正则表达式多结果匹配问题：按起止字符分割字符串失效

Hey there! Let's get that regex working for you so you can pull out those three segments correctly.

Why Your Original Regex Isn't Cutting It #

Your current regex (?:^|\s)format_(.*?)_end(?:\s|$) is looking for whitespace or string boundaries to bookend each segment—but your target string has no spaces. Each end directly runs into the next format, which breaks the regex's expectations:

• The first end isn't followed by whitespace or the string end, so the regex can't finish matching the first segment.
• Every subsequent format doesn't have whitespace before it, so the regex completely ignores those instances.

The Correct Regex #

Instead of relying on whitespace, we just need to match every chunk that starts with format_ and ends at the nearest _end. Here's the simple, effective regex you need:

format_.*?_end

Breakdown of How It Works #

• format_: Matches the literal start of each target segment.
• .*?: Non-greedy match of any characters (this ensures we stop at the first _end instead of skipping ahead to the last one).
• _end: Matches the literal end of each segment.

Example Code (Python) #

Let's test this with your input string:

import re

input_str = "format_abc_endformat_def_endformat_ghi_end"
extracted_segments = re.findall(r'format_.*?_end', input_str)

print(extracted_segments)
# Output: ['format_abc_end', 'format_def_end', 'format_ghi_end']

Bonus: More Precise Match (If Your Use Case Allows) #

If you know the middle part (between format_ and _end) will never contain underscores, you can use a stricter regex to avoid accidental matches:

format_[^_]*_end

This uses [^_]* to match only characters that aren't underscores, which adds an extra layer of safety for specific scenarios.

内容的提问来源于stack exchange，提问作者Erwin Vorenhout