226 lines
9.5 KiB
PHP
226 lines
9.5 KiB
PHP
<?php
|
||
|
||
namespace app\event;
|
||
|
||
use app\model\TiktokAd;
|
||
use GuzzleHttp\Client;
|
||
use GuzzleHttp\Exception\GuzzleException;
|
||
use support\Db;
|
||
use QL\QueryList;
|
||
use support\Redis;
|
||
|
||
|
||
class TiktokAds
|
||
{
|
||
//微博热榜地址
|
||
// const url = 'https://library.tiktok.com/api/v1/search?region=GB&type=1&start_time=1666540800&end_time=1666627200';
|
||
|
||
const userAgent = 'user-agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36';
|
||
|
||
const type = 'tiktokads';
|
||
const limit = 12;
|
||
const sort_order = 'impression,desc';
|
||
|
||
|
||
// const countries = ["AT", "BE", "BG", "HR", "CY", "CZ", "DK", "EE", "FI", "FR", "DE", "GR", "HU", "IS", "IE","IT", "LV", "LI", "LT", "LU", "MT", "NL", "NO", "PL", "PT", "RO", "SK", "SI", "ES", "SE", "CH", "TR", "GB"] ;
|
||
const countries = ["GB", "BE"];
|
||
|
||
|
||
/**
|
||
* 每天爬取tiktok广告
|
||
* @return void
|
||
*/
|
||
public function update()
|
||
{
|
||
try {
|
||
|
||
$client = new Client([
|
||
//允许重定向
|
||
// 'allow_redirects' => true,
|
||
]);
|
||
|
||
|
||
// 获取前两天 0 点的时间戳
|
||
$dayBeforeYesterdayStart = strtotime('-2 days 00:00:00');
|
||
|
||
// 获取前一天 0 点的时间戳
|
||
$yesterdayStart = strtotime('-1 day 00:00:00');
|
||
$countryCache = Redis::get(self::type . 'lastCountry');
|
||
//全部跑完跳出
|
||
if ($countryCache == 'All') {
|
||
dump($countryCache . '国家更新tiktok Ads 完成');
|
||
return;
|
||
}
|
||
|
||
if (empty($countryCache)) {
|
||
$countryCache = 'GB';
|
||
}
|
||
|
||
$searchIdCache = Redis::get(self::type . 'lastSearchId');
|
||
$offsetCache = Redis::get(self::type . 'nextOffset');
|
||
$totalCache = Redis::get(self::type . 'totalCache');
|
||
if (!isset($searchIdCache) || empty($searchIdCache)) {
|
||
$searchIdCache = '';
|
||
}
|
||
if (!isset($offsetCache) || empty($searchIdCache)) {
|
||
$offsetCache = 0;
|
||
}
|
||
if (!isset($totalCache) || empty($searchIdCache)) {
|
||
$totalCache = 0;
|
||
}
|
||
|
||
//判断国家是否跑完。
|
||
if ($totalCache > 0 && $offsetCache > ceil($totalCache / self::limit)) {
|
||
$key = array_search($countryCache, self::countries, true);
|
||
if (in_array($countryCache, self::countries) && isset(self::countries[$key + 1])) {
|
||
$countryCache = self::countries[$key + 1];
|
||
dump('更新' . $countryCache . '国家的tiktok Ads 完成');
|
||
} else {
|
||
$countryCache = 'All'; //赋值非正常国家的code中断定时任务
|
||
dump($countryCache . '国家更新tiktok Ads 完成');
|
||
}
|
||
return;
|
||
}
|
||
|
||
|
||
$start_time = $dayBeforeYesterdayStart;
|
||
$end_time = $yesterdayStart;
|
||
//当前国家爬取
|
||
$currentParams = null;
|
||
$currentParams = ['country' => $countryCache, 'search_id' => $searchIdCache, 'type' => 1, 'start_time' => $start_time, 'end_time' => $end_time];
|
||
|
||
|
||
$url = 'https://library.tiktok.com/api/v1/search?region=' . $currentParams['country'] . '&type=' . $currentParams['type'] . '&start_time=' . $currentParams['start_time'] . '&end_time=' . $currentParams['end_time'];
|
||
|
||
$res = json_decode($client->post($url, [
|
||
'headers' => [
|
||
'accept' => 'application/json, text/plain, */*',
|
||
'accept-language' => 'zh-CN,zh;q=0.9',
|
||
'content-type' => 'application/json',
|
||
'cookie' => 'cookie: _ttp=2ov8Fc4C2CaNscHJd90O9fMhlpE; _ga=GA1.1.1025820618.1731926196; FPID=FPID2.2.Bcgkp%2Fk%2Bbn5w5YeSMR9wd9VpNHJwTUpkkaEqSdCEa0w%3D.1731926196; FPAU=1.2.944915349.1731926193; FPLC=mbVyryI5aG6IVpAvhs1JsgWjA7FVA6QsCJ7VbXhM7zWoXNp4rcD0IK7FNTTf%2FuOrqeOgqEhTd4NB3hY7q3aDVTGQa3WGHqxkGte4%2BBZxsrpaHFas9kb7DPRXM12T5Q%3D%3D; _ga_TEQXTT9FE4=GS1.1.1732097542.7.0.1732097542.0.0.857840528',
|
||
'origin' => 'https://library.tiktok.com',
|
||
'priority' => 'u=1, i',
|
||
'referer' => 'https://library.tiktok.com/ads?region=AT&start_time=1731945600000&end_time=1732032000000&adv_name=&adv_biz_ids=&query_type=&sort_type=last_shown_date,desc',
|
||
'sec-ch-ua' => '"Google Chrome";v="131", "Chromium";v="131", "Not_A Brand";v="24"',
|
||
