464 lines
16 KiB
PHP
464 lines
16 KiB
PHP
<?php
|
||
declare (strict_types = 1);
|
||
|
||
namespace app\controller;
|
||
|
||
use app\BaseController;
|
||
use app\service\CrawlerService;
|
||
use think\facade\View;
|
||
|
||
/**
|
||
* 爬虫控制器
|
||
*/
|
||
class Crawler extends BaseController
|
||
{
|
||
/**
|
||
* 显示爬虫工具首页
|
||
*/
|
||
public function index()
|
||
{
|
||
return View::fetch();
|
||
}
|
||
|
||
/**
|
||
* 获取地区选项(从网页HTML中提取)
|
||
*/
|
||
public function getDsdmOptions()
|
||
{
|
||
try {
|
||
$examid = $this->request->param('examid', '');
|
||
$bmid = $this->request->param('bmid', '');
|
||
$userid = $this->request->param('userid', '');
|
||
$cookiesParam = $this->request->param('cookies', '');
|
||
$aa = $this->request->param('aa', (string)round(microtime(true) * 1000));
|
||
|
||
if (empty($examid) || empty($bmid) || empty($userid)) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => '请先填写examid、bmid和userid',
|
||
]);
|
||
}
|
||
|
||
if (empty($cookiesParam)) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => '请填写Cookie数据',
|
||
]);
|
||
}
|
||
|
||
// 解析JSON格式的cookies
|
||
$cookies = is_string($cookiesParam) ? json_decode($cookiesParam, true) : $cookiesParam;
|
||
if (json_last_error() !== JSON_ERROR_NONE || empty($cookies)) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => 'Cookie数据格式错误,请检查JSON格式',
|
||
]);
|
||
}
|
||
|
||
$cookieString = $this->buildCookieString($cookies);
|
||
|
||
// 构建URL获取HTML - 使用GET请求
|
||
$url = "http://gzrsks.oumakspt.com:62/tyzpwb/stuchooseexam/selectPosition.htm?examstupid=1015&userid={$userid}&bmid={$bmid}&examid={$examid}&aa={$aa}";
|
||
|
||
// 构建Referer URL(与浏览器实际访问一致)
|
||
$refererUrl = "http://gzrsks.oumakspt.com:62/tyzpwb/stuchooseexam/input.htm";
|
||
|
||
$ch = curl_init();
|
||
curl_setopt($ch, CURLOPT_URL, $url);
|
||
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
|
||
curl_setopt($ch, CURLOPT_TIMEOUT, 30);
|
||
curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, 10);
|
||
curl_setopt($ch, CURLOPT_FOLLOWLOCATION, true);
|
||
curl_setopt($ch, CURLOPT_SSL_VERIFYPEER, false);
|
||
curl_setopt($ch, CURLOPT_SSL_VERIFYHOST, false);
|
||
curl_setopt($ch, CURLOPT_ENCODING, ''); // 自动处理gzip编码
|
||
curl_setopt($ch, CURLOPT_HTTPHEADER, [
|
||
'Accept: text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3;q=0.7',
|
||
'Accept-Language: zh-CN,zh;q=0.9',
|
||
'Accept-Encoding: gzip, deflate',
|
||
'Cache-Control: no-cache',
|
||
'Connection: keep-alive',
|
||
'Cookie: ' . $cookieString,
|
||
'User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/143.0.0.0 Safari/537.36',
|
||
'Referer: ' . $refererUrl,
|
||
'Upgrade-Insecure-Requests: 1',
|
||
]);
|
||
|
||
$html = curl_exec($ch);
|
||
$httpCode = curl_getinfo($ch, CURLINFO_HTTP_CODE);
|
||
$error = curl_error($ch);
|
||
curl_close($ch);
|
||
|
||
if ($error) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => '获取网页失败: ' . $error,
|
||
]);
|
||
}
|
||
|
||
if ($httpCode !== 200) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => '获取网页失败,HTTP状态码: ' . $httpCode,
|
||
]);
|
||
}
|
||
|
||
$service = new CrawlerService();
|
||
$options = $service->extractDsdmOptions($html);
|
||
|
||
return json([
|
||
'code' => 1,
|
||
'data' => $options,
|
||
'msg' => '获取成功',
|
||
]);
|
||
|
||
} catch (\Exception $e) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => '获取失败: ' . $e->getMessage(),
|
||
]);
|
||
}
|
||
}
|
||
|
||
/**
|
||
* 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,支持同名键多值(如双JSESSIONID)
|
||
$cookieData = $cookies['请求 Cookie'] ?? $cookies;
|
||
$normalized = [];
|
||
|
||
foreach ($cookieData as $key => $value) {
|
||
