You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
77 lines
2.6 KiB
77 lines
2.6 KiB
<?php
|
|
|
|
namespace App\Services;
|
|
|
|
use App\Utils\CommonUtils;
|
|
use Illuminate\Support\Arr;
|
|
use QL\QueryList;
|
|
|
|
class ComicsService
|
|
{
|
|
|
|
private $queryInstance;
|
|
|
|
public function __construct()
|
|
{
|
|
$this->queryInstance = QueryList::getInstance();
|
|
}
|
|
|
|
/**
|
|
* @return QueryList|null
|
|
*/
|
|
public function getQueryInstance(): ?QueryList
|
|
{
|
|
return $this->queryInstance;
|
|
}
|
|
|
|
public function scrapeAllAlbum($rootUrl)
|
|
{
|
|
$content = $this->getQueryInstance()->get($rootUrl);
|
|
$rules = [
|
|
// 采集文章标题
|
|
'title' => ['.thumbnail', 'title'],
|
|
// 采集链接
|
|
'link' => ['.thumbnail', 'href'],
|
|
// 采集缩略图
|
|
'totalNum' => ['body > section > div.content-wrap > div > div.pagination.pagination-multi > ul > li:nth-child(8) > span', 'text'],
|
|
];
|
|
// $links = $content->find("a.thumbnail")->attrs("href");
|
|
$range = ".pb article";
|
|
$list = $content->rules($rules)->range($range)->query()->getData();
|
|
dump($list->all());
|
|
$parsedList = array_merge([], $list->all());
|
|
$totalPageNumMatchedItems = $content->find("body > section > div.content-wrap > div > div.pagination.pagination-multi > ul > li:nth-child(8) > span")->texts();
|
|
$totalPageNumStr = Arr::get($totalPageNumMatchedItems->all(), 0);
|
|
preg_match("#\d+#", $totalPageNumStr, $totalPageNumPregMatchedItems);
|
|
$totalPageNum = $totalPageNumPregMatchedItems[0];
|
|
// $totalPageNum = body > section > div.content-wrap > div > div.pagination.pagination-multi > ul > li:nth-child(8) > span
|
|
// $links->dump();
|
|
|
|
for ($i = 2; $i < $totalPageNum; $i++) {
|
|
break;
|
|
$pageUrl = $rootUrl . "/page/" . $i;
|
|
$content = $this->getQueryInstance()->get($pageUrl);
|
|
$parsedList[] = $content->rules($rules)->range($range)->query()->getData()->all();
|
|
CommonUtils::randomSleep();
|
|
break;
|
|
}
|
|
dump($parsedList);
|
|
foreach ($parsedList as $item) {
|
|
$this->processAlbum($item);
|
|
exit;
|
|
}
|
|
}
|
|
|
|
public function processAlbum($item)
|
|
{
|
|
$content = $this->getQueryInstance()->get($item["link"]);
|
|
$rules = [
|
|
"link" => [".post-page-numbers", "href"]
|
|
];
|
|
$range = ".article-content > div:nth-child(1) a";
|
|
// $list = $content->rules($rules)->range($range)->query()->getData();
|
|
$list = $content->find(".article-content > div:nth-child(1) a")->attrs("href");
|
|
dump($list->all());
|
|
|
|
}
|
|
}
|