Introduction

A regular expression (regex in short) is a pattern in input text that the regular expression
engine tries to match. One or more character literals, operators, or structures make up a
pattern. It's particularly good at searching for and manipulating text strings, as well as
processing text files. A single regex line can easily replace dozens of lines of programming
code. In this blog, you will come across the basic to advanced regex concepts that would help
you in your development task.

Why Regex?

Searching and replacement operations are made easy by regular expressions. Finding a
substring that matches a pattern and replacing it with something else is a common use case.
Regex is renowned for the IT expertise that dramatically boosts productivity in all computer
tasks.

Basic Regex Commands

1. Characters

The character class is the most basic regex concept. It converts a small group of characters to
match a larger number of characters.

1
Character Description
\ Escape character
. Any character
\s whitespace
\S Not white space
\d Digit
\D Not digit
\w Word character
\W Not word character
\b Word boundary
\B Not word boundary
^ Beginning of string
$ End of string

For example:

1. [xyz] matches x or y or z
2. [^pqr] matches any character _except_ p, q, or r (negation)
3. [a-zA-Z] matches a through z or A through Z, inclusive (range)

2. Groups

To operate on all items in a group, use a group expression. You can use a group expression
to apply an operator to a group or to find a specified text before or after each member of the
group, for example. The grouping operator is the parentheses, while the "|" is used to divide
the elements.

For example:

• My (red|pink|blue) dress
o Here "My red dress", "My blue kite", and "My pink dress" match the expression.
"My yellow kite" or "My dress" do not match.

2
Let’s look into some of these operators.

Group Description
[ ] Characters in brackets are matched.
[^ ] Characters not in brackets are matched.
| Either, or
( ) Capturing the group

More examples:

1. x(yz) parentheses create a capturing group with value yz

2. p(?:qr)* using ?: you disable the capturing group
3. a(?<ok>bc) using ?<ok> we put a name to the group

3. Quantifiers

Quantifiers specify how many characters or expressions should be matched.

Quantifiers Description
* 0 or more (Kleene star)
+ 1 or more (Kleene plus)
? 0 or 1
{ } Exact number of characters
{min,max} Range of characters

The quantifiers ( * + {}) are also known as greedy operators. They expand the match as far as
they can via the input text.

For example:

1. <[^<>]+> matches any of the characters except < or > included one or more times
inside < and >
2. \w+?\d\d\w+ matches abcdef42ghijklmnfhaeij

More examples for quantifiers

3
x{3} Exactly 3 of x
x{3,} 3 or more of x
x{3,6} Between 3 and 6 of x
x* Greedy quantifier
x*? Lazy quantifier
x*+ Possessive quantifier

4. Anchors

Based on the current position in the string, determines whether the match will succeed or fail.

Anchors Description
^ Beginning of the string
$ End of the string
\A The match is at the beginning of the string.
\G Beginning of the match
\Z At the end of the string or before \n at the end of the string, the match occurs.
\z Absolute end of the string
\B No word boundary
\b Word boundary

For example:

1. ^\d{3} matched 444 in 444-888-999-..

2. -\d{3}\Z matches -220 in 110-220
3. \babc\b performs a "whole words only" search
4. \Bend\w*\b matches "ends", "ender" in "end sends endure lender"
5. ^Hello me$ matches the string Hello me

Now, that you have got some brief idea about the characters, quantifiers and groups. Let’s look
into some combined examples.

4
 1. ^(\d*) [.,](\d+)$ matches numbers like 12,3 or 12.3
2. ^[a-zA-Z0-9 ]*$ matches any alphanumeric with spaces.
3. ^[\s]*(.*?)[\s]*$ matches the text by avoiding the extra spaces
4. ^([a-zA-Z0-9._%-]+@[a-zA-Z0-9.-]+\.[a-zA-Z]{2,6})*$ could be used for matching email
5. (https?)://(www)?.?(\\w+).(\\w+)/?(\\w+)? can be used matching URLs

Advanced Regex Commands

1. Lookarounds

When the regex engine processes the lookaround expression, it first creates a substring from
the present place to the beginning (lookbehind) or end (lookahead) of the original string, and
then runs Regex. With the use of the lookaround pattern, IsMatch on that picked substring. A
positive or negative assertion might be used to assess the outcome's success.

Lookaround Description
(?=check) Positive Lookahead
(?!check) Negative Lookahead
(?<=check) Positive Lookbehind
(?<!check) Negative Lookbehind

Examples:

1. (?=\d{10})\d{4} matches 2348 in 2348856787

2. (?<=\d)rat matches mat in 2mat
3. (?!theatre)the\w+ matches theme
4. \w{3}(?<!mon)ster matches munster

Explore InterviewBit’s Exclusive Live Events

2. POSIX commands

A character class is a combination of a small number of characters and a large number of

characters. Only within bracket expressions can we use POSIX character classes. To generate
regular expressions, the POSIX standard supports the following character classes.