$k = trim((string)$key);
|
||
if ($k === '') {
|
||
continue;
|
||
}
|
||
$vals = is_array($value) ? $value : [$value];
|
||
foreach ($vals as $v) {
|
||
$v = trim((string)$v);
|
||
if ($v === '') {
|
||
continue;
|
||
}
|
||
$normalized[$k][] = $v;
|
||
}
|
||
}
|
||
|
||
// 将 JSESSIONID 放在最前(如果存在),其余按键名顺序,多值全部保留
|
||
$parts = [];
|
||
if (isset($normalized['JSESSIONID'])) {
|
||
foreach ($normalized['JSESSIONID'] as $v) {
|
||
$parts[] = 'JSESSIONID=' . $v;
|
||
}
|
||
unset($normalized['JSESSIONID']);
|
||
}
|
||
foreach ($normalized as $k => $vArr) {
|
||
foreach ($vArr as $v) {
|
||
$parts[] = $k . '=' . $v;
|
||
}
|
||
}
|
||
|
||
return implode('; ', $parts);
|
||
}
|
||
|
||
/**
|
||
* 获取职位代码列表
|
||
*/
|
||
public function getZwdmList()
|
||
{
|
||
try {
|
||
$dsdm = $this->request->param('dsdm', '');
|
||
$examid = $this->request->param('examid', '');
|
||
$bmid = $this->request->param('bmid', '');
|
||
$userid = $this->request->param('userid', '');
|
||
$aa = $this->request->param('aa', '');
|
||
$cookiesParam = $this->request->param('cookies', '');
|
||
|
||
if (empty($dsdm)) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => '请选择地区',
|
||
]);
|
||
}
|
||
|
||
if (empty($examid) || empty($bmid) || empty($userid)) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => '请填写examid、bmid和userid',
|
||
]);
|
||
}
|
||
|
||
if (empty($aa)) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => '缺少aa参数,请先获取地区选项再获取职位代码',
|
||
]);
|
||
}
|
||
|
||
if (empty($cookiesParam)) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => '请填写Cookie数据',
|
||
]);
|
||
}
|
||
|
||
// 解析JSON格式的cookies
|
||
$cookies = is_string($cookiesParam) ? json_decode($cookiesParam, true) : $cookiesParam;
|
||
if (json_last_error() !== JSON_ERROR_NONE || empty($cookies)) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => 'Cookie数据格式错误,请检查JSON格式',
|
||
]);
|
||
}
|
||
|
||
$service = new CrawlerService();
|
||
$treeData = $service->getPositionTree($dsdm, $examid, $bmid, $userid, (string)$aa, $cookies);
|
||
|
||
// 提取所有CODE作为zwdm
|
||
$zwdmList = [];
|
||
if (is_array($treeData)) {
|
||
foreach ($treeData as $item) {
|
||
if (isset($item['CODE']) && !empty($item['CODE'])) {
|
||
$zwdmList[] = [
|
||
'zwdm' => $item['CODE'],
|
||
'title' => $item['TITLE'] ?? $item['CODE'],
|
||
];
|
||
}
|
||
}
|
||
}
|
||
|
||
return json([
|
||
'code' => 1,
|
||
'data' => $zwdmList,
|
||
'msg' => '获取成功',
|
||
]);
|
||
|
||
} catch (\Exception $e) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => '获取失败: ' . $e->getMessage(),
|
||
]);
|
||
}
|
||
}
|
||
|
||
/**
|
||
* 自动获取全部职位代码并返回职位详情(无需手动勾选)
|
||
*/
|
||
public function fetchAllPositions()
|
||
{
|
||
try {
|
||
$dsdm = $this->request->param('dsdm', '');
|
||
$examid = $this->request->param('examid', '');
|
||
$bmid = $this->request->param('bmid', '');
|
||
$userid = $this->request->param('userid', '');
|
||
$aa = $this->request->param('aa', '');
|
||
$cookiesParam = $this->request->param('cookies', '');
|
||
|
||
if (empty($dsdm) || empty($examid) || empty($bmid) || empty($userid) || empty($aa)) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => '请填写examid、bmid、userid、dsdm,并先获取地区选项生成aa',
|
||
]);
|
||
}
|
||
|
||
if (empty($cookiesParam)) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => '请填写Cookie数据',
|
||
]);
|
||
}
|
||
|
||
$cookies = is_string($cookiesParam) ? json_decode($cookiesParam, true) : $cookiesParam;
|
||
if (json_last_error() !== JSON_ERROR_NONE || empty($cookies)) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => 'Cookie数据格式错误,请检查JSON格式',
|
||
]);
|
||
}
|
||
|
||
$service = new CrawlerService();
|
||
