Regular Expressions

Course Title: Formal Languages & Theory of

Automata
Course Code : CS224
Credits :3
Instructor : Muhammad Sajid Ali
Last Lecture Exercises
 Show that
– (S+)* = (S*)*
– (S+)+ = S+
– (S*)+ = (S+)*
Last Lecture Exercises
Show that  (S+)* = (S*)*

Λ is not in language generated by S+ however,

language generated by S* includes Λ so, clearly (S+)*
will include Λ as well

(S*)* also generate the same language generated by

(S+)*

Hence (S+)* = (S*)*
Last Lecture Exercises
Show that  (S+)+ = S+

S+ generates the language of all strings by

concatenating all words in S

(S+)+ generates the language of all strings by

concatenating all words in S+

This will not produce any new string

Hence (S+)+ = S+
Introduction – Regular Expression
Previously, we used phrases to describe a language

E.g. L = { xn for n = 1, 2, 3, …}
– This means, language L contains all combination of x of
any length except Λ

E.g. L = { xn for n = 1, 3, 5, …}
– This means, language L will have all strings of odd length

E.g. L = { xn for n = 1, 4, 9, 16, …}

– Defining this language is easy?

E.g. L = { xn for n = 3, 4, 8, 22, …}

– Can you give the clear description of this language?
Introduction – Regular Expression
Let’s learn a new way to define the whole language
which is more precise than just a guess work.

E.g. L = {Λ, x, xx, xxx, xxxx, …}

– Remember Kleene closure *
– let S = {x}  L = S*
– L = {x}*

Let’s define Kleene closure over the letter but not to

a set
– x*  means sequence of x’s of any length or none at all
– * over letter represents any combination of concatenation
of letter x
L = language(x*) ?
Introduction – Regular Expression
 E.g. Σ = {a, b}
– Let L = { a, ab, abb, abbb, abbbb, … }

– Descriptive definition: “All words of the form one ‘a’

followed by some number of b’s or no b at all”

– Regular Expression way to define this language is ab*

– Note * is only on ‘b’ and not on ‘a’

 We can apply * to both ‘ab’ as well

– (ab)* , which means Λ, ab, abab, ababab, …
Regular Expression
 let L = {x, xx, xxx, …}, You may define this
language in multiple ways e.g.
– xx*
– x+
– xx*x*
– x+x*
– x*x+

 * will produce Λ
Regular Expression
 What language is defined by the following
expression ab*a
– L = {aa, aba, abba, abbba}
– Description?

 What language is defined by the following

expression a*b*
– All strings of ‘a’ (if any) and ‘b’ (if any) in which ‘a’ if any
comes before ‘b’ if any
– L={?}
– ba & aba ∈ L(a*b*) ?

– Also note, no needs to be the same number of a’s and b’s

  a*b* ≠ (ab)*
Regular Expression
 Let L = {xodd}

– Regular expression of this language would be

– x(xx)* or (xx)*x

– x*xx*  ?
Regular Expression - Plus Operator
 Let Expression (E) = x + y

– x & y are symbols from alphabet

– + gives a choice to choose one letter, either x or y

 E.g. Σ = {a, b, c}

– L = { a, c, ab, cb, abb, cbb, abbb, cbbb, … }

– Description of (L)  All words start with a or c and may be
followed by any number of b’s

– R.E = (a + c) b*
Regular Expression – Examples
 L = { aaa, aab, aba, abb, baa, bab, bba, bbb }

– Description of (L)  Any word of length 3

– R.E = (a + b) (a + b) (a + b)

 L = All possible strings of a’s and b’s of any length

– R.E = (a + b)*
– (a + b)5  ?
Regular Expression – Examples
 L = All words that start with ‘a’ over Σ = {a, b}

–R.E = a (a + b)*

 L = All words start with ‘a’ and ends with ‘b’ over Σ
= {a, b}

–R.E = a (a + b)* b

 All languages which can describe with R.E are

known as regular languages
Regular Expression – Formal Definition
 Symbols in Regular Expression (RE)

– Alphabet letters, e.g. Σ = {a, b}

– Λ string
– () parentheses for grouping
– * operator
– + operator
Regular Expression – Formal Definition
 Regular Expression (R.E) is defined by the
following rules.
– Every letter from alphabet Σ can be made into RE, Λ itself
is a regular expression
– If r1 and r2 are two regular expressions than
(r1)
r1 + r2
r1 r2
r1 *
– Nothing else is a regular expression
Regular Expression – Formal Definition
 r1* means

