spider_creater/backend/vm/check.go

package vm

import (
	"container/list"
	"errors"
	"fmt"
	"github.com/chromedp/chromedp"
	qu "jygit.jydev.jianyu360.cn/data_processing/common_utils"
	be "spider_creator/backend"
	"time"
)

// 爬虫验证
func (vm *VM) VerifySpiderConfig(sc *be.SpiderConfig) (*be.SpiderConfigVerifyResult, error, []string) {
	if be.Cfg.IsOnly4MainSite {
		return vm.verifySpiderConfig4MainSite(sc) //重点网站
	} else {
		return vm.verifySpiderConfig4Prod(sc) //正式环境
	}
}

// verifySpiderConfig4Prod 验证爬虫配置，支持翻页，列表项数据只提取2条
// 正式环境
func (vm *VM) verifySpiderConfig4Prod(sc *be.SpiderConfig) (*be.SpiderConfigVerifyResult, error, []string) {
	qu.Debug("sc---", *sc)
	var errMsg []string
	verifyResult := list.New()
	be.DataResults[sc.Code] = verifyResult
	ret := &be.SpiderConfigVerifyResult{false, false, false, false, false, false, false}
	_, baseCancelFn, _, _, ctx, incCancelFn := be.NewBrowser(false, true, false, sc.Href, sc.FilterResource)    //列表页使用
	_, baseCancelFn2, _, _, ctx2, incCancelFn2 := be.NewBrowser(false, true, false, sc.Href, sc.FilterResource) //详情页使用
	defer func() {
		incCancelFn2()
		baseCancelFn2()
		incCancelFn()
		baseCancelFn()
	}()

