260 lines
8.9 KiB
PHP
260 lines
8.9 KiB
PHP
<?php
|
||
declare (strict_types = 1);
|
||
|
||
namespace app\service;
|
||
|
||
/**
|
||
* 爬虫服务类
|
||
* 用于处理职位信息爬取相关逻辑
|
||
*/
|
||
class CrawlerService
|
||
{
|
||
/**
|
||
* 基础URL
|
||
*/
|
||
private $baseUrl = 'http://gzrsks.oumakspt.com:62/tyzpwb';
|
||
|
||
/**
|
||
* 获取职位树数据
|
||
* @param string $dsdm 地区代码
|
||
* @param array $cookies Cookie数据
|
||
* @return array
|
||
*/
|
||
public function getPositionTree(string $dsdm, array $cookies): array
|
||
{
|
||
$url = $this->baseUrl . '/tree/getPositionTree.htm';
|
||
|
||
$cookieString = $this->buildCookieString($cookies);
|
||
|
||
$ch = curl_init();
|
||
curl_setopt($ch, CURLOPT_URL, $url);
|
||
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
|
||
curl_setopt($ch, CURLOPT_FOLLOWLOCATION, true);
|
||
curl_setopt($ch, CURLOPT_SSL_VERIFYPEER, false);
|
||
curl_setopt($ch, CURLOPT_SSL_VERIFYHOST, false);
|
||
curl_setopt($ch, CURLOPT_HTTPHEADER, [
|
||
'Accept: application/json, text/javascript, */*; q=0.01',
|
||
'Accept-Language: zh-CN,zh;q=0.9,en;q=0.8',
|
||
'Accept-Encoding: gzip, deflate',
|
||
'Cache-Control: no-cache',
|
||
'Connection: keep-alive',
|
||
'Content-Type: application/x-www-form-urlencoded',
|
||
'Cookie: ' . $cookieString,
|
||
'User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/122.0.0.0 Safari/537.36',
|
||
'Referer: http://gzrsks.oumakspt.com:62/tyzpwb/',
|
||
'X-Requested-With: XMLHttpRequest',
|
||
]);
|
||
curl_setopt($ch, CURLOPT_POST, true);
|
||
curl_setopt($ch, CURLOPT_POSTFIELDS, http_build_query(['dsdm' => $dsdm]));
|
||
curl_setopt($ch, CURLOPT_TIMEOUT, 30);
|
||
curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, 10);
|
||
|
||
$response = curl_exec($ch);
|
||
$httpCode = curl_getinfo($ch, CURLINFO_HTTP_CODE);
|
||
$error = curl_error($ch);
|
||
curl_close($ch);
|
||
|
||
if ($error) {
|
||
throw new \Exception('请求失败: ' . $error);
|
||
}
|
||
|
||
if ($httpCode !== 200) {
|
||
throw new \Exception('请求失败，HTTP状态码: ' . $httpCode);
|
||
}
|
||
|
||
$data = json_decode($response, true);
|
||
|
||
if (json_last_error() !== JSON_ERROR_NONE) {
|
||
throw new \Exception('JSON解析失败: ' . json_last_error_msg());
|
||
}
|
||
|
||
return $data ?: [];
|
||
}
|
||
|
||
/**
|
||
* 获取职位详细信息
|
||
* @param string $zwdm 职位代码
|
||
* @param string $examid 考试ID
|
||
* @param array $cookies Cookie数据
|
||
* @return array
|
||
*/
|
||
public function getPositionInfo(string $zwdm, string $examid, array $cookies): array
|
||
{
|
||
$url = $this->baseUrl . '/stuchooseexam/getPositionInfo.htm';
|
||
|
||
$cookieString = $this->buildCookieString($cookies);
|
||
|
||
$ch = curl_init();
|
||
curl_setopt($ch, CURLOPT_URL, $url);
|
||
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
|
||
curl_setopt($ch, CURLOPT_FOLLOWLOCATION, true);
|
||
curl_setopt($ch, CURLOPT_SSL_VERIFYPEER, false);
|
||
curl_setopt($ch, CURLOPT_SSL_VERIFYHOST, false);
|
||
curl_setopt($ch, CURLOPT_HTTPHEADER, [
|
||
'Accept: application/json, text/javascript, */*; q=0.01',
|
||
'Accept-Language: zh-CN,zh;q=0.9,en;q=0.8',
|
||
'Accept-Encoding: gzip, deflate',
|
||
'Cache-Control: no-cache',
|
||
'Connection: keep-alive',
|
||
'Content-Type: application/x-www-form-urlencoded',
|
||
'Cookie: ' . $cookieString,
|
||
'User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/122.0.0.0 Safari/537.36',
|
||
'Referer: http://gzrsks.oumakspt.com:62/tyzpwb/',
|
||
'X-Requested-With: XMLHttpRequest',
|
||
]);
|
||
curl_setopt($ch, CURLOPT_POST, true);
|
||
curl_setopt($ch, CURLOPT_POSTFIELDS, http_build_query([
|
||
'zwdm' => $zwdm,
|
||
'examid' => $examid
|
||
]));
|
||
curl_setopt($ch, CURLOPT_TIMEOUT, 30);
|
||
curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, 10);
|
||
|
||
$response = curl_exec($ch);
|
||
$httpCode = curl_getinfo($ch, CURLINFO_HTTP_CODE);
|
||
$error = curl_error($ch);
|
||
curl_close($ch);
|
||
|
||
if ($error) {
|
||
throw new \Exception('请求失败: ' . $error);
|
||
}
|
||
|
||
if ($httpCode !== 200) {
|
||
throw new \Exception('请求失败，HTTP状态码: ' . $httpCode);
|
||
}
|
||
|
||
$data = json_decode($response, true);
|
||
|
||
if (json_last_error() !== JSON_ERROR_NONE) {
|
||
throw new \Exception('JSON解析失败: ' . json_last_error_msg());
|
||
}
|
||
|
||
return $data ?: [];
|
||
}
|
||
|
||
/**
|
||
* 批量获取职位信息
|
||
* @param array $zwdmList 职位代码列表
|
||
* @param string $examid 考试ID
|
||
* @param array $cookies Cookie数据
|
||
* @return array
|
||
*/
|
||
public function batchGetPositionInfo(array $zwdmList, string $examid, array $cookies): array
|
||
{
|
||
$results = [];
|
||
|
||
foreach ($zwdmList as $zwdm) {
|
||
try {
|
||
$info = $this->getPositionInfo($zwdm, $examid, $cookies);
|
||
|
||
if (!empty($info)) {
|
||
// 处理单条数据或数组数据
|
||
if (isset($info[0])) {
|
||
$item = $info[0];
|
||
} else {
|
||
$item = $info;
|
||
}
|
||
|
||
// 计算竞争比
|
||
$zprs = isset($item['zprs']) ? intval($item['zprs']) : 0;
|
||
$bkrs = isset($item['bkrs']) ? intval($item['bkrs']) : 0;
|
||
$competitionRatio = $zprs > 0 ? ($bkrs / $zprs) : 0;
|
||
|
||
$results[] = [
|
||
'sbmc' => $item['sbmc'] ?? '', // 省份
|
||
'dsmc' => $item['dsmc'] ?? '', // 地区
|
||
'zpdwmc' => $item['zpdwmc'] ?? '', // 招聘单位/用人司局
|
||
'zwmc' => $item['zwmc'] ?? '', // 职位名称
|
||
'zwdm' => $item['zwdm'] ?? $zwdm, // 职位代码
|
||
'zprs' => $zprs, // 招聘人数
|
||
'bkrs' => $bkrs, // 审核通过人数
|
||
'competition_ratio' => number_format($competitionRatio, 2), // 竞争比
|
||
];
|
||
}
|
||
|
||
// 避免请求过快，添加小延迟
|
||
usleep(200000); // 0.2秒
|
||
|
||
} catch (\Exception $e) {
|
||
// 记录错误但继续处理其他数据
|
||
$results[] = [
|
||
'zwdm' => $zwdm,
|
||
'error' => $e->getMessage(),
|
||
];
|
||
}
|
||
}
|
||
|
||
return $results;
|
||
}
|
||
|
||
/**
|
||
* 构建Cookie字符串
|
||
* @param array $cookies Cookie数组
|
||
* @return string
|
||
*/
|
||
private function buildCookieString(array|string $cookies): string
|
||
{
|
||
// 如果直接传入原始Cookie字符串，优先使用
|
||
if (is_string($cookies)) {
|
||
return trim($cookies);
|
||
}
|
||
|
||
// 只保留用户填写的Cookie，若同名键多值则取最后一个（单JSESSIONID场景）
|
||
$cookieData = $cookies['请求 Cookie'] ?? $cookies;
|
||
$normalized = [];
|
||
|
||
foreach ($cookieData as $key => $value) {
|
||
$k = trim((string)$key);
|
||
if ($k === '') {
|
||
continue;
|
||
}
|
||
$v = is_array($value) ? end($value) : $value;
|
||
$v = trim((string)$v);
|
||
if ($v === '') {
|
||
continue;
|
||
}
|
||
$normalized[$k] = $v;
|
||
}
|
||
|
||
// 将 JSESSIONID 放在最前（如果存在），其余按键名顺序
|
||
$parts = [];
|
||
if (isset($normalized['JSESSIONID'])) {
|
||
// 按需求：输出两个相同的 JSESSIONID
|
||
$parts[] = 'JSESSIONID=' . $normalized['JSESSIONID'];
|
||
$parts[] = 'JSESSIONID=' . $normalized['JSESSIONID'];
|
||
unset($normalized['JSESSIONID']);
|
||
}
|
||
foreach ($normalized as $k => $v) {
|
||
$parts[] = $k . '=' . $v;
|
||
}
|
||
|
||
return implode('; ', $parts);
|
||
}
|
||
|
||
/**
|
||
* 从HTML中提取地区选项
|
||
* @param string $html HTML内容
|
||
* @return array
|
||
*/
|
||
public function extractDsdmOptions(string $html): array
|
||
{
|
||
$options = [];
|
||
|
||
// 使用正则表达式提取option标签
|
||
preg_match_all('/<option\s+value="([^"]+)"[^>]*>([^<]+)<\/option>/i', $html, $matches);
|
||
|
||
if (!empty($matches[1])) {
|
||
foreach ($matches[1] as $index => $value) {
|
||
if (!empty($value)) { // 跳过"请选择"等空值
|
||
$options[] = [
|
||
'value' => $value,
|
||
'text' => $matches[2][$index] ?? $value,
|
||
];
|
||
}
|
||
}
|
||
}
|
||
|
||
return $options;
|
||
}
|
||
}
|