Google通过R中的URL进行翻译而无需API

Question

我正在尝试将多个字符串从本地语言（“泰米尔语”）翻译成英语。我使用了Stackoverflow的示例代码并修改了脚本：

   library(RCurl)
   library(XML)
   library(Unicode)
   getParam = read.csv("file.csv", header=T, encoding = "UTF-8")
   get<-u_to_lower_case(getParam$text)
   get1<-matrix(1:6,nrow = 6, ncol = 1)
   get1<-(as.vector(get))
   data2<-matrix(1:6,nrow=6, ncol = 1)
   translateFrom = "ta"
   translateTo = "en"
   for (i in 1:length(get1))
   {
   URL <- paste("https://translate.google.pl? 
   hl=",translateTo,"&sl=",translateFrom,"&tl=",translateTo,"&ie=UTF- 
   8&prev=_m&q=",get1[i],sep="")
   print(URL)
   page <- getURL(URL)
   tree <-htmlTreeParse(page)
   body <- tree$children$html$children$body 
   body_text <- body$children[[5]]$children[[1]]  
   print(body_text)
   data1[i]<-body_text
   if(nrow(data1) == 0) 
   {data1[1,] <- "N/A"
   } 
   data2<-rbind(data2,data1)
   }

没有循环，我可以翻译它。但是，在循环中，我得到了这样的错误：

    In structure(x$children, class = "XMLNodeList") :
    Calling 'structure(NULL, *)' is deprecated, as NULL cannot have 
    attributes.
    Consider 'structure(list(), *)' instead.

有人可以更正我的代码或提出更好的代码吗？

注意：为了避免陷入Unicode错误，进行了多次修改。但是，终于克服了它。 另外，我不想使用付费的Google翻译API。