|
2
|
|
|
|
|
3
|
|
|
declare(strict_types=1); |
|
4
|
|
|
|
|
5
|
|
|
/** |
|
6
|
|
|
* This file is part of Scout Extended. |
|
7
|
|
|
* |
|
8
|
|
|
* (c) Algolia Team <[email protected]> |
|
9
|
|
|
* |
|
10
|
|
|
* For the full copyright and license information, please view the LICENSE |
|
11
|
|
|
* file that was distributed with this source code. |
|
12
|
|
|
*/ |
|
13
|
|
|
|
|
14
|
|
|
namespace Algolia\ScoutExtended\Splitters\HtmlSplitterComponent; |
|
15
|
|
|
|
|
16
|
|
|
final class Queue |
|
17
|
|
|
{ |
|
18
|
|
|
/** |
|
19
|
|
|
* @var array |
|
20
|
|
|
*/ |
|
21
|
|
|
protected $queue = []; |
|
22
|
|
|
protected $cloneQueue = []; |
|
23
|
|
|
|
|
24
|
|
|
/** |
|
25
|
|
|
* The list of html tags. |
|
26
|
|
|
* |
|
27
|
|
|
* @var string[] |
|
28
|
|
|
*/ |
|
29
|
|
|
protected $tags = [ |
|
30
|
|
|
'h1', |
|
31
|
|
|
'h2', |
|
32
|
|
|
'h3', |
|
33
|
|
|
'h4', |
|
34
|
|
|
'h5', |
|
35
|
|
|
'h6', |
|
36
|
|
|
'p', |
|
37
|
|
|
]; |
|
38
|
|
|
|
|
39
|
|
|
/** |
|
40
|
|
|
* String for key check purpose. |
|
41
|
|
|
*/ |
|
42
|
|
|
private const IMPORTANCE = 'importance'; |
|
43
|
|
|
|
|
44
|
|
|
/** |
|
45
|
|
|
* String for exception purpose. |
|
46
|
|
|
*/ |
|
47
|
|
|
private const PARAGRAPH = 'p'; |
|
48
|
|
|
|
|
49
|
|
|
/** |
|
50
|
|
|
* @return array |
|
51
|
|
|
*/ |
|
52
|
|
|
public function getCloneQueue(): array |
|
53
|
|
|
{ |
|
54
|
|
|
return $this->cloneQueue; |
|
55
|
|
|
} |
|
56
|
|
|
|
|
57
|
|
|
/** |
|
58
|
|
|
* Add object to queue. |
|
59
|
|
|
* |
|
60
|
|
|
* @param ObjectQueue $object |
|
61
|
|
|
*/ |
|
62
|
5 |
|
public function addObjectQueue(ObjectQueue $object): void |
|
63
|
|
|
{ |
|
64
|
5 |
|
if ($this->lengthQueue() === 0) { |
|
65
|
5 |
|
$this->queue[] = $object; |
|
66
|
5 |
|
$this->cloneQueue(); |
|
67
|
5 |
|
} elseif ($this->findWeight($object) > $this->findWeight(end($this->queue))) { |
|
68
|
3 |
|
$this->queue[] = $object; |
|
69
|
3 |
|
$this->cloneQueue(); |
|
70
|
|
|
} else { |
|
71
|
4 |
|
array_pop($this->queue); |
|
72
|
4 |
|
$this->addObjectQueue($object); |
|
73
|
|
|
} |
|
74
|
5 |
|
} |
|
75
|
|
|
|
|
76
|
|
|
/** |
|
77
|
|
|
* Clean Records to have a correct format. |
|
78
|
|
|
* |
|
79
|
|
|
* @return array |
|
80
|
|
|
*/ |
|
81
|
5 |
|
public function sanitizeQueue(): array |
|
82
|
|
|
{ |
|
83
|
5 |
|
$records = []; |
|
84
|
5 |
|
foreach ($this->cloneQueue as $queue) { |
|
85
|
5 |
|
foreach ($queue as $object) { |
|
86
|
5 |
|
if ($object instanceof ObjectQueue) { |
|
87
|
5 |
|
$record[$object->getTag()] = $object->getTagContent(); |
|
88
|
|
|
} else { |
|
89
|
5 |
|
$record[self::IMPORTANCE] = $object; |
|
90
|
5 |
|
$records[] = $record; |
|
91
|
5 |
|
$record = []; |
|
92
|
|
|
} |
|
93
|
|
|
} |
|
94
|
|
|
} |
|
95
|
|
|
|
|
96
|
5 |
|
return $records; |
|
97
|
|
|
} |
|
98
|
|
|
|
|
99
|
|
|
/** |
|
100
|
|
|
* Importance need to be add after to avoid polluted queue. |
|
101
|
|
|
* |
|
102
|
|
|
* @return void |
|
103
|
|
|
*/ |
|
104
|
5 |
|
private function cloneQueue(): void |
|
105
|
|
|
{ |
|
106
|
5 |
|
$this->cloneQueue[] = $this->queue; |
|
107
|
5 |
|
$this->cloneQueue[] = [self::IMPORTANCE => $this->importanceWeight(end($this->queue))]; |
|
108
|
5 |
|
} |
|
109
|
|
|
|
|
110
|
|
|
/** |
|
111
|
|
|
* Importance formula. |
|
112
|
|
|
* Give integer from tags ranking. |
|
113
|
|
|
* |
|
114
|
|
|
* @param ObjectQueue $objectQueue |
|
115
|
|
|
* |
|
116
|
|
|
* @return int |
|
117
|
|
|
*/ |
|
118
|
5 |
|
private function importanceWeight(ObjectQueue $objectQueue): int |
|
119
|
|
|
{ |
|
120
|
5 |
|
if ($objectQueue->getTag() === self::PARAGRAPH) { |
|
121
|
3 |
|
$object = prev($this->queue); |
|
122
|
3 |
|
if (empty(end($this->queue)) || $this->lengthQueue() === 1) { |
|
123
|
|
|
return 0; |
|
124
|
|
|
} |
|
125
|
|
|
|
|
126
|
3 |
|
return (count($this->tags) - 1) + $this->findWeight($object); |
|
127
|
|
|
} |
|
128
|
|
|
|
|
129
|
5 |
|
return $this->findWeight($objectQueue); |
|
130
|
|
|
} |
|
131
|
|
|
|
|
132
|
|
|
/** |
|
133
|
|
|
* Find weight of current nodes. |
|
134
|
|
|
* |
|
135
|
|
|
* @param ObjectQueue $object |
|
136
|
|
|
* |
|
137
|
|
|
* @return int |
|
138
|
|
|
*/ |
|
139
|
5 |
|
private function findWeight(ObjectQueue $object): int |
|
140
|
|
|
{ |
|
141
|
5 |
|
return (int) array_search($object->getTag(), $this->tags, true); |
|
142
|
|
|
} |
|
143
|
|
|
|
|
144
|
|
|
/** |
|
145
|
|
|
* Give the length of the queue. |
|
146
|
|
|
* |
|
147
|
|
|
* @return int |
|
148
|
|
|
*/ |
|
149
|
5 |
|
private function lengthQueue(): int |
|
150
|
|
|
{ |
|
151
|
5 |
|
return count($this->queue); |
|
152
|
|
|
} |
|
153
|
|
|
} |
|
154
|
|
|
|