$treeData = $service->getPositionTree($dsdm, $examid, $bmid, $userid, (string)$aa, $cookies);
|
||
|
||
$zwdmList = [];
|
||
if (is_array($treeData)) {
|
||
foreach ($treeData as $item) {
|
||
if (isset($item['CODE']) && !empty($item['CODE'])) {
|
||
$code = $item['CODE'];
|
||
// 跳过1520开头的职位代码
|
||
if (strpos($code, '1520') === 0) {
|
||
continue;
|
||
}
|
||
$zwdmList[] = $code;
|
||
}
|
||
}
|
||
}
|
||
|
||
if (empty($zwdmList)) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => '未获取到职位代码',
|
||
]);
|
||
}
|
||
|
||
$results = $service->batchGetPositionInfo($zwdmList, $examid, $cookies);
|
||
|
||
return json([
|
||
'code' => 1,
|
||
'data' => $results,
|
||
'msg' => '获取成功',
|
||
]);
|
||
} catch (\Exception $e) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => '获取失败: ' . $e->getMessage(),
|
||
]);
|
||
}
|
||
}
|
||
|
||
/**
|
||
* 获取职位详细信息
|
||
*/
|
||
public function getPositionInfo()
|
||
{
|
||
try {
|
||
$zwdm = $this->request->param('zwdm', '');
|
||
$examid = $this->request->param('examid', '');
|
||
$cookiesParam = $this->request->param('cookies', '');
|
||
|
||
if (empty($zwdm) || empty($examid)) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => '参数不完整',
|
||
]);
|
||
}
|
||
|
||
if (empty($cookiesParam)) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => '请填写Cookie数据',
|
||
]);
|
||
}
|
||
|
||
// 解析JSON格式的cookies
|
||
$cookies = is_string($cookiesParam) ? json_decode($cookiesParam, true) : $cookiesParam;
|
||
if (json_last_error() !== JSON_ERROR_NONE || empty($cookies)) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => 'Cookie数据格式错误,请检查JSON格式',
|
||
]);
|
||
}
|
||
|
||
$service = new CrawlerService();
|
||
$info = $service->getPositionInfo($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;
|
||
|
||
$result = [
|
||
'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
|
||
];
|
||
|
||
return json([
|
||
'code' => 1,
|
||
'data' => $result,
|
||
'msg' => '获取成功',
|
||
]);
|
||
} else {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => '未获取到数据',
|
||
]);
|
||
}
|
||
|
||
} catch (\Exception $e) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => '获取失败: ' . $e->getMessage(),
|
||
]);
|
||
}
|
||
}
|
||
|
||
/**
|
||
* 批量获取职位信息
|
||
*/
|
||
public function batchGetPositionInfo()
|
||
{
|
||
try {
|
||
$zwdmListParam = $this->request->param('zwdm_list', '');
|
||
$examid = $this->request->param('examid', '');
|
||
$cookiesParam = $this->request->param('cookies', '');
|
||
|
||
// 解析JSON格式的zwdm_list
|
||
$zwdmList = is_string($zwdmListParam) ? json_decode($zwdmListParam, true) : $zwdmListParam;
|
||
if (json_last_error() !== JSON_ERROR_NONE || empty($zwdmList) || !is_array($zwdmList)) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => '请选择职位代码',
|
||
]);
|
||
}
|
||
|
||
if (empty($examid)) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => '请填写examid',
|
||
]);
|
||
}
|
||
|
||
if (empty($cookiesParam)) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => '请填写Cookie数据',
|
||
]);
|
||
}
|
||
|
||
// 解析JSON格式的cookies
|
||
$cookies = is_string($cookiesParam) ? json_decode($cookiesParam, true) : $cookiesParam;
|
||
if (json_last_error() !== JSON_ERROR_NONE || empty($cookies)) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => 'Cookie数据格式错误,请检查JSON格式',
|
||
]);
|
||
}
|
||
|
||
$service = new CrawlerService();
|
||
$results = $service->batchGetPositionInfo($zwdmList, $examid, $cookies);
|
||
|
||
return json([
|
||
'code' => 1,
|
||
'data' => $results,
|
||
'msg' => '获取成功',
|
||
]);
|
||
|
||
} catch (\Exception $e) {
|
||
return json([
|
||
'code' => 0,
|
||
'msg' => '获取失败: ' . $e->getMessage(),
|
||
]);
|
||
}
|
||
}
|
||
}
|