5
POSIX Description
[:alpha:] PCRE (C, PHP, R…): ASCII letters A-Z and a-z
[:alpha:] Ruby 2: Unicode letter or ideogram
[:alnum:] PCRE (C, PHP, R…): ASCII digits and letters A-Z and a-z
[:alnum:] Ruby 2: Unicode digit, letter or ideogram
[:punct:] PCRE (C, PHP, R…): ASCII punctuation mark
[:punct:] Ruby: Unicode punctuation mark

Examples:

1. [8[:alpha:]]+ sample match could beWellDone88

2. [[:alpha:]\d]+ sample match could beкошка99
3. [[:alnum:]]{10} sample match could beABC1275851
4. [[:alnum:]]{10} sample match could be кошка67810
5. [[:punct:]]+ sample match could be ?!.,:;
6. [[:punct:]]+ sample match could be ‽,:〽⁆

3. Unicode property escapes

The following are examples of Unicode property escapes:

1. \p{prop=value}: All characters with the prop property have the value value.
2. \P{prop=value}: All characters without a property prop with the value value are matched.
Match \p{bin_prop}: all characters with the bin prop binary property set to True.
3. \P{bin_prop}: All characters with the binary attribute bin prop set to False will be
matched.

4. Flags

A flag is a parameter that can be added to a regex to change how it searches. A flag modifies
a regular expression's default searching behaviour. It does a regex search in a unique method.
A single lowercase alphabetic character is used to represent a flag. There are six flags in the
JavaScript regex, each providing a different purpose.

6
Flag Description
i Makes the expression search case-insensitively.
g Makes the expression search for all occurrences.
s Makes a wild character. match newlines as well.
Instead of matching the beginning and conclusion of the entire string, the boundary
m characters ^ and $ match the beginning and ending of each individual line.
y Starts the expression's search from the index specified by the lastIndex attribute.
Assumes that individual characters are code points rather than code units, and so matches
u 32-bit characters.

When employing the forward slashes / to build an expression, flags come after the second
slash. This can be expressed in general notation as follows: \pattern\flag

For example, if the flag i was added to the regex /a/, the result would be /a/i.

To provide a regex with many flags, we write them one by one (without any spaces or other
delimiters).

If we gave the flags i and g to the regex /a/, for example, we'd write /a/ig (or equivalently /a/gi,
as the order doesn't matter).

Note: The sequence in which flags appear is irrelevant; flags simply change the behaviour of
searching, thus placing one before the other makes no difference.

5. Recurse

The following engines enable recursion: PCRE (C, PHP, R...) Perl Ruby 2+ Python via the
alternate regex package JGSoft (not available in a programming language)

The most common application of recursion is to match balanced constructions.

Command Description
(?R) Recurse entire pattern
(?1) Recurse the first subpattern
(?+1) Recurse first relative subpattern

7
Command Description
(?&name) Recurse subpattern name
(?P=name) Match subpattern name
(?P>name) Recurse subpattern name

a(?R)?z, a(?0)?z, and a\g<0>?z are all regexes that match one or more letters followed by the
same number of letters z.

Examples:

• (\((?R)?\)) match parentheses like ((()))

• (\((?R)*\)) match parentheses like (()()())
• \w{3}\d{4}(?R)? matches patterns like ccc8888ggg9999

Possible Performance pitfalls

1. Regular Expression Performance Pitfalls

Because two "equivalent" regexes might have substantial changes in processing performance,
you should understand how your regex engine works.

1. It is feasible to construct regexes that match in exponential time, but you must essentially
TRY to do so.
2. Regexes that run in quadratic time are more commonly created by accident.
3. Problems of many kinds
• Recompilation (from forgetting to compile regexes used multiple times)
• The Middle Dot-star (which causes backtracking)
o The first approach, use a character class that is negated.
o Use reluctant quantifiers as a second option.

Tips to increase the performance

1. Improving the performance of regular expressions

• When you require parentheses but not capture, use non-capturing groups.

8
 • Do a quick check before attempting a match, if the regex is very complex, e.g.
o Does an email address contain '@'?
• Present the most likely option(s) first, e.g.
o light green|dark green|brown|yellow|green|pink leaf
• Minimize the amount of looping
o \d\d\d\d\d\d is faster than \d{6}
o aaaaaa+ is faster than a{6,}
• Avoid obvious backtracking, e.g.
o Mr|Ms|Mrs should be M(?:rs?|s)
o Good night|Good morning should be Good (?:night|morning)

Conclusion

In this blog, you came across some of the interesting regex command concepts that would be
very helpful in the matching of different kinds of strings. You also learned about some of the
pitfalls that could happen when using regex and that could affect the performance of your
application too. Further to get over these pitfalls, the blog also discusses some of the common
tips that could help you in overcoming these obstacles.

9
10