Question

我试图在this page中获取选择器列表：

$("#Lastname"),$(".intro"),....

我尝试使用xpathSApply：

library(XML)
library(RCurl)
a <- getURL('http://www.w3schools.com/jquery/trysel.asp')
doc <- htmlParse(a)
xpathSApply(doc,'//*[@id="selectorOptions"]') ## I can't get the right xpath

我也试过但没有成功：

xpathSApply(doc,'//*[@id="selectorOptions"]/div[i]')

编辑我添加了python标签，因为我也接受了python解决方案。

Answer 1

以下是获取此类javascript页面的R方式。您需要使用@Peyton所述的浏览器。 Selenium服务器是控制浏览器的一种好方法。我为硒服务器写了一些R的绑定 https://github.com/johndharrison/RSelenium

以下内容允许用户访问帖子javascript来源：

require(devtools)
devtools::install_github("RSelenium", "johndharrison")
library(RSelenium)
library(RJSONIO)

# one needs to have an active server running
# the following commented out lines source the latest java binary
# RSelenium::checkForServer()
# RSelenium::startServer()
# a selenium server is assummed to be running now

remDR <- remoteDriver$new()
remDR$open() # opens a browser usually firefox with default settings
remDR$navigate('http://www.w3schools.com/jquery/trysel.asp') # navigate to your page
webElem <- remDR$findElements(value = "//*[@id='selectorOptions']") # find your elememts

# display the appropriate quantities
cat(fromJSON(webElem[[1]]$getElementText())$value)
> cat(fromJSON(webElem[[1]]$getElementText())$value)
$("#Lastname")
$(".intro")
$(".intro, #Lastname")
$("h1")
$("h1, p")
$("p:first")
$("p:last")
$("tr:even")
$("tr:odd")
$("p:first-child")
$("p:first-of-type")
$("p:last-child")
$("p:last-of-typ
.....................

更新：

在这种情况下访问信息的一种更简单的方法是使用executeScript方法

library(RSelenium)
RSelenium:startServer()
remDr$open()
remDR$navigate('http://www.w3schools.com/jquery/trysel.asp')
remDr$executeScript("return w3Sels;")[[1]]

> remDr$executeScript("return w3Sels;")[[1]]
 [1] "#Lastname"              ".intro"                
 [3] ".intro, #Lastname"      "h1"                    
 [5] "h1, p"                  "p:first"               
 [7] "p:last"                 "tr:even"               
 [9] "tr:odd"                 "p:first-child"         
[11] "p:first-of-type"        "p:last-child"          
[13] "p:last-of-type"         "li:nth-child(1)"       
[15] "li:nth-last-child(1)"   "li:nth-of-type(2)"     
[17] "li:nth-last-of-type(2)" "b:only-child"          
[19] "h3:only-of-type"        "div > p"               
[21] "div p"                  "ul + h3"               
[23] "ul ~ table"             "ul li:eq(0)"           
[25] "ul li:gt(0)"            "ul li:lt(2)"           
[27] ":header"                ":header:not(h1)"       
[29] ":animated"              ":focus"                
[31] ":contains(Duck)"        "div:has(p)"            
[33] ":empty"                 ":parent"               
[35] "p:hidden"               "table:visible"         
[37] ":root"                  "p:lang(it)"            
[39] "[id]"                   "[id=my-Address]"       
[41] "p[id!=my-Address]"      "[id$=ess]"             
[43] "[id|=my]"               "[id^=L]"               
[45] "[title~=beautiful]"     "[id*=s]"               
[47] ":input"                 ":text"                 
[49] ":password"              ":radio"                
[51] ":checkbox"              ":submit"               
[53] ":reset"                 ":button"               
[55] ":image"                 ":file"                 
[57] ":enabled"               ":disabled"             
[59] ":selected"              ":checked"              
[61] "*"

Answer 2

感谢jdharrison评论我解析了javascript代码以提取所有选择器。正如Peyton所提到的，这适用于这种特殊情况，因为所有选择器都在代码中。

capture.output(xpathSApply(doc,'//*/script')[[6]],
               file='test.js')
ll <- readLines('test.js')
ll <- ll[grepl('w3Sels.push',ll)]
ll <- unlist(regmatches(ll, gregexpr("(?<=\\().*?(?=\\))", ll, perl=T)))

 cat(head(ll))
"#Lastname" ".intro" ".intro, #Lastname" "h1" "h1, p" "p:first"

刮掉这个网页的xpath是什么？

2 个答案: