Text Processing Skill

Master text manipulation with grep, sed, awk, and regular expressions

Learning Objectives

After completing this skill, you will be able to:

• Search files efficiently with grep and ripgrep
• Transform text with sed substitutions
• Process structured data with awk
• Write and debug regular expressions
• Build efficient text processing pipelines

Prerequisites

• Bash basics (variables, control flow)
• Command line navigation
• Understanding of stdin/stdout

Core Concepts

1. Grep Essentials

# Basic search
grep 'pattern' file.txt
grep -i 'pattern' file.txt      # Case insensitive
grep -v 'pattern' file.txt      # Invert match
grep -n 'pattern' file.txt      # Line numbers
grep -c 'pattern' file.txt      # Count only

# Extended regex
grep -E 'pat1|pat2' file.txt
grep -E '^start.*end$' file.txt

# Recursive search
grep -r 'pattern' ./
grep -rn --include='*.py' 'def ' ./

2. Sed Essentials

# Substitution
sed 's/old/new/' file           # First match
sed 's/old/new/g' file          # All matches
sed -i 's/old/new/g' file       # In-place

# Line operations
sed -n '5p' file                # Print line 5
sed '5d' file                   # Delete line 5
sed '/pattern/d' file           # Delete matching

# Multiple operations
sed -e 's/a/b/' -e 's/c/d/' file

3. Awk Essentials

# Field processing
awk '{print $1}' file           # First field
awk -F: '{print $1}' file       # Custom delimiter
awk '{print $NF}' file          # Last field

# Patterns
awk '/pattern/' file            # Match lines
awk '$3 > 100' file             # Condition

# Calculations
awk '{sum+=$1} END{print sum}' file
awk 'NR>1 {total++} END{print total}' file

4. Regex Quick Reference

# Metacharacters
.     # Any character
^     # Start of line
$     # End of line
*     # Zero or more
+     # One or more (ERE)
?     # Zero or one (ERE)

# Character classes
[abc]     # Any of a, b, c
[^abc]    # Not a, b, c
[a-z]     # Range
\d        # Digit (PCRE)
\w        # Word char (PCRE)
\s        # Whitespace (PCRE)

Common Patterns

Log Analysis

# Count requests by IP
awk '{print $1}' access.log | sort | uniq -c | sort -rn

# Find errors
grep -E 'ERROR|FATAL' app.log | tail -20

# Extract timestamps
grep 'ERROR' app.log | sed 's/.*\[\([^]]*\)\].*/\1/'

Data Transformation

# CSV to TSV
sed 's/,/\t/g' data.csv

# JSON value extraction
grep -oP '"name":\s*"\K[^"]+' data.json

# Remove blank lines
sed '/^$/d' file.txt

Anti-Patterns

Don't	Do	Why
cat file | grep	grep pattern file	Useless use of cat
Multiple sed calls	Single sed with -e	Reduces overhead
grep -E ".*"	Omit if not needed	Slower with regex

Practice Exercises

1. Log Parser: Extract top 10 IPs from access log
2. CSV Filter: Filter CSV rows by column value
3. Config Editor: Update config values with sed
4. Report Generator: Summarize data with awk

Troubleshooting

Common Errors

Error	Cause	Fix
Invalid regex	Bad pattern	Escape special chars
No match	Wrong case	Use -i flag
sed delimiter	/ in pattern	Use # or |

Debug Techniques

# Test regex online
# https://regex101.com/

# Print matched groups
echo "test" | sed -n 's/\(.*\)/\1/p'

# Debug awk
awk '{print NR, NF, $0}' file

Performance Tips

# Use ripgrep for speed
rg 'pattern' --type py

# Set locale for speed
LC_ALL=C grep 'pattern' file

# Limit output
grep -m 10 'pattern' file

Resources

• GNU Grep Manual
• Sed One-Liners
• AWK Tutorial
• Regex101