如何匹配一列中的多个字符串与另一列中的多个字符串--删除R中的匹配？

Question

这是我的代码：

A <- c("ruler measure", "measure rulers", "rulers")
B <- c("you can measure things with rulers", "you can measure things with rulers", "you can measure things with rulers")
df <- data.frame(as.character(A), as.character(B))

df_new <- df %>%
  mutate(
    new_B = str_replace_all(B, A, "")
  )

我想让这些列看起来像这样：

A                          B   
ruler measure             you can things with
measure rulers            you can things with
rulers                    you can measures things with  

但是，str_replace_all()似乎只替换了A和B的一个匹配(例如，标尺)，而不是另一个匹配(例如度量)。

谢谢你的帮助！！

Answer 1

这里有一个基本的R解

df <- within(df,
             new_B <- mapply(gsub, 
                             sapply(strsplit(as.character(A),"\\s+"),
                                    function(v) paste0(paste0("\\s+?",v,".*?\\b"),collapse = "|")),
                             "",
                             B))

这样的话

> df
               A                                   B                       new_B
1  ruler measure  you can measure things with rulers         you can things with
2 measure rulers you can measures things with rulers         you can things with
3         rulers  you can measure things with rulers you can measure things with

Answer 2

我们可以用|替换这个空间

library(dplyr)
library(stringr)
df %>% 
   mutate(new_B = str_replace_all(B, str_replace(A, " ", "|"), ''))