Properties of Context-free

Languages

1
Topics
1) Simplifying CFGs, Normal forms
2) Pumping lemma for CFLs
3) Closure and decision properties of
CFLs

2
How to “simplify” CFGs?

3
Three ways to simplify/clean a CFG
(clean)
1. Eliminate useless symbols

(simplify)
2. Eliminate -productions A => 

3. Eliminate unit productions A => B

4
Eliminating useless symbols

Grammar cleanup

5
Eliminating useless symbols
A symbol X is reachable if there exists:
 S *  X 

A symbol X is generating if there exists:

 X * w,

for some w  T*

For a symbol X to be “useful”, it has to be both

reachable and generating


S *  X  * w’, for some w’  T*
reachable generating

6
Algorithm to detect useless
symbols
1. First, eliminate all symbols that are not
generating

2. Next, eliminate all symbols that are not

reachable

Is the order of these steps important,

or can we switch?

7
Example: Useless symbols
 SAB | a
 A b

1. A, S are generating
2. B is not generating (and therefore B is useless)
3. ==> Eliminating B… (i.e., remove all productions that involve
B)
1. S a
2. Ab
4. Now, A is not reachable and therefore is useless

5. Simplified G: What would happen if you reverse the order:

1. Sa i.e., test reachability before generating?
Will fail to remove:
Ab
8
 X * w

Algorithm to find all generating symbols

 Given: G=(V,T,P,S)
 Basis:
 Every symbol in T is obviously generating.
 Induction:
 Suppose for a production A , where  is
generating
 Then, A is also generating

9
 S *  X 
Algorithm to find all reachable symbols
 Given: G=(V,T,P,S)
 Basis:
 S is obviously reachable (from itself)
 Induction:
 Suppose for a production A 1 2… k,
where A is reachable
 Then, all symbols on the right hand side,
{1, 2 ,… k} are also reachable.

10
Eliminating -productions
A => 

11
What’s the point of removing -productions?
A
Eliminating -productions
Caveat: It is not possible to eliminate -productions for
languages which include  in their word set
So we will target the grammar for the rest of the language
Theorem: If G=(V,T,P,S) is a CFG for a language L, then
L\ {} has a CFG without -productions

Definition: A is “nullable” if A* 

 If A is nullable, then any production of the form

“B CAD” can be simulated by:

 B  CD | CAD

 This can allow us to remove  transitions for A

12
Algorithm to detect all nullable
variables
 Basis:
 If A  is a production in G, then A is
nullable
(note: A can still have other productions)
 Induction:
 If there is a production B C1C2…Ck,
where every Ci is nullable, then B is also
nullable

13
Eliminating -productions
Given: G=(V,T,P,S)
Algorithm:
1. Detect all nullable variables in G
2. Then construct G1=(V,T,P1,S) as follows:
i. For each production of the form: AX1X2…Xk, where
k≥1, suppose m out of the k Xi’s are nullable symbols
ii. Then G1 will have 2m versions for this production
i. i.e, all combinations where each X i is either present or absent
iii. Alternatively, if a production is of the form: A, then
remove it

14
Example: Eliminating -
productions
 Let L be the language represented by the following CFG G:
i. SAB
ii. AaAA | 
iii. BbBB |  Simplified
grammar
Goal: To construct G1, which is the grammar for L-{}

 Nullable symbols: {A, B}

 G1 can be constructed from G as follows: G1:

 B  b | bB | bB | bBB • S  A | B | AB
 ==> B  b | bB | bBB • A  a | aA | aAA
 Similarly, A  a | aA | aAA • B  b | bB | bBB
 Similarly, S  A | B | AB
+
 Note: L(G) = L(G1) U {} • S

15
 Eliminating unit productions
A => B B has to be a variable

What’s the point of removing unit transitions ?

Will save #substitutions
E.g., A=>B | … A=>xxx | yyy | zzz | …
B=>C | … B=> xxx | yyy | zzz | …
C=>D | … C=> xxx | yyy | zzz | …
D=>xxx | yyy | zzz D=>xxx | yyy | zzz
after 16
before
 AB
Eliminating unit productions
 Unit production is one which is of the form A B, where both A & B
are variables
 E.g.,
1. E  T | E+T
2. T  F | T*F
3. F  I | (E)
4. I  a | b | Ia | Ib | I0 | I1
 How to eliminate unit productions?

