Drill 01: Fundamentals

Before anything else, master these basics: what matches literally, what’s special, and how to escape.

Core Concepts

Concept What You Need to Know

Concept	What You Need to Know
Literal matching	Most characters match themselves: `foo` matches "foo"
Metacharacters	Special characters with regex meaning: `. * + ? ^ $ \| ( ) [ ] { } \`
Escaping	Use backslash to match metacharacters literally: `\.` matches a period
Case sensitivity	Regex is case-sensitive by default; use flags to change

Literal matching

Most characters match themselves: foo matches "foo"

Metacharacters

Special characters with regex meaning: . * + ? ^ $ | ( ) [ ] { } \

Escaping

Use backslash to match metacharacters literally: \. matches a period

Case sensitivity

Regex is case-sensitive by default; use flags to change

Metacharacter Reference

Char Meaning To Match Literally

Char	Meaning	To Match Literally
`.`	Any single character (except newline)	`\.`
`*`	Zero or more of previous	`\*`
`+`	One or more of previous	`\+`
`?`	Zero or one of previous	`\?`
`^`	Start of line/string	`\^`
`$`	End of line/string	`\$`
`\|`	Alternation (OR)	`\\|`
`(`	Group start	`\(`
`)`	Group end	`\)`
`[`	Character class start	`\[`
`]`	Character class end	`\]`
`{`	Quantifier start	`\{`
`}`	Quantifier end	`\}`
`\`	Escape character	`\\`

.

Any single character (except newline)

\.

*

Zero or more of previous

\*

+

One or more of previous

\+

?

Zero or one of previous

\?

^

Start of line/string

\^

$

End of line/string

\$

|

Alternation (OR)

\|

(

Group start

\(

)

Group end

\)

[

Character class start

\[

]

Character class end

\]

{

Quantifier start

\{

}

Quantifier end

\}

\

Escape character

\\

Interactive CLI Drill

Run the practice script:

bash ~/atelier/_bibliotheca/domus-captures/docs/modules/ROOT/examples/regex-drills/01-fundamentals.sh

Exercise Set 1: Literal Matching

cat << 'EOF' > /tmp/ex01.txt
The server responded with 200 OK
Error: Connection failed
Warning: Low disk space
server.example.com
10.50.1.20
EOF

Ex 1.1: Match "server"

Solution

grep 'server' /tmp/ex01.txt

Output:

The server responded with 200 OK
server.example.com

Ex 1.2: Case-insensitive "error"

Solution

grep -i 'error' /tmp/ex01.txt

Output:

Error: Connection failed

Ex 1.3: Match exact IP

Solution

grep '10\.50\.1\.20' /tmp/ex01.txt

Note: Escape periods to match literally, not "any character"

Exercise Set 2: The Dot Trap

cat << 'EOF' > /tmp/ex02.txt
192.168.1.100
192X168Y1Z100
10.50.1.20
version1.2.3
versionABC
EOF

Ex 2.1: Why does this match both?

grep -E '192.168' /tmp/ex02.txt

Explanation

The unescaped . matches ANY character: - 192.168 matches "192.168" (dot is any char, happens to be period) - 192.168 also matches "192X168" (dot matches X)

This is the #1 regex bug. Always escape dots in IPs, domains, versions.

Ex 2.2: Match only valid IP format

Solution

grep -E '192\.168\.1\.100' /tmp/ex02.txt

Exercise Set 3: Escaping Special Characters

cat << 'EOF' > /tmp/ex03.txt
[ERROR] Database offline
[WARN] Low memory
Price: $99.99
Path: C:\Users\Admin
What? Really? Yes!
2+2=4
5*5=25
EOF

Ex 3.1: Match [ERROR] including brackets

Solution

grep -E '\[ERROR\]' /tmp/ex03.txt

Ex 3.2: Match the $99.99 price

Solution

grep -E '\$99\.99' /tmp/ex03.txt

Ex 3.3: Match Windows path

Solution

grep -E 'C:\\Users' /tmp/ex03.txt

Note: In shell, \\ becomes one backslash passed to grep

Ex 3.4: Match "2+2=4"

Solution

grep -E '2\+2=4' /tmp/ex03.txt

Real-World Applications

Professional: Finding Log Levels

# Match [INFO], [WARN], [ERROR] in logs
grep -E '\[(INFO|WARN|ERROR)\]' /var/log/app.log

Professional: ISE Authentication Logs

# Match Passed-Authentication or Failed-Attempt
grep -E 'Passed-Authentication|Failed-Attempt' /var/log/ise-psc.log

Personal: Find Prices in Documents

# Find any dollar amounts
grep -Eo '\$[0-9,]+(\.[0-9]{2})?' ~/Documents/receipts/*.txt

Personal: Search for URLs

# Find URLs (escape the dots and slashes)
grep -Eo 'https?://[a-zA-Z0-9\./\-]+' ~/notes/*.md

Tool Variants

sed: Escape for Substitution

# Replace [ERROR] with [CRITICAL]
echo '[ERROR] System failure' | sed 's/\[ERROR\]/[CRITICAL]/'

awk: Double Escaping

# In awk, backslashes need extra escaping
echo '192.168.1.100' | awk '/192\.168\.1\.100/ {print "Found IP"}'

vim: Same Escaping Rules

# Search for [ERROR]
/\[ERROR\]

# Substitute $99 with $49
:%s/\$99/\$49/g

Python: Raw Strings Help

import re

# Use r'' raw strings to avoid double-escaping
pattern = r'\[ERROR\]'
text = '[ERROR] Connection failed'
if re.search(pattern, text):
    print('Found error')

Key Takeaways

Concept Remember

Concept	Remember
Dot `.`	Matches ANY character. Always escape in IPs, domains, filenames.
Brackets `[]`	Define character classes. Escape with `\[` and `\]` for literal.
Dollar `$`	End anchor. Escape with `\$` for currency.
Question `?`	Optional quantifier. Escape with `\?` for literal question marks.
Backslash `\`	Escape character. Use `\\` to match a literal backslash.

Dot .

Matches ANY character. Always escape in IPs, domains, filenames.

Brackets []

Define character classes. Escape with \[ and \] for literal.

Dollar $

End anchor. Escape with \$ for currency.

Question ?

Optional quantifier. Escape with \? for literal question marks.

Backslash \

Escape character. Use \\ to match a literal backslash.

Common Mistakes

Mistake Wrong Correct

Mistake	Wrong	Correct
Unescaped dot in IP	`192.168.1.1`	`192\.168\.1\.1`
Unescaped brackets	`[ERROR]`	`\[ERROR\]`
Forgetting shell quoting	`grep $99 file`	`grep '\$99' file`
Windows paths	`C:\Users`	`C:\\Users` (in shell)

Unescaped dot in IP

192.168.1.1

192\.168\.1\.1

Unescaped brackets

[ERROR]

\[ERROR\]

Forgetting shell quoting

grep $99 file

grep '\$99' file

Windows paths

C:\Users

C:\\Users (in shell)

Self-Test

Answer these without looking:

What does . match in regex?
List 5 metacharacters that need escaping
How do you match a literal backslash?
What flag makes matching case-insensitive?
Why is 192.168.1.1 a BAD pattern for matching IPs?

Answers

Any single character except newline
. * + ? ^ $ | ( ) [ ] { } \ (any 5)
\\
-i flag in grep, re.IGNORECASE in Python
The dots match ANY character, so it would also match "192X168Y1Z1"

Next Drill

Drill 02: Character Classes - Learn [abc], [0-9], and PCRE shorthand.