	listRunJs, contentRunJs := sc.ListJSCode, sc.ContentJSCode
	//2. 执行JS代码，获取列表页信息
	if be.RegSpace.ReplaceAllString(listRunJs, "") == "" {
		listRunJs = renderJavascriptCoder(loadListItemsJS, sc)
	}
	if be.RegSpace.ReplaceAllString(contentRunJs, "") == "" {
		contentRunJs = renderJavascriptCoder(loadContentJS, sc)
	}
	qu.Debug("获取列表页JS代码：", listRunJs)
	qu.Debug("获取详情页JS代码：", contentRunJs)
	//3.打开列表，获取条目清单
	chromedp.Run(ctx, chromedp.Tasks{
		chromedp.Navigate(sc.Href),
		chromedp.WaitReady("document.body", chromedp.ByJSPath),
		//chromedp.Sleep(1000 * time.Millisecond),
		chromedp.Sleep(time.Duration(sc.ListDelayTime) * time.Millisecond),
	})
	//初始化列表页信息
	if !vm.InitListPage(ctx, sc) {
		qu.Debug("初始化列表页失败，退出")
		return ret, errors.New("初始化列表页失败"), errMsg
	}
	no := 1
	ret.ListTrunPage = true
T:
	for j := 0; j < VERIVY_MAX_TRUN_PAGE && j < int(sc.MaxPages); j++ { //最多检查2页
		qu.Debug("开始检查第" + fmt.Sprint(j+1) + "页...")
		listResult := make(be.ResultItems, 0)
		err := chromedp.Run(ctx, chromedp.Tasks{
			chromedp.Evaluate(listRunJs, &listResult),
		})
		if err != nil {
			qu.Debug("执行列表页JS代码失败", err.Error())
			errMsg = append(errMsg, fmt.Sprintf("%s%d%s%s", "第", j+1, "页列表数据获取失败：", err.Error()))
			continue
		}
		//TODO 5.操作详情页
		qu.Debug("列表采集条数：", len(listResult))
		for contentIndex, r := range listResult {
			qu.Debug("当前列表页第" + fmt.Sprint(contentIndex+1) + "条")
			if contentIndex > 1 { //每页只校验2条
				break
			}
			//打开详情页
			err = chromedp.Run(ctx2, chromedp.Tasks{
				chromedp.Navigate(r.Href),
				chromedp.WaitReady("document.body", chromedp.ByJSPath),
				//chromedp.Sleep(2000 * time.Millisecond),
				chromedp.Sleep(time.Duration(sc.ContentDelayTime) * time.Millisecond),
			})
			if err != nil {
				qu.Debug("当前列表页第" + fmt.Sprint(contentIndex+1) + "条详情页打开异常")
				errMsg = append(errMsg, fmt.Sprintf("%s%d%s%d%s%s", "第", j+1, "页，第", contentIndex+1, "条打开详情页失败：", err.Error()))
				continue
			}
			//获取详情页内容
			err = chromedp.Run(ctx2, chromedp.Tasks{
				chromedp.Evaluate(contentRunJs, r),
			})
			if err != nil {
				qu.Debug("当前列表页第" + fmt.Sprint(contentIndex+1) + "条详情页内容获取失败")
				errMsg = append(errMsg, fmt.Sprintf("%s%d%s%d%s%s", "第", j+1, "页，第", contentIndex+1, "条详情页内容获取失败：", err.Error()))
				continue
			}
			//下载附件
			if sc.AttachCss != "" {
				downloadAttaches(r, vm.attachesDir)
			}
			r.Site = sc.Site
			r.Channel = sc.Channel
			if r.Title == "" {
				r.Title = r.ListTitle
			}
			if r.PublishTime == "" {
				r.PublishTime = r.ListPubTime
			}
			r.No = no
			no += 1
			//结果放入缓存
			verifyResult.PushBack(r)
		}
		qu.Debug("第"+fmt.Sprint(j+1)+"页校验成功数据条数：", verifyResult.Len())
		//翻页
		if verifyResult.Len() > 0 {
			if sc.MaxPages > 1 && j < VERIVY_MAX_TRUN_PAGE-1 && j < int(sc.MaxPages)-1 { //&& !ret.ListTrunPage {
				if err = trunPage(sc, sc.ListTurnDelayTime, ctx); err != nil { //翻页失败
					qu.Debug("第" + fmt.Sprint(j+1) + "页翻页失败")
					errMsg = append(errMsg, fmt.Sprintf("%s%d%s%s", "第", j+1, "页翻页失败：", err.Error()))
					ret.ListTrunPage = false
					break T
				}
			}
		} else {
			ret.ListTrunPage = false
			break T
		}
	}
	//检查
	for el := verifyResult.Front(); el != nil; el = el.Next() {
		r, _ := el.Value.(*be.ResultItem)
		ret.Title = r.Title != ""
		ret.PublishUnit = r.PublishUnit != ""
		ret.PublishTime = r.PublishTime != "" && Reg_Date.MatchString(r.PublishTime)
		ret.Content = r.Content != ""
		ret.Attaches = len(r.AttachLinks) > 0
	}
	qu.Debug(verifyResult.Len())
	ret.ListItems = (sc.MaxPages == 1 && verifyResult.Len() > 0) || (sc.MaxPages > 1 && verifyResult.Len() > 2)
	return ret, nil, errMsg
}

// verifySpiderConfig4MainSite 只验证列表标注
// 重点网站测试环境
func (vm *VM) verifySpiderConfig4MainSite(sc *be.SpiderConfig) (*be.SpiderConfigVerifyResult, error, []string) {
	qu.Debug("sc---", *sc)
	var errMsg []string
	verifyResult := list.New()
	be.DataResults[sc.Code] = verifyResult
	ret := &be.SpiderConfigVerifyResult{false, true, false, true, true, true, false}
	_, baseCancelFn, _, _, ctx, incCancelFn := be.NewBrowser(false, true, false, sc.Href, sc.FilterResource) //列表页使用
	defer func() {
		incCancelFn()
		baseCancelFn()
	}()