 Replace E T with E  F | T*F

 Then, upon recursive application wherever there is a unit production:

 E F | T*F | E+T (substituting for T)
 E I | (E) | T*F| E+T (substituting for F)
 E a | b | Ia | Ib | I0 | I1 | (E) | T*F | E+T (substituting for I)
 Now, E has no unit productions

 Similarly, eliminate for the remainder of the unit productions

17
The Unit Pair Algorithm:
to remove unit productions
 Suppose AB1 B2  …  Bn  
 Action: Replace all intermediate productions to produce 
directly

i.e., A ; B1 ; … Bn  ;

Definition: (A,B) to be a “unit pair” if A*B

 We can find all unit pairs inductively:

 Basis: Every pair (A,A) is a unit pair (by definition). Similarly, if AB
is a production, then (A,B) is a unit pair.

 Induction: If (A,B) and (B,C) are unit pairs, and AC is also a unit
pair.

18
The Unit Pair Algorithm:
to remove unit productions
Input: G=(V,T,P,S)
Goal: to build G1=(V,T,P1,S) devoid of unit
productions
Algorithm:
1. Find all unit pairs in G
2. For each unit pair (A,B) in G:
1. Add to P1 a new production A, for every
B which is a non-unit production
2. If a resulting production is already there in P,
then there is no need to add it.
19
 Example: eliminating unit
productions
Unit pairs Only non-unit
productions to be
added to P1
G:
1. E  T | E+T (E,E) E  E+T
2. T  F | T*F
3. F  I | (E) (E,T) E  T*F
4. I  a | b | Ia | Ib | I0 | I1 (E,F) E  (E)
(E,I) E  a|b|Ia | Ib | I0 | I1
(T,T) T  T*F
(T,F) T  (E)
(T,I) T  a|b| Ia | Ib | I0 | I1
G1 :
1. E  E+T | T*F | (E) | a| b | Ia | Ib | I0 | I1 (F,F) F  (E)
2. T  T*F | (E) | a| b | Ia | Ib | I0 | I1
(F,I) F  a| b| Ia | Ib | I0 |
3. F  (E) | a| b | Ia | Ib | I0 | I1
4. I  a | b | Ia | Ib | I0 | I1 I1
(I,I) I  a| b | Ia | Ib | I0 |
I1
20
Putting all this together…
 Theorem: If G is a CFG for a language that
contains at least one string other than , then there
is another CFG G1, such that L(G1)=L(G) - , and
G1 has:
 no  -productions
 no unit productions
 no useless symbols

 Algorithm:
Step 1) eliminate  -productions Again,
Step 2) eliminate unit productions the order is
Step 3) eliminate useless symbols important!

Why?
21
Normal Forms

22
Why normal forms?
 If all productions of the grammar could be
expressed in the same form(s), then:

a. It becomes easy to design algorithms that use

the grammar

b. It becomes easy to show proofs and properties

23
Chomsky Normal Form (CNF)
Let G be a CFG for some L-{}
Definition:
G is said to be in Chomsky Normal Form if all
its productions are in one of the following
two forms:
i. A  BC where A,B,C are variables, or
ii. Aa where a is a terminal
 G has no useless symbols
 G has no unit productions
 G has no -productions
24
CNF checklist
Is this grammar in CNF?
G1 :
1. E  E+T | T*F | (E) | Ia | Ib | I0 | I1
2. T  T*F | (E) | Ia | Ib | I0 | I1
3. F  (E) | Ia | Ib | I0 | I1
4. I  a | b | Ia | Ib | I0 | I1

Checklist:
• G has no -productions
• G has no unit productions
• G has no useless symbols
• But…
• the normal form for productions is violated

So, the grammar is not in CNF

25
How to convert a G into CNF?
 Assumption: G has no -productions, unit productions or useless
symbols

1) For every terminal a that appears in the body of a production:

i. create a unique variable, say Xa, with a production Xa  a, and
ii. replace all other instances of a in G by Xa

