如何向量化stringr::str_count的pattern参数以正确统计匹配次数？

原代码尝试向量化stringr::str_count的pattern参数，统计指定子串在目标字符串中的出现次数，但实际结果与预期不符：

library(stringr)

# Define the patterns you want to count
patterns <- c("apple", "banana", "orange")

# Create a vectorized version of str_count
vectorized_str_count <- Vectorize(str_count, vectorize.args = "pattern")

# Input string
string <- "I like apples, bananas, and oranges. Apples are my favorite."

# Count the occurrences of patterns in the string
counts <- vectorized_str_count(string, pattern = patterns)

# Print the counts
print(counts)

预期输出为2, 1, 1，但实际返回1, 1, 1——核心原因是str_count默认区分大小写，原字符串中首字母大写的Apples无法被小写的apple模式匹配，且apples（复数）包含apple子串，仅被统计1次。

调整方案（无需构造正则） #

方法1：修改向量化函数，添加忽略大小写参数 #

直接在Vectorize包装的逻辑中指定ignore.case = TRUE，让匹配不区分大小写：

library(stringr)

patterns <- c("apple", "banana", "orange")
string <- "I like apples, bananas, and oranges. Apples are my favorite."

# 向量化时嵌入忽略大小写的匹配逻辑
vectorized_str_count <- Vectorize(function(str, pat) {
  str_count(str, pattern = pat, ignore.case = TRUE)
}, vectorize.args = "pat")

counts <- vectorized_str_count(string, pat = patterns)
print(counts)
# 输出：apple banana orange 
#          2      1      1 

方法2：用purrr包实现更简洁的向量化（推荐） #

Vectorize本质是循环语法糖，用purrr::map_int可更直观地遍历模式并统计：

library(stringr)
library(purrr)

patterns <- c("apple", "banana", "orange")
string <- "I like apples, bananas, and oranges. Apples are my favorite."

counts <- map_int(patterns, ~str_count(string, pattern = .x, ignore.case = TRUE))
names(counts) <- patterns
print(counts)
# 输出：apple banana orange 
#          2      1      1 

内容的提问来源于stack exchange，提问作者Fat Pat