regex - 如何在R中仅grep 10场比赛？

Question

我有一个数据文件，我可以在其中搜索“你”:

> dtq_ml_wuv[grep("you", dtq_ml_wuv$rn), "rn"]

"you can take"         "you can tell"         "you can thank"       
"you can try"          "you can turn"         "you can use"         
"you can visit"        "you can work"         "you donet know"      
"you donet need"       "you dont know"        "you get enough"      
"you get see"          "you go back"          "you got keep"        
"you guys can"         "you heard right"      "you just go"         
"you just gotta"       "you just look"        "you just need"       
"you just stay"        "you know better"      "you know else"       
"you know got"         "you know i"           "you know if"         
"you know im"          "you know it"          "you know just"       
"you know many"        "you know means"       "you know one"        
"you know really"      "you know right"       "you like see"    

我怎样才能得到 grep至 停止找到后，说 0 到 25 个匹配项？

我试过

> dtq_ml_wuv[grep("you{0, 25}", dtq_ml_wuv$rn), "rn"]

但它告诉我表达式无效，因为 {} 的内容无效.

任何提示表示赞赏。

Answer 1

这是一个分块版本。调整块大小以充分利用内部编译代码的速度，而不是搜索太多。

grepn<-function(pattern,x,n,chunk.size=32) {
  N<-length(x);
  chunk<-1:chunk.size;
  k<-1;
  M<-vector("integer",n+chunk.size);
  while(k < n && chunk[1] <= N) {
    i<-na.omit(grep(pattern,x[chunk]));
    if(length(i)) M[k:(k+length(i)-1)]<-i+chunk[1]-1
    k<-k+length(i)
    chunk<-chunk+chunk.size
  }
  return(M[1:(min(k-1,n))])
}

示例与 data作为问题中的字符向量

grepn("e",data,n=10,chunk.size=16)
[1]  1  2  6  9 10 12 13 15 17 21