R Regular Expressions

A regular expression (regex) is a pattern that describes a set of strings. Instead of searching for an exact word, regex lets you search for patterns — "any 10-digit number", "any email address", "any word starting with A". R uses regex with functions like grepl(), gsub(), and regexpr().

Basic Pattern Building Blocks

Pattern   Meaning                          Example
────────────────────────────────────────────────────────────
.         Any single character             "c.t" matches "cat","cut","c9t"
*         Previous item zero or more times "go*" matches "g","go","goo"
+         Previous item one or more times  "go+" matches "go","goo"
?         Previous item zero or one time   "colou?r" matches "color","colour"
^         Start of string                  "^R" matches "R is fun"
$         End of string                    "csv$" matches "data.csv"
[abc]     One of these characters          "[aeiou]" any vowel
[^abc]    None of these characters         "[^0-9]" any non-digit
[a-z]     Range of characters             "[a-z]" lowercase letter
\\d       Digit (0-9)                      "\\d{3}" three digits
\\D       Non-digit                        
\\w       Word character (letter/digit/_)  
\\s       Whitespace                       

Quantifiers

{n}       Exactly n times           "\\d{4}" → exactly 4 digits
{n,}      At least n times          "\\d{2,}" → 2 or more digits
{n,m}     Between n and m times     "\\d{2,4}" → 2 to 4 digits

grepl() — Pattern Matching

# Check if emails are valid (simplified pattern)
emails <- c("user@example.com", "badEmail", "test@domain.org", "noatsign")

valid <- grepl("^[a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\\.[a-zA-Z]{2,}$", emails)
print(valid)
# TRUE FALSE TRUE FALSE

Practical Regex Patterns

Pattern                    What It Matches
────────────────────────────────────────────────────────────
"^\\d{10}$"                Exactly 10 digits (phone number)
"^[A-Z]{2}\\d{6}$"        2 uppercase + 6 digits (passport ID)
"\\d{1,2}/\\d{1,2}/\\d{4}" Date in DD/MM/YYYY format
"^[A-Za-z ]+$"             Only letters and spaces
"[0-9]+(\\.[0-9]+)?"       Integer or decimal number
"\\b\\w{5}\\b"             Exactly 5-letter words

gsub() and sub() — Replace with Regex

# Remove all non-numeric characters
messy <- "Phone: (91) 98765-43210"
clean <- gsub("[^0-9]", "", messy)
# "919876543210"

# Remove extra whitespace
text <- "Too   many    spaces   here"
clean <- gsub("\\s+", " ", text)
# "Too many spaces here"

# Mask email domain for privacy
email <- "user@private.com"
masked <- gsub("@.*$", "@***", email)
# "user@***"

regmatches() — Extract Matches

text <- "Call us at 9876543210 or 8765432109 for help"

matches <- regmatches(text, gregexpr("\\d{10}", text))[[1]]
print(matches)
# "9876543210" "8765432109"

Groups and Capture with regmatches

# Extract date parts
dates <- c("15/08/2024", "01/01/2025", "25/12/2024")
pattern <- "(\\d{2})/(\\d{2})/(\\d{4})"

m <- regmatches(dates, regexec(pattern, dates))
m[[1]]  # "15/08/2024" "15" "08" "2024"

Real Use Case: Validate and Extract PIN Codes

addresses <- c(
  "12 MG Road, Bengaluru 560001",
  "45 Park Street, Kolkata 700016",
  "No PIN here",
  "Chennai 600001"
)

# Check if PIN code present (6 digits)
has_pin <- grepl("\\b[1-9][0-9]{5}\\b", addresses)
print(has_pin)
# TRUE TRUE FALSE TRUE

# Extract PIN codes
pins <- regmatches(addresses, regexpr("\\b[1-9][0-9]{5}\\b", addresses))
print(pins)
# "560001" "700016" "600001"

Quick Reference: Common Anchors and Groups

^          start of string
$          end of string
\\b        word boundary (between word char and non-word char)
( )        grouping and capture
|          OR  ("cat|dog" matches "cat" or "dog")
(?i)       case-insensitive (in Perl regex: perl=TRUE)

Regular expressions are one of the most powerful text-processing tools in programming. A single regex line can validate thousands of email addresses, extract all phone numbers from a document, or clean an entire column of messy text. Start with simple patterns and build up gradually — each new pattern you learn dramatically expands what you can do with text data.