– If r1 = aa + b  (aa + b)*
Regular Expression – Union of Two Languages
Let R.E  (a + b)* a (a + b)*

– This means language defines all words that have an ‘a’ in

them somewhere.
– You may generate word abbaab by these different
choices Λ a (bbaab) or (abb) a (ab) or (abba) a (b)
– Think about ‘Λ’ and words that have only ‘b’s’ ?
• These are the only words left out by the above regular expression
– These omitted left out words can be generated by the b*
language

let combine these two languages

– (all strings with an a) + (all strings without an a)
– (a + b)* a (a + b)* + b*

Notice: we combined two languages to produce an expression that defines the

union of two languages.
Regular Expression – Union of Two Languages

 Don’t forget + means choice in R.E

 When we added + between two languages, it

means first choose the left set or the right set and
then find the desired word in that set.

 For plus as union or plus as choice, these all make

sense but for algebra these are misguided
Regular Expression – Examples

 Write the regular expression of all words that have

at least two a’s.

– (a + b)* a (a + b)* a (a + b)*

Regular Expression – Equivalent Expressions

 Another expression to define the same language

that have at least two a’s.

– b*ab*a(a + b)*

 Two expressions are equivalent if they describe

the same language
Regular Expression – Examples

 Write the regular expression of all words that have

at exactly two a’s.

– (a + b)* a (a + b)* a (a + b)*  is it correct ?

– b*ab*ab*
Regular Expression – Examples

 Write the regular expression of the language that

have at least one ‘a’ and one ‘b’.

– (a+b)* a (a+b)* b (a+b)* is this correct ?

– This expression saying that strings like ba, bbaaa are not
included, however ‘ba’ is a valid word as we define our
language
– Let’s include these words as well, can you think a way ?
– (a+b)* b (a+b)* a (a+b)*
– (a+b)* a (a+b)* b (a+b)* + (a+b)* b (a+b)* a (a+b)*
– We can also define these words exceptions as bb*aa*
– (a+b)* a (a+b)* b (a+b)* + bb*aa* defines the same lang

 (a+b)* a (a+b)* b (a+b)* + (a+b)* b (a+b)* a (a+b)* = (a+b)* a (a+b)* b (a+b)* + bb*aa*
Regular Expression – Examples

 Can you change previous expression to generate

all words that include all a’s and b’s as well

– Find out what left (i.e. cannot be generated)?

– Words that do not contain both ‘a’ and ‘b’ are the words óf
all a’s, b’s or Λ

– (a+b)* a (a+b)* b (a+b)* + bb*aa* + a* + b*

 See some expression equivalence examples at

page# 40
Regular Expression & Algebra

 In algebra ab = ba

 In formal languages ab ≠ ba

 Let T = { a, c, ab, cb, abb, cbb, … } can be define

with R.E  (a + c) b*

 This can also be written as ab* + cb*  distributive

law

 Expression may be distributed but operation

cannot, i.e. (ab)* ≠ a*b*
Regular Expression & Algebra

 In algebra ab = ba

 In formal languages ab ≠ ba

 Let T = { a, c, ab, cb, abb, cbb, … } can be define

with R.E  (a + c) b*

 This can also be written as ab* + cb*  distributive

law

 Expression may be distributed but operation

cannot, i.e. (ab)* ≠ a*b*
Regular Expression & Algebra

Product of two sets

E.g  S = {a, aa, aaa} & T = {bb, bbb}

– ST  {abb, abbb, aabb, aabbb, aaabb, aaabbb}

E.g  S = {a, bb, bab} & T = {a, ab}

– ST  {aa, aab, bba, bbab, baba, babab}

E.g  S = {Λ, x, xx} & T = {Λ, y, yy, yyy, yyyy, …}

– ST  {
Λ, y, yy, yyy, yyyy, …
x, xy, xyy, xyyy, xyyyy, …
xx, xxy, xxyyy, xxyyyy
}
Languages Associated With RE

Following rules define the language associated with

any regular expression

The language associated with the regular

expression that is just a single letter is that one-letter
word alone and the language associated with Λ is just
{Λ}. A one-word language.