'sec-ch-ua-mobile' => '?0',
|
||
'sec-ch-ua-platform' => '"Windows"',
|
||
'sec-fetch-dest' => 'empty',
|
||
'sec-fetch-mode' => 'cors',
|
||
'sec-fetch-site' => 'same-origin',
|
||
'user-agent' => self::userAgent,
|
||
// 'user-agent' => 'user-agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36',
|
||
],
|
||
'json' => [
|
||
'query' => '',
|
||
'query_type' => '',
|
||
'adv_biz_ids' => '',
|
||
'order' => self::sort_order,
|
||
'offset' => (int) $offsetCache,
|
||
'search_id' => $currentParams['search_id'],
|
||
'limit' => self::limit,
|
||
],
|
||
|
||
])->getBody()->getContents(), true);
|
||
// dump($res);return; //调试点
|
||
|
||
if ($res['search_id'] != $searchIdCache) {
|
||
$searchIdCache = $res['search_id'];
|
||
dump( 'search_id更新 ' . $searchIdCache . ' 成功');
|
||
}
|
||
|
||
|
||
if ($res['total'] == 0 || $res['code'] != 0) {
|
||
dump('更新tiktok Ads接口响应异常：' . json_encode($res, JSON_UNESCAPED_UNICODE));
|
||
return;
|
||
}
|
||
$listAdsIds = [];
|
||
foreach ($res['data'] as $ad) {
|
||
if($ad['audit_status'] == 2 || empty($ad['videos'])){
|
||
continue; //审核不过或者没视频不采集
|
||
}
|
||
|
||
$imagesJson = is_array($ad['image_urls']) ? json_encode($ad['image_urls']):json_encode([]);
|
||
$rejection_info = is_array($ad['rejection_info']) ? json_encode($ad['rejection_info']):null;
|
||
// dump($rejection_info);
|
||
$insertData[$ad['id']] = [
|
||
'ad_id' => $ad['id'],
|
||
'name' => $ad['name'],
|
||
'audit_status' => $ad['audit_status'],
|
||
'type' => $ad['type'],
|
||
'first_shown_date' => $ad['first_shown_date'],
|
||
'last_shown_date' => $ad['last_shown_date'],
|
||
'image_urls' => $imagesJson,
|
||
'estimated_audience' => $ad['estimated_audience'],
|
||
'spent' => $ad['spent'],
|
||
'impression' => $ad['impression'],
|
||
'show_mode' => $ad['show_mode'],
|
||
'rejection_info' => $rejection_info,
|
||
'sor_audit_status' => $ad['sor_audit_status'],
|
||
'country_code' => $countryCache,
|
||
];
|
||
if(isset($ad['videos']) && !empty($ad['videos'])) {
|
||
// 遍历 "videos" 数组
|
||
foreach ($ad['videos'] as $video) {
|
||
$insertData[$ad['id']]['video_url'] = $video['video_url'];
|
||
$insertData[$ad['id']]['cover_img'] = $video['cover_img'];
|
||
}
|
||
}
|
||
$listAdsIds = array_column($insertData, 'ad_id');
|
||
|
||
}
|
||
// dump($insertData);return;
|
||
|
||
|
||
if (empty($insertData)) return;
|
||
|
||
//开启事务
|
||
Db::beginTransaction();
|
||
|
||
|
||
//删除原来的旧数据
|
||
TiktokAd::query()->whereIn('ad_id', array_keys($insertData))->delete();
|
||
//添加新的数据
|
||
TiktokAd::query()->insert($insertData);
|
||
|
||
//redis缓存
|
||
Redis::set(self::type, json_encode($insertData, JSON_UNESCAPED_UNICODE));
|
||
|
||
//redis缓存 记录更新时间
|
||
$time = date('Y-m-d H:i:s');
|
||
Redis::set(self::type . 'time', $time);
|
||
Redis::set(self::type . 'lastCountry', $countryCache);
|
||
Redis::set(self::type . 'nextOffset', ++$offsetCache); //记录下一次的offset
|
||
Redis::set(self::type . 'totalCache', $res['total']);
|
||
Redis::set(self::type . 'lastSearchId', $searchIdCache);
|
||
if(!empty($listAdsIds)){
|
||
Redis::rPush(self::type . 'AdsIds',...$listAdsIds);
|
||
}
|
||
|
||
//提交事务
|
||
Db::commit();
|
||
//销毁$res
|
||
unset($res);
|
||
|
||
dump(date('Y-m-d H:i:s') . '更新' . self::type . '成功');
|
||
} catch (GuzzleException|\Exception $exception) {
|
||
//回滚事务
|
||
Db::rollBack();
|
||
dump('更' . self::type . '异常：' . $exception->getMessage());
|
||
dump($exception);
|
||
}
|
||
// } catch (ClientExceptionInterface $e) {
|
||
// // 捕获 4xx 错误
|
||
// dump( 'Client error: ' . $e->getMessage() . "\n");
|
||
// } catch (ServerExceptionInterface $e) {
|
||
// // 捕获 5xx 错误
|
||
// dump('Server error: ' . $e->getMessage() . "\n");
|
||
// } catch (TransportExceptionInterface $e) {
|
||
// // 捕获网络传输错误
|
||
// dump('Transport error: ' . $e->getMessage() . "\n") ;
|
||
// } catch (\Exception $e) {
|
||
// // 捕获所有其他错误
|
||
// dump('General error: ' . $e->getMessage() . "\n") ;
|
||
// }
|
||
}
|
||
}
|