如何删除R中字符串中不同位置以特定字符(@)开头的任何内容？

Question

我有一个包含多条tweet的专栏：

ID | Tweet
1    @ChipotleTweets @ChipotleTweets Becky is very nice
2    Happy Halloween! I now look forward to $3 booritos at @ChipotleTweets
3    Considering walking to @.ChipotleTweets in my llama onesie.

这样做的目的是删除“@___”和@之后的所有内容，但不删除该字符串之外的空格文本。

目前正在使用这段代码来检测'@‘，但是如果它不在句子的第一个位置，我就不会发现任何东西

tweet_pattern <- " @\\w+"

Customer <- Customer %>% 
           clean_Tweet = ifelse(str_detect(text, tweet_pattern), 
                                str_remove(text, tweet_pattern), 
                                NA_character_))

所需输出：

ID | Tweet                                                                  | cleaned_tweet 
1    @ChipotleTweets @ChipotleTweets Becky is very nice                       Becky is very nice
2    Happy Halloween! I now look forward to $3 booritos at @ChipotleTweets    Happy Halloween! I now look forward to $3 booritos at
3    Considering walking to @.ChipotleTweets in my llama onesie.              Considering walking to in my llama onesie.

Answer 1

我们可以更改模式以匹配零个或多个空格(\\s*)，后跟@和str_remove_all中的一个或多个非空格(\\S+)，以删除这些子字符串

library(stringr)
library(dplyr)
Customer %>%
     mutate(Cleaned_Tweet = str_remove_all(Tweet, "\\s*@\\S+"))

-output

 ID                                                                 Tweet                                         Cleaned_Tweet
1  1                    @ChipotleTweets @ChipotleTweets Becky is very nice                                    Becky is very nice
2  2 Happy Halloween! I now look forward to $3 booritos at @ChipotleTweets Happy Halloween! I now look forward to $3 booritos at
3  3           Considering walking to @.ChipotleTweets in my llama onesie.            Considering walking to in my llama onesie.

注意：str_remove只删除了match的第一个实例，也就是说，如果一个字符串中有多个匹配，它会跳过其他的，只匹配第一个。我们需要str_remove_all来删除匹配模式的所有实例。

数据

Customer <- structure(list(ID = 1:3, Tweet = c("@ChipotleTweets @ChipotleTweets Becky is very nice", 
"Happy Halloween! I now look forward to $3 booritos at @ChipotleTweets", 
"Considering walking to @.ChipotleTweets in my llama onesie."
)), class = "data.frame", row.names = c(NA, -3L))