|
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
|
|
|
/** |
|
17
|
|
|
* Naive SEO sanitization of web page title. |
|
18
|
|
|
* pretty domain name as "google.com" or "google.co.uk" |
|
19
|
|
|
*/ |
|
20
|
|
|
public function cleanSEOTitle(string $prettyDomainName, ?string $title): ?string |
|
21
|
|
|
{ |
|
22
|
|
|
if (empty(trim($title))) { |
|
|
|
|
|
|
23
|
|
|
return null; |
|
24
|
|
|
} |
|
25
|
|
|
$title = str_replace(['–','—','\\'],['-','-','/'], $title); // replace em dash with hyphen |
|
26
|
|
|
|
|
27
|
|
|
$seoSegments = $this->extractSEOSegments($title); |
|
28
|
|
|
// No SEO segmentation found |
|
29
|
|
|
if (count($seoSegments) < 2) { |
|
30
|
|
|
return $title; |
|
31
|
|
|
} |
|
32
|
|
|
|
|
33
|
|
|
// remove seo segments containing same words as the website domain name |
|
34
|
|
|
$seoSegmentsFiltered = array_values(array_filter($seoSegments, function ($segment) use ($prettyDomainName) { |
|
35
|
|
|
// strip string after last dot in prettyDomainName : blabla.com => blabla |
|
36
|
|
|
$domainMain = preg_replace('/\.[^.]*$/', '', $prettyDomainName); |
|
37
|
|
|
// strip string if only 2 chars after dot : like in 'blabla.co.uk' |
|
38
|
|
|
$domainMain = preg_replace('/\.[^.]{2}$/', '', $domainMain); |
|
39
|
|
|
|
|
40
|
|
|
$strippedSegment = str_replace([' ', '-'], '', mb_strtolower(TextUtil::stripPunctuation(TextUtil::stripAccents($segment)))); |
|
41
|
|
|
|
|
42
|
|
|
return !empty(trim($segment)) |
|
43
|
|
|
&& false === strpos($strippedSegment, str_replace(['.', '-'], '', $prettyDomainName)) |
|
44
|
|
|
&& false === strpos($strippedSegment, str_replace(['.', '-'], '', $domainMain)); |
|
45
|
|
|
})); |
|
46
|
|
|
if (count($seoSegmentsFiltered) === 0) { |
|
47
|
|
|
return trim($seoSegments[0]); // before filtering |
|
48
|
|
|
} |
|
49
|
|
|
|
|
50
|
|
|
// if only one segment or first segment is long enough, return it |
|
51
|
|
|
if (count($seoSegmentsFiltered) === 1 || mb_strlen($seoSegmentsFiltered[0]) >= 25) { |
|
52
|
|
|
return trim($seoSegmentsFiltered[0]); |
|
53
|
|
|
} |
|
54
|
|
|
|
|
55
|
|
|
// rebuild bestTitle but keep only the first 2 SEO segments |
|
56
|
|
|
return trim($seoSegmentsFiltered[0]) . ' - ' . trim($seoSegmentsFiltered[1]); |
|
57
|
|
|
} |
|
58
|
|
|
|
|
59
|
|
|
private function extractSEOSegments(string $title): array |
|
60
|
|
|
{ |
|
61
|
|
|
$seoSeparator = $this->getSEOSeparator($title); |
|
62
|
|
|
if (null === $seoSeparator) { |
|
63
|
|
|
return [$title]; |
|
64
|
|
|
} |
|
65
|
|
|
|
|
66
|
|
|
return explode($seoSeparator, $title); |
|
67
|
|
|
} |
|
68
|
|
|
|
|
69
|
|
|
private function getSEOSeparator(string $title): ?string |
|
70
|
|
|
{ |
|
71
|
|
|
$title = str_replace(' — ', ' - ', $title); |
|
72
|
|
|
if (strpos($title, ' | ') !== false) { |
|
73
|
|
|
return ' | '; |
|
74
|
|
|
} |
|
75
|
|
|
if (strpos($title, ' - ') !== false) { |
|
76
|
|
|
return ' - '; |
|
77
|
|
|
} |
|
78
|
|
|
if (strpos($title, ' / ') !== false) { |
|
79
|
|
|
return ' / '; |
|
80
|
|
|
} |
|
81
|
|
|
if (strpos($title, ' : ') !== false) { |
|
82
|
|
|
return ' : '; |
|
83
|
|
|
} |
|
84
|
|
|
|
|
85
|
|
|
return null; |
|
86
|
|
|
} |
|
87
|
|
|
} |