CS411-2010F-08 Context-Free Grammars 1

08-0: Context-Free Grammars

• Set of Terminals (Σ)

• Set of Non-Terminals

• Set of Rules, each of the form:

<Non-Terminal> → <Terminals & Non-Terminals>
• Special Non-Terminal – Initial Symbol
08-1: Generating Strings with CFGs

• Start with the initial symbol

• Repeat:
• Pick any non-terminal in the string
• Replace that non-terminal with the right-hand side of some rule that has that non-terminal as a left-hand
side
Until all elements in the string are terminals

08-2: CFG Example

S → aS
S → Bb
B → cB
B→ǫ
Generating a string:
S replace S with aS
aS replace S wtih Bb
aBb replace B wtih cB 08-3: CFG Example
acBb replace B wtih ǫ
acb Final String
S → aS
S → Bb
B → cB
B→ǫ
Generating a string:
S replace S with aS
aS replace S wtih aS
aaS replace S wtih Bb
aaBb replace B wtih cB 08-4: CFG Example
aacBb replace B wtih cB
aaccBb replace B wtih ǫ
aaccb Final String
S → aS
S → Bb
B → cB
B→ǫ
Regular Expression equivalent to this CFG:
08-5: CFG Example
CS411-2010F-08 Context-Free Grammars 2

S → aS
S → Bb
B → cB
B→ǫ
Regular Expression equivalent to this CFG:
a ∗ c∗ b
08-6: CFG Example
CFG for L = {0n 1n : n > 0}
08-7: CFG Example
CFG for L = {0n 1n : n > 0}
S → 0S1 or S → 0S1|01
S → 01

(note – can write:

A→α
A→β
as
A → α|β )
(examples: 01, 0011, 000111) 08-8: CFG Formal Definition
G = (V, Σ, R, S)

• V = Set of symbols, both terminals & non-terminals

• Σ ⊂ V set of terminals (alphabet for the language being described)
• R ⊂ ((V − Σ) × V ∗ ) Finite set of rules
• S ∈ (V − Σ) Start symbol

08-9: Derivation
A Derivation is a listing of how a string is generated – showing what the string looks like after every replacement.
S → AB
A → aA|ǫ
B → bB|ǫ
S ⇒ AB
⇒ aAB
⇒ aAbB
⇒ abB
⇒ abbB
⇒ abb
08-10: Parse Tree
A Parse Tree is a graphical representation of a derivation.
CS411-2010F-08 Context-Free Grammars 3

A B

a A b B

S ⇒ AB ε b B
⇒ aAB
⇒ aAbB ε
⇒ abB
⇒ abbB
⇒ abb
08-11: Parse Tree
A Parse Tree is a graphical representation of a derivation.
S

A B

a A b B

A ε

S ⇒ AB a A
⇒ AbB
⇒ aAbB ε
⇒ aaAbB
⇒ aaAb
⇒ aab
08-12: Fun with CFGs

• Create a Context-Free Grammar for all strings over a,b which contain the substring “aba”

08-13: Fun with CFGs

• Create a Context-Free Grammar for all strings over a,b which contain the substring “aba”

S → AabaA
A →aA
A →bA
A→ǫ
08-14: LCF G
The Context-Free Languages, LCF G , is the set of all languages that can be described by some CFG:

• LCF G = {L : ∃ CFG G ∧ L[G] = L}

We already know LCF G 6⊆ LREG (why)?

• LREG ⊂ LCF G ?
CS411-2010F-08 Context-Free Grammars 4

08-15: LREG ⊆ LCF G

We will prove LREG ⊆ LCF G in two different ways:

• Prove by induction that, given any regular expression r, we create a CFG G such that L[G] = L[r]
• Given any NFA M , we create a CFG G such that L[G] = L[M ]

08-16: LREG ⊆ LCF G

• To Prove: Given any regular expression r, we can create a CFG G such that L[G] = L[r]
• By induction on the structure of r

08-17: LREG ⊆ LCF G

Base Cases:

• r = a, a ∈ Σ

08-18: LREG ⊆ LCF G

Base Cases:

• r = a, a ∈ Σ

S→a
08-19: LREG ⊆ LCF G
Base Cases:

• r=ǫ

08-20: LREG ⊆ LCF G

Base Cases:

• r=ǫ

S→ǫ
08-21: LREG ⊆ LCF G
Base Cases:

• r=∅

08-22: LREG ⊆ LCF G

Base Cases:

• r=∅

S → SS
08-23: LREG ⊆ LCF G
Recursive Cases:

• r = (r1 r2 )

L[G1 ] = L[r1 ], Start symbol of G1 = S1

L[G2 ] = L[r2 ], Start symbol of G2 = S2
08-24: LREG ⊆ LCF G
Recursive Cases:
CS411-2010F-08 Context-Free Grammars 5

• r = (r1 r2 )

L[G1 ] = L[r1 ], Start symbol of G1 = S1

L[G2 ] = L[r2 ], Start symbol of G2 = S2
G = all rules from G1 and G2 , plus plus new non-terminal S, and new rule:
S → S1 S2
New start symbol S
08-25: LREG ⊆ LCF G
Recursive Cases:

• r = (r1 + r2 )

L[G1 ] = L[r1 ], Start symbol of G1 = S1

L[G2 ] = L[r2 ], Start symbol of G2 = S2
08-26: LREG ⊆ LCF G
Recursive Cases:

• r = (r1 + r2 )

L[G1 ] = L[r1 ], Start symbol of G1 = S1

L[G2 ] = L[r2 ], Start symbol of G2 = S2
G = all rules from G1 and G2 , plus new non-terminal S, and new rules:
S → S1
S → S2
Start symbol = S
08-27: LREG ⊆ LCF G
Recursive Cases:

• r = (r1∗ )

L[G1 ] = L[r1 ], Start symbol of G1 = S1

08-28: LREG ⊆ LCF G
Recursive Cases:

• r = (r1∗ )

L[G1 ] = L[r1 ], Start symbol of G1 = S1

G = all rules from G1 , plus new non-terminal S, and new rules:
S → S1 S
S→ǫ
Start symbol = S
(Example) 08-29: LREG ⊆ LCF G II

• Given any NFA

• M = (K, Σ, ∆, s, F )
CS411-2010F-08 Context-Free Grammars 6

• Create a grammar
• G = (V, Σ, R, S) such that L[G] = L[M ]
• Idea: Derivations like “backward NFA configurations”, showing past instead of future

• Example for all strings over {a, b} that contain aa, not bb

08-30: LREG ⊆ LCF G II

• M = (K, Σ, ∆, s, F )
• G = (V, Σ′ , R, S)
• V
• Σ′
• R
• S

08-31: LREG ⊆ LCF G II

• M = (K, Σ, ∆, s, F )
• G = (V, Σ′ , R, S)
S
• V =K Σ
• Σ′ = Σ
• R = {(q1 → aq2 ) : q1 , q2 ∈ K (and V ),
a ∈ Σ, ((q1 , a), q2 ) ∈ ∆}∪
{(q → ǫ) : q ∈ F }
• S=s

(Example)
08-32: CFG – Ambiguity

• A CFG is ambiguous if there exists at least one string generated by the grammar that has ¿ 1 different parse tree
S → AabaA
A →aA
• Previous CFG is ambiguous (examples)
A →bA
A→ǫ

08-33: CFG – Ambiguity

• Consider the following CFG:

E → E + E|E − E|E ∗ E|N

N → 0|1|2|3|4|5|6|7|8|9

• Is this CFG ambiguous?

• Why is this a problem?
CS411-2010F-08 Context-Free Grammars 7

08-34: CFG – Ambiguity

E → E + E|E − E|E ∗ E|N
N → 0|1|2|3|4|5|6|7|8|9

E
E
E * E
E + E

E E
E + E N
N *

3 N N N N 5

4 5 3 4
08-35: CFG – Ambiguity
E → E + E|E − E|E ∗ E|N
N → 0|1|2|3|4|5|6|7|8|9

• If all we care about is removing ambiguity, there is a (relatively) easy way to make this unambiguous (make all
operators right-associative)

08-36: CFG – Ambiguity

E → E + E|E − E|E ∗ E|N
N → 0|1|2|3|4|5|6|7|8|9
Non-ambiguous:
E → N |N + E|N − E|N ∗ E
N → 0|1|2|3|4|5|6|7|8|9

• If we were writing a compiler, would this be a good CFG?

