R语言双条件填充new列：字段提取正则匹配问题

Let's tackle this problem step by step. You've already nailed the first part (using values from column b when available), so we just need to fix the regex extraction for when b is empty.

The Core Issue with Your Regex #

Your previous attempts didn't work because:

• The first gsub pattern targets underscores and dot-separated endings, which don't align with your actual data structure.
• The second sub looks for TASK= (with an equals sign), but your data uses TASK (a space) instead—so it never finds a valid match.

Correct Approach #

We'll use ifelse() to handle the two cases, and a precise regex to pull the content between TASK and the next space when b is empty.

Base R Solution

Here's the complete code that delivers your desired output:

# Your original data
a <- c("11-010 Bla", "TASK 21 MMM", "TASK 03-11-11 Hah")
b <- c("11-010","","")
df <- data.frame(a, b, new = "")

# Populate the new column
df$new <- ifelse(
  df$b != "",  # Check if column b has a value
  df$b,        # Use b's value if yes
  sub(".*TASK (\\S+).*", "\\1", df$a)  # Extract from a if no
)

Let's break down the regex .*TASK (\\S+).*:

• .*: Matches any characters leading up to TASK
• (\\S+): Captures one or more non-space characters (this is the value we want)
• .*: Matches any remaining characters after the captured value
• \\1: Replaces the entire string with the captured group

Alternative with stringr (More Readable)

If you're using the stringr package, you can use a positive lookbehind for even cleaner code:

library(stringr)

df$new <- ifelse(
  df$b != "",
  df$b,
  str_extract(df$a, "(?<=TASK )\\S+")
)

The (?<=TASK ) part is a positive lookbehind—it tells str_extract to find the \\S+ (non-space characters) that come right after TASK .

Final Output #

After running either code, your df$new will be exactly what you wanted:
"11-010", "21", "03-11-11"

内容的提问来源于stack exchange，提问作者Kalenji