If r1 is a regular expression associated with the

language L1 and r2 is a regular expression
associated with language L2 then
– Language(r1r2)  L1L2
– Language(r1 + r2)  L1 + L2
– Language(r1*)  L1*
Languages Associated With RE – Questions to Think

Following rules define the language associated with

any regular expression

We see previously that two different RE presents

same language, can we define a way to determine
when that happens?

We have seen that every regular expression is

associated with a language, is it also true that every
language can be defined by RE
Languages Asociated With RE – Questions to Think

Following rules define the language associated with

any regular expression

We see previously that two different RE presents

same language, can we define a way to determine
when that happens?

We have seen that every regular expression is

associated with a language, is it also true that every
language can be defined by RE
Finite Languages are Regular

If L is a finite language  a language with only finitely

many words

Then L can be defined by a regular expression

L = { abba, baaa, bbbb }

– R.E  abba + baaa + bbbb

L = { baa, abbba, bababa }

– R.E  baa + abbba + bababa

L = { aa, ab, ba, bb }

– R.E  aa + ab + ba + bb
Finite Languages are Regular

If L is a finite language  a language with only finitely

many words

Then L can be defined by a regular expression

L = { abba, baaa, bbbb }

– R.E  abba + baaa + bbbb

L = { baa, abbba, bababa }

– R.E  baa + abbba + bababa

L = { aa, ab, ba, bb }

– R.E  aa + ab + ba + bb
Finite Languages are Regular

 L = { Λ, x, xx, xxx, xxxx, xxxxx }

– R.E  Λ + x + xx + xxx + xxxx + xxxxx
– More elegant way  (Λ + x)5
How Hard IT Is To Understand RE
Let (a + b)*(aa + bb)(a + b)*

This is the set of strings of a’s and b’s that at some

point contain double letter.

Question: What strings do not contain double

letter?
– Λ, a, ab, ba, aba, bab, abab, baba, …
– (ab)* covers all of these kind of words except those which
starts with b and end with a. let’s add this choice as well
– (Λ + b)(ab)*(Λ + a)
Combining these REs will give the following
– (a + b)*(aa + bb)(a + b)* + (Λ + b)(ab)*(Λ + a)

– Can you immediately tell that this expression represents

all strings?
How Hard IT Is To Understand RE
Let (a + b)* a (a + b)* (a + Λ) (a + b)* a (a + b)*

All words that have at least two a’s.

Let’s break this RE into two cases ?

– (a + b)* a (a + b)* a (a + b)* a (a + b)*
+
– (a + b)* a (a + b)* Λ (a + b)* a (a + b)*

The first part of RE says, all words that have at least

three a’s
In second part of RE,
– do you see that (a + b)* Λ (a + b)* is (a + b)* ?
– (a + b)* a (a + b)* a (a + b)*
– Have you seen this kind of RE before?
How Hard IT Is To Understand RE
 So,
– (a + b)* a (a + b)* a (a + b)* a (a + b)*
+
– (a + b)* a (a + b)* a (a + b)*

 This language is the union of all words with atleast

three a’s in them and lanaguage of all words with
atleast two a’s in them

 But all strings with three or more a’s are

themselves the strings with two or more a’s
How Hard IT Is To Understand RE

 Star operator can be applied to an expression that

already have a star in it,
– (a + b*)*
– (aa + ab*)*
– ((a + bbba*) + ba*b)*

 (a + b*)*, internal * is adding something new?

– (a + b*)* = (a + b)*

 What do you see about second expression?

– (aa + ab*) = (aa + ab)* ?
Even Even Language
 Consider the following RE,
RE = [ aa + bb + (ab + ba) (aa + bb)* (ab + ba) ]*

 All collection of words are made up of the following

– aa
– bb
– (ab + ba) (aa + bb)* (ab + ba)

 Every word of the language of E contains an even

number of a’s and an even number of b’s

 All strings with an even number of a’s and an even

number of b’s belong to the language of E
Even Even Language
 Can you write and algorithm to find if the given
word belongs to E or not?

 Use two flags

– One for a
– One for b

 Each time ‘a’ or ‘b’ is encountered, the respective

flag is inverted

 If starting values of flag matche, the value at the

end of the end of the string than the respective
string belongs to Even-Even language.
Even Even Language
 Can you do the same using one flag?

 Use one flags

 Read two letters at a time

 If both letters mismatch

– Invert the flag

 If starting values of flag matche, the value at the

end of the end of the string than the respective
string belongs to Even-Even language.