Practical — Logs and Formatting

Putting groups + alternation + replacement together

You now have everything for production-grade text transformation. Let's combine it on three real tasks.

Task 1: Normalize log timestamps

Logs from different sources use different timestamp formats: 2026-05-04T14:32:11Z, 04/May/2026 14:32:11, 14:32:11 04 May 2026. Pick one canonical format and convert all to it.

Strategy: write three patterns matching each input format with named groups, then sub each one to the canonical format with backreferences.

Task 2: CSV with awkward quoting

Real CSV is messy: Pippa,"3, going on 30",warm. The middle field has a comma INSIDE quotes. Pattern that handles both quoted and bare: (?:"([^"]*)"|([^,]*))(?:,|$). Capture each field with finditer.

Task 3: Markdown to HTML conversion (subset)

Convert **bold** to <strong>bold</strong>:

re.sub(r'\*\*([^*]+)\*\*', r'<strong>\1</strong>', text)

One line per Markdown construct. Real Markdown parsers exist for production; for ad-hoc rewrites, regex is faster.

The combined principle

Describe the input shape with named groups.
Use non-capturing groups for scoping.
Use alternation for known variants.
Use backreferences in replacement to reorder.
Test with at least one adversarial input per pattern.

Five steps. They scale from one-line transforms to full log pipelines.

Code

Three real transformations·python

import re

# 1. Normalize ISO date to MM/DD/YYYY
re.sub(
    r'(?P<y>\d{4})-(?P<m>\d{2})-(?P<d>\d{2})',
    r'\g<m>/\g<d>/\g<y>',
    'logged 2026-05-04'
)
# 'logged 05/04/2026'

# 2. CSV field extraction with quoting
def parse_csv_line(line):
    pattern = r'(?:"([^"]*)"|([^,]*))(?:,|$)'
    fields = []
    for m in re.finditer(pattern, line):
        quoted, bare = m.groups()
        fields.append(quoted if quoted is not None else bare)
        if not m.group():  # avoid infinite loop on empty
            break
    return fields

parse_csv_line('Pippa,"3, going on 30",warm')
# ['Pippa', '3, going on 30', 'warm', '']

# 3. Markdown bold → HTML
re.sub(r'\*\*([^*]+)\*\*', r'<strong>\1</strong>', 'this is **bold** text')
# 'this is <strong>bold</strong> text'

Practical — Logs and Formatting

Putting groups + alternation + replacement together

Task 1: Normalize log timestamps

Task 2: CSV with awkward quoting

Task 3: Markdown to HTML conversion (subset)

The combined principle

Code

External links

Exercise

Progress

Comments 0