|
1
|
|
|
<?php |
|
2
|
|
|
|
|
3
|
|
|
namespace PhpTek\Exodus\Crawl; |
|
4
|
|
|
|
|
5
|
|
|
use PHPCrawl\PHPCrawler; |
|
6
|
|
|
use PHPCrawl\PHPCrawlerDocumentInfo; |
|
7
|
|
|
use PHPCrawl\PHPCrawlerURLDescriptor; |
|
8
|
|
|
use SilverStripe\Core\Injector\Injectable; |
|
9
|
|
|
use SilverStripe\Core\Config\Configurable; |
|
10
|
|
|
use PhpTek\Exodus\Tool\StaticSiteUrlList; |
|
11
|
|
|
use PhpTek\Exodus\Tool\StaticSiteUtils; |
|
12
|
|
|
|
|
13
|
|
|
/** |
|
14
|
|
|
* Extends PHPCrawler essentially to override its handleDocumentInfo() method. |
|
15
|
|
|
* |
|
16
|
|
|
* @see {@link PHPCrawler} |
|
17
|
|
|
*/ |
|
18
|
|
|
class StaticSiteCrawler extends PHPCrawler |
|
19
|
|
|
{ |
|
20
|
|
|
use Injectable; |
|
21
|
|
|
use Configurable; |
|
22
|
|
|
|
|
23
|
|
|
/** |
|
24
|
|
|
* |
|
25
|
|
|
* @var StaticSiteUrlList |
|
26
|
|
|
*/ |
|
27
|
|
|
protected $urlList; |
|
28
|
|
|
|
|
29
|
|
|
/** |
|
30
|
|
|
* |
|
31
|
|
|
* @var boolean |
|
32
|
|
|
*/ |
|
33
|
|
|
protected $verbose = false; |
|
34
|
|
|
|
|
35
|
|
|
/* |
|
36
|
|
|
* Holds the StaticSiteUtils object on construct |
|
37
|
|
|
* |
|
38
|
|
|
* @var StaticSiteUtils |
|
39
|
|
|
*/ |
|
40
|
|
|
protected $utils; |
|
41
|
|
|
|
|
42
|
|
|
/** |
|
43
|
|
|
* Set this by using the yml config system |
|
44
|
|
|
* |
|
45
|
|
|
* Example: |
|
46
|
|
|
* <code> |
|
47
|
|
|
* StaticSiteContentExtractor: |
|
48
|
|
|
* log_file: ../logs/crawler-log.txt |
|
49
|
|
|
* </code> |
|
50
|
|
|
* |
|
51
|
|
|
* @var string |
|
52
|
|
|
*/ |
|
53
|
|
|
private static $log_file = null; |
|
|
|
|
|
|
54
|
|
|
|
|
55
|
|
|
/** |
|
56
|
|
|
* |
|
57
|
|
|
* @param StaticSiteUrlList $urlList |
|
58
|
|
|
* @param number $limit |
|
59
|
|
|
* @param boolean $verbose |
|
60
|
|
|
* @return void |
|
61
|
|
|
*/ |
|
62
|
|
|
public function __construct(StaticSiteUrlList $urlList, $limit = false, $verbose = false) |
|
63
|
|
|
{ |
|
64
|
|
|
parent::__construct(); |
|
65
|
|
|
|
|
66
|
|
|
$this->urlList = $urlList; |
|
67
|
|
|
$this->verbose = $verbose; |
|
68
|
|
|
$this->utils = singleton(StaticSiteUtils::class); |
|
69
|
|
|
} |
|
70
|
|
|
|
|
71
|
|
|
/** |
|
72
|
|
|
* After checking raw status codes out of PHPCrawler we continue to save each URL to our cache file. |
|
73
|
|
|
* |
|
74
|
|
|
* $PageInfo gives us: |
|
75
|
|
|
* |
|
76
|
|
|
* $PageInfo->url |
|
77
|
|
|
* $PageInfo->http_status_code |
|
78
|
|
|
* $PageInfo->links_found_url_descriptors |
|
79
|
|
|
* |
|
80
|
|
|
* @param PHPCrawlerDocumentInfo $PageInfo |
|
81
|
|
|
* @return int |
|
82
|
|
|
* @todo Can we make use of PHPCrawlerDocumentInfo#error_occured instead of manually checking server codes?? |
|
83
|
|
|
* @todo The comments below state that badly formatted URLs never make it to our caching logic. Wrong! |
|
84
|
|
|
* - Pass the preg_replace() call for "fixing" $mossBracketRegex into StaticSiteUrlProcessor#postProcessUrl() |
|
85
|
|
|
* @todo Processor-specific logic (MOSS) should be ported into dedicated class under "Process" namespace |
