查找单词匹配，允许在R中最多分隔n个单词

Question

希望在word2之前匹配word1，允许在word1和word2之间最多分隔5个单词。例如，如果word1是苹果，word2是芒果，那么模式应该匹配“apple是像芒果一样的水果”，但不匹配“芒果是像苹果一样的水果”(word2在word1之前)或“苹果和橙子是水果，像芒果”(长度超过5个单词)。python中的正则表达式示例是\b(?:word1\W+(?:\w+\W+){0,5}?word2|word2\W+(?:\w+\W+){0,5}?word1)\b。在R中识别此模式的类似模式和功能是什么？

Answer 1

#DATA
word1 = "apple"
word2 = "mango"
p1 = "apple is a fruit like mango"
p2 = "apple and orange are fruits, like mango"
p3 = "mango is a fruit like apple"

#FUNCTION
foo = function(word1, word2, string){
    ind2 = unlist(gregexpr(word2, string))[1]
    ind1 = unlist(gregexpr(word1, string))[1] 
    nwords = length(unlist(gregexpr(" ", substr(string, ind1, ind2))))
    if(ind2 > ind1 & nwords <= 5){
        substr(string, ind1, ind2 + nchar(word2))
    }else{
        NA
    }
}

#USAGE
foo(word1, word2, p1)
#[1] "apple is a fruit like mango"

foo(word1, word2, p2)
#[1] NA

foo(word1, word2, p3)
#[1] NA

Answer 2

这个可以用。将第一个单词计数为apple，此正则表达式搜索下一个4个单词，如果在定义的单词限制中找到芒果，则进行匹配。

library(stringr)
> stri <- c('apple is a fruit like mango','apple and orange are fruits, like mango','apple is not a fruit like orange or mango')
> stri_extract_all(str = stri, regex = 'apple(\\s\\w+){1,4}?.mango')

[[1]]
[1] "apple is a fruit like mango"

[[2]]
[1] NA

[[3]]
[1] NA