COP 4020 Programming Languages Computing LL1 parsing table

COP 4020 Programming Languages Computing LL(1) parsing table Prof. Xin Yuan

Overview n n LL(1) parsing in action (Top-down parsing) Computing LL(1) parsing table 9/25/2020 COP 4020 Spring 2014 2

n Using the parsing table, the predictive parsing program works like this: ¨ ¨ ¨ A stack of grammar symbols ($ on the bottom) A string of input tokens ($ at the end) A parsing table, M[NT, T] of productions Algorithm: put ‘$ Start’ on the stack ($ is the end of input string). 1) if top == input == $ then accept 2) if top == input then pop top of the stack; advance to next input symbol; goto 1; 3) if top is nonterminal if M[top, input] is a production then replace top with the production; goto 1 else error 4) else error

¨ Example: (1) E->TE’ (2) E’->+TE’ (3) E’-> (4) T->FT’ (5) T’->*FT’ (6) T’-> (7) F->(E) (8) F->id Stack $E $E’T’F $E’T’id $E’T’ …. . . E E’ T T’ F id (1) + * ( (1) (2) (4) input id+id*id$ +id*id$ $ (3) (6) (4) (6) (8) ) (5) (7) production E->TE’ T->FT’ F->id This produces leftmost derivation: E=>TE’=>FT’E’=>id. T’E’=>…. =>id+id*id

(1) E->TE’ (2) E’->+TE’ (3) E’-> (4) T->FT’ (5) T’->*FT’ (6) T’-> (7) F->(E) (8) F->id E E’ T T’ F id (1) + * ( (1) (2) (4) ) $ (3) (6) (4) (6) (5) (8) (7) How to compute the parsing table for LL(1) grammar? Key: We need to make choice for every production When can E be expanded with production E->TE’? Intuitively, any token that can be the first token by expanding TE’. This should include all first token by expanding T, what are they? What if T can derive empty string ( ) , we should also include the first token that can be derived from E’ What if E’ can also derive empty string? We should all possible tokens that can potentially follow E? When should E’ be expanded with production E’-> ?

(1) E->TE’ (2) E’->+TE’ (3) E’-> (4) T->FT’ (5) T’->*FT’ (6) T’-> (7) F->(E) (8) F->id E E’ T T’ F id (1) + * ( (1) (2) (4) ) $ (3) (6) (4) (6) (5) (8) (7) How to compute the parsing table for LL(1) grammar? Intuition: We need to make choice for every production • • Case 1 (easy): E’->+TE’: expand for all tokens that can be the first token after expanding the right hand side of the production (expanding +TE’) Case 1 (harder): E->TE’: expand for all tokens that can be the first token after expanding TE’ We call this First set. Case 2: E’-> : no first token? Whenever we see a token that can potential follow E’ in a sentential form. (Follow set)

n For a production that can derive a string of tokens, find all possible first tokens. ¨ n A production N -> X Y Z should be expanded when the token can be the first of X Y Z (after derivation): First(X Y Z). For a production that can derive empty string, find all possible tokens that can follow the nonterminal. ¨ When should we expand with E’-> n ? Anything token that can potentially follow E’: Follow(E’).

n First set and follow set ¨ First( ): Here, is a string of symbols. The set of terminals that begin strings derived from a. n If a is empty string or generates empty string, then empty string is in First( ). ¨ Follow(A): Here, A is a nonterminal symbol. Follow(A) is the set of terminals that can immediately follow A in a sentential form. ¨ Example: S->i. Et. S | i. Et. Se. S|a E->b First(a) = ? , First(i. Et. S) = ? , First(S) = ? Follow(E) = ? Follow(S) = ?

n Compute FIRST(X) If a is a terminal then FIRST(a) = {a} (Case 1) ¨ If X-> , add to FIRST(X). (Case 2) ¨ If and add every none FIRST( ) to FIRST(X). If , add to FIRST(X). (Case 3) ¨ ¨ FIRST( E->TE’ E’->+TE’| T->FT’ T’->*FT’ | F->(E) | id ): similar to the third case. FIRST(E) = ? FIRST(E’)= ? FIRST(T) = ? FIRST(T’) = ? FIRST(F) = ? in

Computing first set E->TE’ E’->+TE’| T->FT’ T’->*FT’ | F->(E) | id 9/25/2020 FIRST(E) = {(, id} FIRST(E’)={+, } FIRST(T) = {(, id} FIRST(T’) = {*, } FIRST(F) = {(, id} COP 4020 Spring 2014 10

n Compute Follow(A) If S is the start symbol, add $ to Follow(S). ¨ If A-> B , add First( )-{ } to Follow(B). ¨ If A-> B or A-> B and => , add Follow(A) to Follow(B). ¨ ¨ Note: you are looking at the right hand side of productions!!! E->TE’ E’->+TE’| T->FT’ T’->*FT’ | F->(E) | id First(E) = {(, id}, Follow(E)={), $} First(E’)={+, e}, Follow(E’) = {), $} First(T) = {(, id}, Follow(T) = {+, ), $} First(T’) = {*, e}, Follow(T’) = {+, ), $} First(F) = {(, id}, Follow(F) = {*, +, ), $}

n How to construct the parsing table? ¨ With first(a) and follow(A), we can build the parsing table. For each production A-> : n n Add A-> to M[A, t] for each t in First( ). If First( ) contains empty string ¨ ¨ n n Add A-> to M[A, t] for each t in Follow(A) if $ is in Follow(A), add A-> to M[A, $] Make each undefined entry of M error. Construct parsing table for the following grammar: E->TE’ E’->+TE’| T->FT’ T’->*FT’ | F->(E) | id First(E) = {(, id}, Follow(E)={), $} First(E’)={+, e}, Follow(E’) = {), $} First(T) = {(, id}, Follow(T) = {+, ), $} First(T’) = {*, e}, Follow(T’) = {+, ), $} First(F) = {(, id}, Follow(F) = {*, +, ), $}

n LL(1) grammar: ¨ A grammar whose parsing table has no multiply-defined entries is a LL(1) grammar. n use one input symbol of lookahead at each step to make a parsing decision. No ambiguous or left-recursive grammar can be LL(1) ¨ A grammar is LL(1) iff for each set of A productions, where ¨ The following conditions hold: ¨