• How can we get correct associativity

08-37: CFG – Ambiguity

• Ambiguous:
E → E + E|E − E|E ∗ E|N
N → 0|1|2|3|4|5|6|7|8|9
• Unambiguous:
E → E + T |E − T |T
T → T ∗ N |N
N → 0|1|2|3|4|5|6|7|8|9
Can add parentheses, other operators, etc. (More in Compilers)
08-38: Fun with CFGs

• Create a CFG for all strings over {(,)} that form balanced parenthesis
• ()
CS411-2010F-08 Context-Free Grammars 8

• ()()
• (()())((()()()))
• ((((()))))

08-39: Fun with CFGs

• Create a CFG for all strings over {(,)} that form balanced parenthesis
S → (S)
S → SS
S→ǫ
• Is this grammar ambiguous?

08-40: Fun with CFGs

• Create a CFG for all strings over {(,)} that form balanced parenthesis
S → (S)
S → SS
S→ǫ
• Is this grammar ambiguous?
• YES! (examples)

08-41: Fun with CFGs

• Create an unambiguous CFG for all strings over {(,)} that form balanced parenthesis

08-42: Fun with CFGs

• Create an unambiguous CFG for all strings over {(,)} that form balanced parenthesis

S → AS
S→ǫ
A → (S)
08-43: Ambiguous Languages

• A language L is ambiguous if all CFGs G that generate it are ambiguous

• Example:

• L1 = {ai bi cj dj |i, j > 0}

• L2 = {ai bj cj di |i, j > 0}
• L3 = L1 ∪ L2

• L3 is inherently ambiguous

(Create a CFG for L3 ) 08-44: Ambiguous Languages

• L1 = {ai bi cj dj |i, j > 0}

• L2 = {ai bj cj di |i, j > 0}
• L3 = L1 ∪ L2
CS411-2010F-08 Context-Free Grammars 9

S → S1 |S2
S1 → AB
A → aAb|ab
B → cBd|cd
S2 → aS2 d|aCd
C → bCc|bc
What happens when i = j? 08-45: (More) Fun with CFGs

• Create an CFG for all strings over {a, b} that have the same number of a’s as b’s (can be ambiguous)

08-46: (More) Fun with CFGs

• Create an CFG for all strings over {a, b} that have the same number of a’s as b’s (can be ambiguous)

S → aSb
S → bSa
S → SS
S→ǫ
08-47: (More) Fun with CFGs

• Create an CFG for L = {wwR : w ∈ (a + b)∗ }

08-48: (More) Fun with CFGs

• Create an CFG for L = {wwR : w ∈ (a + b)∗ }

S → aSa
S → bSb 08-49: (More) Fun with CFGs
S→ǫ

• Create an CFG for all palindromes over {a, b}. That is, create a CFG for:
• L = {w : w ∈ (a + b)∗ , w = wR }

08-50: (More) Fun with CFGs

• Create an CFG for all palindromes over {a, b}. That is, create a CFG for:
• L = {w : w ∈ (a + b)∗ , w = wR }

S → aSa
S → bSb
S →ǫ 08-51: (More) Fun with CFGs
S →a
S →b

• Create an CFG for L = {ai bj ck : j > i + k}

08-52: (More) Fun with CFGs

• Create an CFG for L = {ai bj ck : j > i + k}

HINT: We may wish to break this down into 3 different langauges ...
08-53: (More) Fun with CFGs
CS411-2010F-08 Context-Free Grammars 10

• Create an CFG for L = {ai bj ck : j > i + k}

S → ABC
A → aAb
A→ǫ
B → bB
B→b
C → bCc|ǫ
08-54: (More) Fun with CFGs

• Create an CFG for all strings over {0, 1} that have the an even number of 0’s and an odd number of 1’s.
• HINT: It may be easier to come up with 4 CFGs – even 0’s, even 1’s, odd 0’s odd 1’s, even 0’s odd 1’s, odd
1’s, even 0’s – and combine them ...

08-55: (More) Fun with CFGs

• Create an CFG for all strings over {0, 1} that have the an even number of 0’s and an odd number of 1’s.

S1 = Even 0’s Even 1’s

S2 = Even 0’s Odd 1’s
S3 = Odd 0’s Even 1’s
S4 = Odd 0’s Odd 1’s

S1 → 0S3 |1S2
S2 → 0S4 |1S1
S3 → 0S1 |1S4
S4 → 0S2 |1S3