Question

比方說，我用下面的表達

library(XML) 
url.df_1 = htmlTreeParse("http://www.appannie.com/app/android/com.king.candycrushsaga/", useInternalNodes = T) 

如果我運行下面的代碼解析了一個網站，

xpathSApply(url.df_1, "//div[@class='app_content_section']/h3", function(x) c(xmlValue(x), xmlAttrs(x)[["href"]])) 

我會得到如下 -

[1] "Description"      "What's new"      
[3] "Permissions"      "More Apps by King.com All Apps »" 
[5] "Customers Also Viewed"   "Customers Also Installed"  

現在，我感興趣的只是「客戶還安裝」部分。但是，當我運行下面的代碼，

xpathSApply(url.df_1, "//div[@class='app_content_section']/ul/li/a", function(x) c(xmlValue(x), xmlAttrs(x)[["href"]])) 

它吐出全部列入「由King.com所有應用更多應用程序的」應用程序「客戶還看」和「客戶還安裝」。

所以，我想，

xpathSApply(url.df_1, "//div[h3='Customers Also Installed']」, function(x) c(xmlValue(x), xmlAttrs(x)[["href"]])) 

但這並沒有工作。所以我試了

xpathSApply(url.df_1, "//div[contains(.,'Customers Also Installed')]",xmlValue) 

但是這也行不通。 （輸出應該像下面這樣）

[,1]             
[1,] "Christmas Candy Free\n Daniel Development\n " 
[2,] "/app/android/xmas.candy.free/"      
[,2]           
[1,] "Jewel Candy Maker\n Nutty Apps\n "  
[2,] "/app/android/com.candy.maker.jewel.nuttyapps/" 
[,3]          
[1,] "Pogz 2\n Terry Paton\n "   
[2,] "/app/android/com.terrypaton.unity.pogz2/" 

任何指導將非常感謝！

Answer 1

這裏是一個選項（你真的很接近）：

xpathSApply(url.df_1,"//div[contains(.,'Customers Also Installed')]/*/li/a",xmlGetAttr,'href') 

[1] "/app/android/xmas.candy.free/"     
[2] "/app/android/com.candy.maker.jewel.nuttyapps/" 
[3] "/app/android/com.terrypaton.unity.pogz2/"