Python RegEx Module

Python re Module: Compiled Patterns, Named Groups, and Flags

When a regex runs inside a tight loop or parses complex text formats, Python's re module lets you pre-compile patterns with re.compile(), add readable inline comments with re.VERBOSE, and extract fields using named capture groups.

Named groups replace brittle positional indexes like match.group(2) with self-documenting dictionary lookups like match.group("level").

Compiled Patterns and Named Groups

Compile complex patterns once at module level so Python parses the regex syntax only once rather than on every function call.

import re

LOG_PATTERN = re.compile(
    r'''
    ^(?P<date>\d{4}-\d{2}-\d{2})\s+   # ISO date
    (?P<level>INFO|WARN|ERROR)\s+     # Severity level
    (?P<message>.+)$                  # Log body
    ''',
    re.VERBOSE,
)

match = LOG_PATTERN.match("2026-11-12 ERROR Database pool exhausted")
if match:
    print(match.groupdict())

Dynamic Replacements with re.sub() Callbacks

Instead of a static replacement string, pass a function to re.sub() to compute each replacement dynamically from the Match object.

def mask_digits(m: re.Match) -> str:
    return "*" * len(m.group(0))

print(re.sub(r"\d{4}(?=\d{4})", mask_digits, "Card 411122223333"))

Common Mistakes to Avoid

• Catastrophic backtracking (ReDoS): Nested quantifiers like (a+)+$ can freeze your CPU for seconds on non-matching input; keep quantifiers unambiguous.
• Using greedy .* when you want lazy .*?: r'"(.*)"' grabs from the first quote to the last quote on the line; use r'"(.*?)"' or r'"([^"]*)"'.
• Re-compiling identical patterns inside million-row loops: Call re.compile() once outside the loop.

Frequently Asked Questions

What does re.finditer() do compared to re.findall()?

re.finditer() yields Match objects lazily one at a time, giving you start/end positions and named groups without building a large list in memory.

What does re.IGNORECASE (re.I) affect?

It makes letter matching case-insensitive ([a-z] matches A-Z). Combine multiple flags using bitwise OR: re.IGNORECASE | re.MULTILINE.

When should I avoid regex altogether?

Never parse nested HTML, XML, or JSON with regex. Use dedicated parsers like json, xml.etree.ElementTree, or BeautifulSoup.