2) Now, all productions will be in one of the following

two forms:

A  B1B2… Bk (k≥3) or Aa

B2 C2 and so on…
3) Replace each production of the form A  BB1B B … Bk by:
1 2 3C1

26
 Example #1
G in CNF:
G:
X0 => 0
S => AS | BABC
X1 => 1
A => A1 | 0A1 | 01
S => AS | BY1
B => 0B | 0
Y1 => AY2
C => 1C | 1 Y2 => BC
A => AX1 | X0Y3 | X0X1
Y3 => AX1
B => X0B | 0
C => X1C | 1

All productions are of the form: A=>BC or A=>a

27
 Example #2
G: 1. E  EX+T | TX*F | X(EX) | IXa | IXb | IX0 | IX1
1. E  E+T | T*F | (E) | Ia | Ib | I0 | I1 2. T  TX*F | X(EX) | IXa | IXb | IX0 | IX1
2. T  T*F | (E) | Ia | Ib | I0 | I1 3. F  X(EX) | IXa | IXb | IX0 | IX1
3. F  (E) | Ia | Ib | I0 | I1 4. I  Xa | Xb | IXa | IXb | IX0 | IX1
4. I  a | b | Ia | Ib | I0 | I1 Step (1) 5. X+  +
6. X*  *
7. X+  +
)
(2 8. X(  (
t ep 9. …….
S

1. E  EC1 | TC2 | X(C3 | IXa | IXb | IX0 | IX1

2. C1  X+T
3. C2  X*F
4. C3  EX)
5. T  ..…….
6. ….

28
 Languages with 
 For languages that include ,
 Write down the rest of grammar in CNF
 Then add production “S => ” at the end
E.g., consider: G in CNF:
G: X0 => 0
S => AS | BABC X1 => 1
A => A1 | 0A1 | 01 |  S => AS | BY1 |
B => 0B | 0 |  Y1 => AY2
C => 1C | 1 |  Y2 => BC

A => AX1 | X0Y3 | X0X1

Y3 => AX1
B => X0B | 0
C => X1C | 1 29
Other Normal Forms
 Griebach Normal Form (GNF)
 All productions of the form
A==>a 

CNF, eliminate left recursion, GNF.

30
Return of the Pumping Lemma !!

Think of languages that cannot be CFL

== think of languages for which a stack will not be enough

e.g., the language of strings of the form ww

31
Why pumping lemma?
 A result that will be useful in proving
languages that are not CFLs
 (just like we did for regular languages)

 But before we prove the pumping

lemma for CFLs ….
 Let us first prove an important property
about parse trees
32
Observe that any parse tree generated by a CNF will be a
binary tree, where all internal nodes have exactly two children
(except those nodes connected to the leaves).

The “parse tree theorem”

Given: Parse tree for w
 Suppose we have a S = A0
parse tree for a
string w, according A1
to a CNF grammar, A2
G=(V,T,P,S)
. h
= tree height
.
 Let h be the height of
.
the parse tree
Ah-1
Implies:
 |w| ≤ 2h-1
a

In other words, a CNF parse tree’s string yield (w) w

can no longer be 2h-1 33
To show: |w| ≤ 2h-1

Proof…The size of parse trees

Proof: (using induction on h) Parse tree for w
Basis: h = 1
 Derivation will have to be S = A0
“Sa”
 |w|= 1 = 21-1 .

