Strict code review that ignores style
codingReviews a diff for correctness, edge cases, and regressions — and refuses to comment on formatting.
Builds a pattern from positive examples and then states what it will wrongly match.
You have a list of strings that should match and need to know what else will.
Write a regular expression that matches the following examples.
Then, before showing the pattern:
1. List the general form you inferred, including the edge cases the examples do not cover.
2. List at least three kinds of string that would match this pattern by accident.
3. State where the pattern is the wrong tool — nested structures, escaped quotes, anything not regular.
Only then output the pattern, with a one-line explanation per group.
Examples that should match:
<positive>
Examples that should not match:
<negative>
Flavour (PCRE, JS, Python):
<flavour>
Always supply negatives. Without them the model writes a pattern that matches your examples and half of everything else.
Item 3 is the honest part: it tells you when you need a parser instead, which is the answer more often than people want to hear.
Reviews a diff for correctness, edge cases, and regressions — and refuses to comment on formatting.
Extracts or restructures code while proving the public behavior stays identical.
Writes the failing test first, then the fix, and refuses to skip the reproduction.