获取 URL 目录的自定义函数

Question

看起来很简单，

考虑以下 URLs，

[1] "scripts.iucr.org/cgi-bin/paper?S1600536812045886"                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                        
[2] "cpa-seoadvisors.com/cvv/auth/auth/view/pdf/index.html/"                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                  
[3] "www.scirp.org/journal/PaperDownload.aspx?DOI=10.4236/csta.2012.13014"                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                    
[4] "www.google.com.cy/search?q=DNS+traffic&es_..."                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                      
[5] "seesaa.net/pede/lobortis/ligula/sit/amet.png?semper=vitae&est=..."

我想得到第一个 '/' 和用 ? 分隔令牌的部分。我写了下面的函数

get_directory <- function(x){
  dir <- sapply(strsplit(x, '/'), function(i)sum(grepl('\?', i)))
  ifelse(dir > 0, sapply(strsplit(x, '/'), function(i) paste(i[-c(1, length(i))], collapse = '/')), 0)
}

但它在 [3] 和 [4] URL 失败了。

预期输出应该是

"cgi-bin"
"0"
"journal"
"0"
"pede/lobortis/liguls/sit"

数据

dput(df)
structure(list(V1 = c("scripts.iucr.org/cgi-bin/paper?S1600536812045886", 
"cpa-seoadvisors.com/cvv/auth/auth/view/pdf/index.html/", "www.scirp.org/journal/PaperDownload.aspx?DOI=10.4236/csta.2012.13014", 
"www.google.com.cy/search?q=DNS+traffic&es_...", "seesaa.net/pede/lobortis/ligula/sit/amet.png?semper=vitae&est=..."
)), .Names = "V1", row.names = c(NA, -5L), class = "data.frame")

Answer 1

我们可以使用str_extract。使用正则表达式环视，我们匹配一个或多个字符 (.*) 在 / 后跟一个 / 和一个或多个不是 ? 的字符 ([^?]+) 后跟 ?。

library(stringr)
res <- str_extract(df$V1, "(?<=\/).*(?=\/[^?]+[?])")
res[is.na(res)] <- 0
res
#[1] "cgi-bin"                  "0"                        "journal"                 
#[4] "0"                        "pede/lobortis/ligula/sit"

获取 URL 目录的自定义函数

Custom Function for obtaining URL directory

regex

r

strsplit