Ind. Hyp: h = k-1 A B

 |w|≤ 2k-2
h
Ind. Step: h = k = height

S will have exactly two children:

SAB

 Heights of A & B subtrees are

at most h-1

wA wB
 w = wA wB, where |wA| ≤ 2 k-2

and |wB| ≤ 2k-2

w
34
k-1
Implication of the Parse Tree
Theorem (assuming CNF)
Fact:
 If the height of a parse tree is h, then
 ==> |w| ≤ 2h-1

Implication:
 If |w| ≥ 2m, then
 Its parse tree’s height is at least m+1

35
The Pumping Lemma for CFLs
Let L be a CFL.
Then there exists a constant N, s.t.,
 if z L s.t. |z|≥N, then we can write
z=uvwxy, such that:
1. |vwx| ≤ N
2. vx≠
3. For all k≥0: uvkwxky  L

Note: we are pumping in two places (v & x)

36
Proof: Pumping Lemma for CFL
 If L=Φ or contains only , then the lemma is
trivially satisfied (as it cannot be violated)

 For any other L which is a CFL:

 Let G be a CNF grammar for L
 Let m = number of variables in G
 Choose N=2m.
 Pick any z  L s.t. |z|≥ N
 the parse tree for z should have a height ≥ m+1
(by the parse tree theorem)

37
 Meaning:
Repetition in the
last m+1 variables

Parse tree for z

h-m≤ i < j ≤ h
S = A0 S = A0
+

A1 Ai = Aj
A2 Ai
, > m levels

. h ≥ m+1 h ≥ m+1

.
Aj
m variables

. m+1
Ah-1
u v x y
Ah=a
w

z z = uvwxy
• Therefore, vx≠
38
 Extending the parse tree…
S = A0
S = A0

Replacing Or, replacing

Aj with Ai Ai with Aj
(k times) Aj
Ai=Aj
h ≥ m+1

Ai
w

Ai u y
u v x y
z = uwy
v x
…

w ==> For all k≥0: uvkwxky L

z = uvkwxky 39
Proof contd..
Also, since Ai’s subtree no taller than m+1

==> the string generated under Ai‘s subtree, which is

vwx, cannot be longer than 2m (=N)

But, 2m =N

==> |vwx| ≤ N

This completes the proof for the pumping lemma.

40
Application of Pumping
Lemma for CFLs
Example 1: L = {ambmcm | m>0 }
Claim: L is not a CFL
Proof:
 Let N <== P/L constant
 Pick z = aNbNcN
 Apply pumping lemma to z and show that there
exists at least one other string constructed from z
(obtained by pumping up or down) that is  L

41
Proof contd…
 z = uvwxy
 As z = aNbNcN and |vwx| ≤ N and vx≠
 ==> v, x cannot contain all three symbols
(a,b,c)
 ==> we can pump up or pump down to build
another string which is  L

42
Example #2 for P/L application
 L = { ww | w is in {0,1}*}

 Show that L is not a CFL

 Try string z = 0N0N

 what happens?
 Try string z = 0N1N0N1N
 what happens?

43
Example 3
2
 L = { 0 | k is any integer)
k

 Prove L is not a CFL using Pumping

Lemma

44
Example 4
 L = {aibjck | i<j<k }

 Prove that L is not a CFL

45
CFL Closure Properties

46
Closure Property Results
 CFLs are closed under:
 Union
 Concatenation
 Kleene closure operator
 Substitution
 Homomorphism, inverse homomorphism
 reversal
 CFLs are not closed under: Note: Reg languages
 Intersection are closed
under
 Difference these
 Complementation operators
47
 Strategy for Closure Property
Proofs
 First prove “closure under substitution”
 Using the above result, prove other closure properties
 CFLs are closed under:
 Union
 Concatenation
 Kleene closure operator
Prove  Substitution
this first  Homomorphism, inverse homomorphism
 Reversal

48
 Note: s(L) can use
a different alphabet

The Substitution operation

For each a  ∑, then let s(a) be a language
If w=a1a2…an  L, then:

s(w) = { x1x2 … }  s(L), s.t., xi  s(ai)
Example:
 Let ∑={0,1}
 Let: s(0) = {anbn | n ≥1}, s(1) = {aa,bb}
 If w=01, s(w)=s(0).s(1)
 E.g., s(w) contains a1 b1 aa, a1 b1bb,
a2 b2 aa, a2 b2bb,
… and so on.

