grep Cheatsheet

Top patterns.

Overview

grep is the line-matching tool that anchors Unix text search. Five primitive surfaces cover almost every operational use: regex pattern matching, recursive directory traversal, surrounding context, inverted matches, and word/line anchors for precision. rg (ripgrep) and ag (the silver searcher) are faster modern alternatives on large trees; the BSD/GNU grep on every Unix box covers everywhere else.

• Pattern matching. Lines matching a regex. The filter that turns logs into investigation data.
• Recursive search. -r walks directory trees. Finds occurrences across an entire codebase.
• Context flags. -A, -B, -C show surrounding lines. Matches become understandable in context.
• Inverted match plus anchors. -v for exclusion patterns; -w for whole word, -x for whole line.

The approach

Five idioms cover almost every operational use of grep. Memorising them moves the team from copy-pasted commands to confident text investigation.

• grep -i pattern. Case-insensitive search. Catches the mixed-case log lines that exact-match misses.
• grep -r --include="*.py" pattern. Recursive with file glob. Limits search scope to relevant files.
• grep -A 5 pattern. Five lines of trailing context. Stack traces and surrounding log lines become readable.
• grep -E "p1|p2" plus rg. Extended regex with alternation; rg or ag on large trees for speed.

Why this compounds

Each grep one-liner replaces a manual scroll through a log file. The team’s text-search fluency deepens; regex skills transfer to sed, awk, log queries, and code-aware tooling.

• Faster investigation. One-liners replace full-file scrolling. Root cause arrives faster.
• Composable filters. grep in pipelines produces precise multi-stage filtering.
• Universal availability. Present on every Unix box. No installation, no dependency.
• Year-one investment, year-two habit. First year builds fluency. By year two, grep is the default for any text-search problem.