	listRunJs, contentRunJs := sc.ListJSCode, sc.ContentJSCode
	//2. 执行JS代码，获取列表页信息
	if be.RegSpace.ReplaceAllString(listRunJs, "") == "" {
		listRunJs = renderJavascriptCoder(loadListItemsJS, sc)
	}
	if be.RegSpace.ReplaceAllString(contentRunJs, "") == "" {
		contentRunJs = renderJavascriptCoder(loadContentJS, sc)
	}
	qu.Debug("列表页JS:", listRunJs)
	//3.打开列表，获取条目清单
	chromedp.Run(ctx, chromedp.Tasks{
		chromedp.Navigate(sc.Href),
		chromedp.WaitReady("document.body", chromedp.ByJSPath),
		//chromedp.Sleep(1000 * time.Millisecond),
		chromedp.Sleep(time.Duration(sc.ListDelayTime) * time.Millisecond),
	})
	//4.初始化列表页信息
	if !vm.InitListPage(ctx, sc) {
		qu.Debug("初始化列表页失败，退出")
		return ret, errors.New("初始化列表页失败"), errMsg
	}
	no := 1
	ret.ListTrunPage = true
T:
	for j := 0; j < VERIVY_MAX_TRUN_PAGE && j < int(sc.MaxPages); j++ { //最多检查2页
		qu.Debug("开始检查第" + fmt.Sprint(j+1) + "页...")
		listResult := make(be.ResultItems, 0)
		err := chromedp.Run(ctx, chromedp.Tasks{
			chromedp.Evaluate(listRunJs, &listResult),
		})
		if err != nil {
			qu.Debug("执行列表页JS代码失败", err.Error())
			errMsg = append(errMsg, fmt.Sprintf("%s%d%s%s", "第", j+1, "页列表数据获取失败：", err.Error()))
			continue
		}
		//5.操作详情页
		qu.Debug("列表采集条数：", len(listResult))
		for contentIndex, r := range listResult {
			if contentIndex > 1 { //每页只校验2条
				break
			}
			qu.Debug("当前列表页第" + fmt.Sprint(contentIndex+1) + "条")
			r.Site = sc.Site
			r.Channel = sc.Channel
			//qu.Debug(r.Title, r.ListTitle)
			if r.Title == "" {
				r.Title = r.ListTitle
			}
			//qu.Debug(r.PublishTime, r.ListPubTime)
			if r.PublishTime == "" {
				r.PublishTime = r.ListPubTime
			}
			r.No = no
			no += 1
			//结果放入缓存
			verifyResult.PushBack(r)
		}
		qu.Debug("列表采集条数结果：", verifyResult.Len())
		//6.翻页
		if verifyResult.Len() > 0 {
			if sc.MaxPages > 1 && j < VERIVY_MAX_TRUN_PAGE-1 && j < int(sc.MaxPages)-1 { //&& !ret.ListTrunPage {
				if err = trunPage(sc, sc.ListTurnDelayTime, ctx); err != nil { //翻页失败
					qu.Debug("第" + fmt.Sprint(j+1) + "页翻页失败")
					errMsg = append(errMsg, fmt.Sprintf("%s%d%s%s", "第", j+1, "页翻页失败：", err.Error()))
					ret.ListTrunPage = false
					break T
				}
			}
		} else {
			ret.ListTrunPage = false
			break T
		}
	}
	//检查
	for el := verifyResult.Front(); el != nil; el = el.Next() {
		r, _ := el.Value.(*be.ResultItem)
		ret.Title = r.Title != ""
		qu.Debug("Check Title:", ret.Title, r.Title, r.ListTitle)

		ret.PublishTime = r.PublishTime != "" && Reg_Date.MatchString(r.PublishTime)
		qu.Debug("Check PublishTime:", ret.PublishTime, r.PublishTime, r.ListPubTime)
	}
	if ret.ListItems {
		ret.ListItems = (sc.MaxPages == 1 && verifyResult.Len() > 0) || (sc.MaxPages > 1 && verifyResult.Len() > 2)
	}

	return ret, nil, errMsg
}