|
86
|
|
|
*/ |
|
87
|
|
|
public function handleDocumentInfo(PHPCrawlerDocumentInfo $PageInfo): int |
|
88
|
|
|
{ |
|
89
|
|
|
$info = $PageInfo; // upgraded phpcrawler compatibility |
|
90
|
|
|
/* |
|
91
|
|
|
* MOSS has many URLs with brackets, e.g. http://www.stuff.co.nz/news/cat-stuck-up-tree/(/ |
|
92
|
|
|
* These result in 4xx response-codes returned from curl requests for it, and won't filter down to our |
|
93
|
|
|
* caching or URL Processor logic. We can "recover" these URLs by stripping and replacing |
|
94
|
|
|
* with a trailing slash. This allows us to be able to fetch all the URL's children, if any. |
|
95
|
|
|
*/ |
|
96
|
|
|
$isRecoverableUrl = (bool) preg_match('#(\(|%28)+(.+)?$#i', $info->url); |
|
97
|
|
|
// Ignore errors and redirects, they'll get logged for later analysis |
|
98
|
|
|
$badStatusCode = (($info->http_status_code < 200) || ($info->http_status_code > 299)); |
|
99
|
|
|
/* |
|
100
|
|
|
* We're checking for a bad status code AND for "recoverability", becuase we might be able to recover the URL |
|
101
|
|
|
* when re-requesting it during the import stage, as long as we cache it correctly here. |
|
102
|
|
|
*/ |
|
103
|
|
|
if ($badStatusCode && !$isRecoverableUrl) { |
|
104
|
|
|
$message = $info->url . " Skipped. We got a bad status-code and URL was irrecoverable" . PHP_EOL; |
|
105
|
|
|
$this->utils->log($message); |
|
106
|
|
|
|
|
107
|
|
|
return 1; |
|
108
|
|
|
} |
|
109
|
|
|
|
|
110
|
|
|
// Continue building our cache |
|
111
|
|
|
$this->urlList->addAbsoluteURL($info->url, $info->content_type); |
|
112
|
|
|
$this->urlList->saveURLs(); |
|
113
|
|
|
|
|
114
|
|
|
return 0; |
|
115
|
|
|
} |
|
116
|
|
|
|
|
117
|
|
|
/** |
|
118
|
|
|
* |
|
119
|
|
|
* @return void |
|
120
|
|
|
* @throws \InvalidArgumentException |
|
121
|
|
|
*/ |
|
122
|
|
|
protected function initCrawlerProcess(): void |
|
123
|
|
|
{ |
|
124
|
|
|
parent::initCrawlerProcess(); |
|
125
|
|
|
|
|
126
|
|
|
// Add additional URLs to crawl to the crawler's LinkCache |
|
127
|
|
|
// NOTE: This is using an undocumented API |
|
128
|
|
|
if ($extraURLs = $this->urlList->getExtraCrawlURLs()) { |
|
129
|
|
|
foreach ($extraURLs as $extraURL) { |
|
130
|
|
|
$this->LinkCache->addUrl(new PHPCrawlerURLDescriptor($extraURL)); |
|
131
|
|
|
} |
|
132
|
|
|
} |
|
133
|
|
|
|
|
134
|
|
|
// Prevent URLs that match the exclude patterns from being fetched |
|
135
|
|
|
if ($excludePatterns = $this->urlList->getExcludePatterns()) { |
|
136
|
|
|
foreach ($excludePatterns as $pattern) { |
|
137
|
|
|
$validRegExp = $this->addURLFilterRule('|' . str_replace('|', '\|', $pattern) . '|'); |
|
138
|
|
|
|
|
139
|
|
|
if (!$validRegExp) { |
|
140
|
|
|
throw new \InvalidArgumentException('Exclude url pattern "' . $pattern . '" is not a valid regular expression.'); |
|
141
|
|
|
} |
|
142
|
|
|
} |
|
143
|
|
|
} |
|
144
|
|
|
} |
|
145
|
|
|
} |
|
146
|
|
|
|
phptek /
silverstripe-exodus
Loading history...