|
1
|
|
|
<?php |
|
2
|
|
|
/* |
|
3
|
|
|
* This file is part of dispositif/wikibot application (@github) |
|
4
|
|
|
* 2019-2023 © Philippe M./Irønie <[email protected]> |
|
5
|
|
|
* For the full copyright and MIT license information, view the license file. |
|
6
|
|
|
*/ |
|
7
|
|
|
|
|
8
|
|
|
declare(strict_types=1); |
|
9
|
|
|
|
|
10
|
|
|
namespace App\Domain\Publisher; |
|
11
|
|
|
|
|
12
|
|
|
use App\Domain\Utils\TextUtil; |
|
13
|
|
|
|
|
14
|
|
|
class SeoSanitizer |
|
15
|
|
|
{ |
|
16
|
|
|
private const MAX_LENGTH_FIRST_SEG_ALLOWING_SECOND_SEG = 30; |
|
17
|
|
|
private const REBUILD_SEPARATOR = ' - '; |
|
18
|
|
|
|
|
19
|
|
|
/** |
|
20
|
|
|
* Naive SEO sanitization of web page title. |
|
21
|
|
|
* pretty domain name as "google.com" or "google.co.uk" |
|
22
|
|
|
*/ |
|
23
|
|
|
public function cleanSEOTitle(string $prettyDomainName, ?string $title): ?string |
|
24
|
|
|
{ |
|
25
|
|
|
if (empty(trim($title))) { |
|
|
|
|
|
|
26
|
|
|
return null; |
|
27
|
|
|
} |
|
28
|
|
|
$title = str_replace(['–', '—', '\\'], ['-', '-', '/'], $title); // replace em dash with hyphen |
|
29
|
|
|
|
|
30
|
|
|
$seoSegments = $this->extractSEOSegments($title); |
|
31
|
|
|
// No SEO segmentation found |
|
32
|
|
|
if (count($seoSegments) < 2) { |
|
33
|
|
|
return $title; |
|
34
|
|
|
} |
|
35
|
|
|
$seoSegmentsFiltered = $this->deleteSegmentsContainingSitename($prettyDomainName, $seoSegments); |
|
36
|
|
|
if (count($seoSegmentsFiltered) === 0) { |
|
37
|
|
|
return trim($seoSegments[0]); |
|
38
|
|
|
} |
|
39
|
|
|
|
|
40
|
|
|
return $this->buildNewTitle($seoSegmentsFiltered); |
|
41
|
|
|
} |
|
42
|
|
|
|
|
43
|
|
|
/** |
|
44
|
|
|
* @param string[] $titleSegments |
|
45
|
|
|
*/ |
|
46
|
|
|
protected function buildNewTitle(array $titleSegments): string |
|
47
|
|
|
{ |
|
48
|
|
|
// if only one segment or first segment is long enough, return it |
|
49
|
|
|
if ( |
|
50
|
|
|
count($titleSegments) === 1 |
|
51
|
|
|
|| mb_strlen($titleSegments[0]) >= self::MAX_LENGTH_FIRST_SEG_ALLOWING_SECOND_SEG |
|
52
|
|
|
) { |
|
53
|
|
|
return trim($titleSegments[0]); |
|
54
|
|
|
} |
|
55
|
|
|
|
|
56
|
|
|
// rebuild title but keep only the first 2 segments |
|
57
|
|
|
return trim($titleSegments[0]) . self::REBUILD_SEPARATOR . trim($titleSegments[1]); |
|
58
|
|
|
} |
|
59
|
|
|
|
|
60
|
|
|
/** |
|
61
|
|
|
* @return string[] |
|
62
|
|
|
*/ |
|
63
|
|
|
private function extractSEOSegments(string $title): array |
|
64
|
|
|
{ |
|
65
|
|
|
$seoSeparator = $this->getSEOSeparator($title); |
|
66
|
|
|
|
|
67
|
|
|
return (null === $seoSeparator) ? [$title] : explode($seoSeparator, $title); |
|
68
|
|
|
} |
|
69
|
|
|
|
|
70
|
|
|
private function getSEOSeparator(string $title): ?string |
|
71
|
|
|
{ |
|
72
|
|
|
// order is important. '-' before '/' ? see date, etc |
|
73
|
|
|
if (strpos($title, ' | ') !== false) { |
|
74
|
|
|
return ' | '; |
|
75
|
|
|
} |
|
76
|
|
|
if (strpos($title, ' / ') !== false) { |
|
77
|
|
|
return ' / '; |
|
78
|
|
|
} |
|
79
|
|
|
if (strpos($title, ' - ') !== false) { |
|
80
|
|
|
return ' - '; |
|
81
|
|
|
} |
|
82
|
|
|
if (strpos($title, ' : ') !== false) { |
|
83
|
|
|
return ' : '; |
|
84
|
|
|
} |
|
85
|
|
|
|
|
86
|
|
|
return null; |
|
87
|
|
|
} |
|
88
|
|
|
|
|
89
|
|
|
/** |
|
90
|
|
|
* Remove SEO segments as containing same words as the website domain name. |
|
91
|
|
|
* |
|
92
|
|
|
* @param string[] $seoSegments |
|
93
|
|
|
* |
|
94
|
|
|
* @return string[] |
|
95
|
|
|
*/ |
|
96
|
|
|
private function deleteSegmentsContainingSitename(string $prettyDomainName, array $seoSegments): array |
|
97
|
|
|
{ |
|
98
|
|
|
$siteName = TextUtil::stripPunctuation($this->extractSiteName($prettyDomainName)); |
|
99
|
|
|
$siteName = str_replace(['.', '-', ' '], '', $siteName); |
|
100
|
|
|
$prettyDomainName = TextUtil::stripPunctuation($prettyDomainName); |
|
101
|
|
|
$prettyDomainName = str_replace(['.', '-', ' '], '', $prettyDomainName); |
|
102
|
|
|
|
|
103
|
|
|
return array_values(array_filter( |
|
104
|
|
|
$seoSegments, |
|
105
|
|
|
function ($segment) use ($prettyDomainName, $siteName) { |
|
106
|
|
|
$strippedSegment = mb_strtolower(TextUtil::stripPunctuation(TextUtil::stripAccents($segment))); |
|
107
|
|
|
$strippedSegment = str_replace(['.', '-', ' '], '', $strippedSegment); |
|
108
|
|
|
|
|
109
|
|
|
return !empty(trim($segment)) |
|
110
|
|
|
&& false === strpos($strippedSegment, $prettyDomainName) |
|
111
|
|
|
&& false === strpos($strippedSegment, $siteName); |
|
112
|
|
|
} |
|
113
|
|
|
)); |
|
114
|
|
|
} |
|
115
|
|
|
|
|
116
|
|
|
/** |
|
117
|
|
|
* Get site name from pretty domain name. |
|
118
|
|
|
* Ex: "google.com" => "google" |
|
119
|
|
|
* Ex: "my-news.co.uk" => "my-news" |
|
120
|
|
|
*/ |
|
121
|
|
|
private function extractSiteName(string $prettyDomainName): string |
|
122
|
|
|
{ |
|
123
|
|
|
// strip string after last dot |
|
124
|
|
|
$siteName = preg_replace('/\.[^.]*$/', '', $prettyDomainName); |
|
125
|
|
|
|
|
126
|
|
|
// strip string if only 2 chars after dot |
|
127
|
|
|
return preg_replace('/\.[^.]{2}$/', '', $siteName); |
|
128
|
|
|
} |
|
129
|
|
|
} |
Dispositif /
Wikibot
Loading history...