49
CFLs are closed under
Substitution
IF L is a CFL and a substititution defined
on L, s(L), is s.t., s(a) is a CFL for every
symbol a, THEN:
 s(L) is also a CFL

What is s(L)?
L s(L)
w1 s(w1) Note: each s(w)
w2 s(L) s(w2) is itself a set of strings
w3 s(w3)
w4 s(w4)
50
 CFLs are closed under
Substitution
 G=(V,T,P,S) : CFG for L
 Because every s(a) is a CFL, there is a CFG for each s(a)
 Let Ga = (Va,Ta,Pa,Sa)
 Construct G’=(V’,T’,P’,S) for s(L)
 P’ consists of:
 The productions of P, but with every occurrence of terminal “a” in
their bodies replaced by Sa.
 All productions in any Pa, for any a  ∑

Parse tree for G’: S

Sa1 Sa2 Sa
n

…
x1 x2 xn 51
Substitution of a CFL:
example
 Let L = language of binary palindromes s.t., substitutions for 0
and 1 are defined as follows:
 s(0) = {anbn | n ≥1}, s(1) = {xx,yy}
 Prove that s(L) is also a CFL.

CFG for L: CFG for s(0): CFG for s(1):

S=> 0S0|1S1| S0=> aS0b | ab S1=> xx | yy

Therefore, CFG for s(L):

S=> S0SS0 | S1 S S1 |
S0=> aS0b | ab
S1=> xx | yy 52
 CFLs are closed under union
Let L1 and L2 be CFLs
To show: L2 U L2 is also a CFL
Let us show by using the result of Substitution

 Make a new language:

 Lnew = {a,b} s.t., s(a) = L1 and s(b) = L2
==> s(Lnew) == same as == L1 U L2

 A more direct, alternative proof

 Let S1 and S2 be the starting variables of the
grammars for L1 and L2 53
 CFLs are closed under
concatenation
 Let L1 and L2 be CFLs
Let us show by using the result of Substitution

 Make Lnew= {ab} s.t.,

s(a) = L1 and s(b)= L2
==> L1 L2 = s(Lnew)

 A proof without using substitution?

54
CFLs are closed under
Kleene Closure
 Let L be a CFL

 Let Lnew = {a}* and s(a) = L1

 Then, L* = s(Lnew)

55
 We won’t use substitution to prove this result

CFLs are closed under

Reversal
 Let L be a CFL, with grammar
G=(V,T,P,S)
 For LR, construct GR=(V,T,PR,S) s.t.,
 If A==>  is in P, then:
 A==> R is in PR

 (that is, reverse every production)

56
 Some negative closure results

CFLs are not closed under

Intersection
 Existential proof:
 L1 = {0n1n2i | n≥1,i≥1}
 L2 = {0i1n2n | n≥1,i≥1}
 Both L1 and L2 are CFLs
 Grammars?
 But L1  L2 cannot be a CFL
 Why?
 We have an example, where intersection is
not closed.
 Therefore, CFLs are not closed under
intersection 57
 Some negative closure results

CFLs are not closed under

complementation
 Follows from the fact that CFLs are not
closed under intersection

 L1  L2 = L1 U L2
Logic: if CFLs were to be closed under complementation
 the whole right hand side becomes a CFL (because
CFL is closed for union)
 the left hand side (intersection) is also a CFL
 but we just showed CFLs are
NOT closed under intersection!
 CFLs cannot be closed under complementation.
58
 Some negative closure results

CFLs are not closed under

difference
 Follows from the fact that CFLs are not
closed under complementation

 Because, if CFLs are closed under

difference, then:
 L = ∑* - L
 So L has to be a CFL too
 Contradiction

59
Decision Properties
 Emptiness test
 Generating test
 Reachability test
 Membership test
 PDA acceptance

60
“Undecidable” problems for
CFL
 Is a given CFG G ambiguous?
 Is a given CFL inherently ambiguous?
 Is the intersection of two CFLs empty?
 Are two CFLs the same?
 Is a given L(G) equal to ∑*?

61
Summary
 Normal Forms
 Chomsky Normal Form
 Griebach Normal Form
 Useful in proroving P/L
 Pumping Lemma for CFLs
 Main difference: z=uviwxiy
 Closure properties
 Closed under: union, concatentation, reversal, Kleen
closure, homomorphism, substitution
 Not closed under: intersection, complementation,
difference

62