基于RSelenium爬取中国裁判文书网文书数据

传送：RSelenium基本用法总结

          Python与R协同完成【中国裁判文书网】文书内容爬取

#以执业机构（即律师事务所）作为条件输入
system("java -jar F:/.../SeleniumSever/selenium-server-standalone-3.8.1.jar",wait=F)
system("java -Dwebdriver.chrome.driver=F:/.../SeleniumSever/chromedriver.exe",wait=F)
setwd("F:/...")
library("readxl")
library("magrittr")
library("RSelenium")
library("rvest")
library("dplyr")
#读取之前爬的律师事务所名称
law_firm_info<-read_xls("law_firm_info.xls") %>% select(c(1,4)) %>% unique()
remDr<-remoteDriver(browserName="chrome")
remDr$open()
url<-"http://wenshu.court.gov.cn/"

#函数1:审理程序+文书名称+审理法院+文书字号+审理日期
get_info_one<-function(tian){
trial_procedure<-tian %>% '[['(1) %>% read_html() %>% html_nodes(xpath="//*[@id='resultList']/div/div[1]") %>% html_text() #审理程序
document_name<-tian %>% '[['(1) %>% read_html() %>% html_nodes(xpath="//*[@id='resultList']/div/table/tbody/tr[1]/td/div/a[2]") %>% html_text() #文书名称
tmp<-tian %>% '[['(1) %>% read_html() %>% html_nodes(xpath="//*[@id='resultList']/div/table/tbody/tr[2]/td/div") %>% html_text() #审理法院+文书字号+审理日期
tmp_2<-data.frame()
for (j in seq_along(tmp)) {
tmp_1<-unlist(strsplit(tmp[j],"\u00A0\u00A0\u00A0\u00A0"))
tmp_1<-data.frame(court_name=tmp_1[1],document_number=tmp_1[2],date_trial=tmp_1[3],stringsAsFactors=F)
tmp_2<-rbind(tmp_2,tmp_1)
rm(tmp_1)
}
gg<-cbind(trial_procedure,document_name,tmp_2,stringsAsFactors=F)
return(gg)
}

#函数2:窗口切换,只允许在两个窗口间不断切换
window_switch_gettext<-function(tian) {
data_frame_huizong<-data.frame()
     for (j in seq_along(as.character(tian$document_number))) {
	   css_tmp<-paste0("#resultList > div:nth-child(",j,") > table > tbody > tr:nth-child(1) > td > div > a:nth-child(4)")
       btn<-remDr$findElement(using='css selector',value=css_tmp)
       btn$clickElement()      #此时两个窗口
       Sys.sleep(runif(1,9,15))
	   rm(css_tmp)
	   #将窗口切换至新打开的窗口-即第二个窗口-使用窗口句柄函数
	   remDr$switchToWindow(windowId=remDr$getWindowHandles()[[2]])
	   #在第二个打开窗口-获取文书的详细信息-字符串格式
	   text_tmp<-remDr$getPageSource() %>% '[['(1) %>% read_html() %>% html_nodes("#DivContent div") %>% html_text()
data_frame_tmp<-as.data.frame(cbind(as.character(tian$document_number)[j],text_tmp),stringsAsFactors=F)
       rm(text_tmp)	   
	   colnames(data_frame_tmp)<-c("document_number","text")
data_frame_tmp<-data_frame_tmp[which(nchar(as.character(data_frame_tmp$text))>1),]
	   data_frame_huizong<-rbind(data_frame_huizong,data_frame_tmp)
	   rm(data_frame_tmp)
	   #将第二个窗口关闭
	   remDr$closeWindow()
	   #将窗口切换到初始窗口
	   remDr$switchToWindow(windowId=remDr$getWindowHandles()[[1]])   
}
    return(data_frame_huizong)
}

document_basic_info<-data.frame()     #存储文书基本信息
document_text_info<-data.frame()      #存储文书详细信息

remDr$navigate(url)
btn<-remDr$findElement(using='css selector',value='#head_maxsearch_btn')  #点击高级搜索
btn$clickElement()
#定位律所文本框+输入文本
btn<-remDr$findElement(using='css selector',value='#adsearch_LS')
btn$sendKeysToElement(as.list(as.character(law_firm_info[1,1]))) 
#定位搜索按钮+点击
btn<-remDr$findElement(using='css selector',value='#list_btnmaxsearch')
btn$clickElement()
Sys.sleep(runif(1,5,10))
rm(btn)

#判断搜索结果是否有数据,如果没有数据则进入下个事务所（这个语句要写在循环里面，作为一个条件判断）
if ((remDr$getPageSource() %>% '[['(1) %>% read_html() %>% html_nodes(xpath="//*[@id='resultList']") %>% html_text())=="无符合条件的数据...") 
next

#搜索结果有数据:获取当前界面的文书基本信息
tmp_tian<-get_info_one(remDr$getPageSource())
tmp_tian$law_firm<-as.character(law_firm_info[1,1])
document_basic_info<-rbind(document_basic_info,tmp_tian)

#获取文书详细内容
document_text_tmp<-window_switch_gettext(tmp_tian)
document_text_info<-rbind(document_text_info,document_text_tmp)    
rm(tmp_tian)
rm(document_text_tmp)
gc()

#以上信息均是首页数据
#循环点击下一页爬取数据
j=1
while (1==1) {
j<-j+1
trycatch_value_2<-tryCatch(
          {
		  #定位[下一页]按钮，仍然为一个窗口
          btn<-remDr$findElement(using='css selector',value='#pageNumber > a.next');
          btn$clickElement();1+1
		  },
		  error=function(e) return(paste0(law_firm_info[1,1],"-律师事务所的第",j,"页，在点击下一页时出现错误，错误信息：",e$message))
)
          if (trycatch_value_2 != 2) {print(trycatch_value_2);break} else {
Sys.sleep(runif(1,10,17))
#判断页面是否刷新出来--当前页的页码 ?=j  
trycatch_value_3<-tryCatch(
          {
		  m<-0;
		  for (n in 1:10) {
		  m<-m+1
		  page_tmp<-remDr$getPageSource() %>% '[['(1) %>% read_html() %>% html_nodes("#pageNumber > span") %>% html_text()
		  if (as.numeric(page_tmp)==j) break else
		  {Sys.sleep(runif(1,10,17)+m);next}
		  };1+1
		  },
		  error=function(e) return(paste0(law_firm_info[1,1],"-律师事务所的第",j,"页，页面未刷新出来，错误信息：",e$message))
)
          if (trycatch_value_3 != 2) {print(trycatch_value_3);break} else {
	  
trycatch_value_1<-tryCatch(
          {
		  tmp_tian<-get_info_one(remDr$getPageSource());
          tmp_tian$law_firm<-as.character(law_firm_info[1,1]);      
          document_basic_info<-rbind(document_basic_info,tmp_tian);
		  document_text_tmp<-window_switch_gettext(tmp_tian);
          document_text_info<-rbind(document_text_info,document_text_tmp);    
          rm(tmp_tian);
          rm(document_text_tmp);
		  1+1},
          error=function(e) return(paste0(law_firm_info[1,1],"-律师事务所的第",j,"页在抓取详细信息时出现错误，错误信息：",e$message))
)
          if (trycatch_value_1 != 2) {print(trycatch_value_1);break}
}
}
}

#数据存储
write.table(document_basic_info,"document_basic_info.csv",append = F,col.names = T,row.names = F,sep=",")
write.table(document_text_info,"document_text_info.csv",append = F,col.names = T,row.names = F,sep=",")