如何用正则表达式检测标题前置编号？正则匹配问题排查

Let's break down what's wrong with your current regex and fix it step by step to match prefixes like II -, IV., a), etc.

What's Wrong with Your Original Pattern #

1. Incorrect character group syntax
   You used curly braces {} instead of square brackets [] to define sets of possible characters. Curly braces are for quantifiers (e.g., {1,3} to match 1-3 instances), not for listing characters like spaces or hyphens. This broke how the regex parses your intended separators.

2. Broken Roman numeral matching
   Your pattern fragments like IX? only match I or IX, and VI{0,3}? is a messy structure that can't properly match valid Roman numerals like II, III, or VI. It's why you're only getting a single I instead of the full II.

3. Disjointed separator logic
   Mixing numeral patterns with separators in an unstructured way caused the regex to stop at the first matching character instead of capturing the full numeral plus separator.

Corrected Regex #

Here's a refined pattern that handles the Roman numeral cases you're targeting, plus common separators:

import re

# Matches Roman numerals (I-IX) + optional valid separators at string start
pattern = r'^(I{1,3}|IV|V|VI{1,3}|IX)(?:[\s\-.,)]|\-\s)?'

# Test with your input
m = re.search(pattern, "II - Informations supplementaires")
if m:
    print(m.group(0))  # Output: "II -" (full matched prefix)
    print(m.group(1))  # Output: "II" (just the numeral)

How This Works #

• Roman numeral capture: ^(I{1,3}|IV|V|VI{1,3}|IX)
  This matches all valid single-digit Roman numerals at the start of the string:

  • I{1,3}: Matches I, II, III
  • IV: Matches IV
  • V: Matches V
  • VI{1,3}: Matches VI, VII, VIII
  • IX: Matches IX

• Optional separator: (?:[\s\-.,)]|\-\s)?
  This non-capturing group handles the various separators you listed:

  • [\s\-.,)]: Matches a single space, hyphen, dot, comma, or closing parenthesis
  • |\-\s: Also matches a hyphen followed by a space (like your II - example)
  • ?: Makes the separator optional, so it will match just the numeral if no separator exists

Extending to Letter-Based Prefixes #

If you also need to match letter-based prefixes like a), C., or C-, expand the pattern to include letter matches:

pattern = r'^([A-Za-z]|I{1,3}|IV|V|VI{1,3}|IX)(?:[\s\-.,)]|\-\s)?'

# Test with a letter prefix
m = re.search(pattern, "C- Informations supplementaires")
print(m.group(0))  # Output: "C-"

内容的提问来源于stack exchange，提问作者dada