LL(1) parsing

LL(1) parsing

Top-down parsing

A top-down parsing algorithm parses an input string in such a way that the implied traversal of the parse tree occurs from the root to the leaves.

LL(1) parsing

Top-down parsing

A top-down parsing algorithm parses an input string in such a way that the implied traversal of the parse tree occurs from the root to the leaves. Top-down parsers come in two forms: backtracking parsers and predictive parsers.

LL(1) parsing

Top-down parsing

A top-down parsing algorithm parses an input string in such a way that the implied traversal of the parse tree occurs from the root to the leaves. Top-down parsers come in two forms: backtracking parsers and predictive parsers. A predictive parser attempts to predict the next construction using one or more lookahead tokens.

LL(1) parsing

Top-down parsing

A top-down parsing algorithm parses an input string in such a way that the implied traversal of the parse tree occurs from the root to the leaves. Top-down parsers come in two forms: backtracking parsers and predictive parsers. A predictive parser attempts to predict the next construction using one or more lookahead tokens. Two well-known top-down parsing methods are recursive-descent parsing and LL(1) parsing.

LL(1) parsing

Top-down parsing

A top-down parsing algorithm parses an input string in such a way that the implied traversal of the parse tree occurs from the root to the leaves. Top-down parsers come in two forms: backtracking parsers and predictive parsers. A predictive parser attempts to predict the next construction using one or more lookahead tokens. Two well-known top-down parsing methods are recursive-descent parsing and LL(1) parsing. Recursive descent parsing is the most suitable method for a handwritten parser.

LL(1) parsing

LL(1) parsing

The rst L in LL(1) refers to the fact that the input is processed from left to right.

LL(1) parsing

LL(1) parsing

The rst L in LL(1) refers to the fact that the input is processed from left to right. The second L refers to the fact that LL(1) parsing determines a leftmost derivation for the input string.

LL(1) parsing

LL(1) parsing

The rst L in LL(1) refers to the fact that the input is processed from left to right. The second L refers to the fact that LL(1) parsing determines a leftmost derivation for the input string. The 1 in parentheses implies that LL(1) parsing uses only one symbol of input to predict the next grammar rule that should be used.

LL(1) parsing

LL(1) parsing - Example 1

We start by considering the following grammar that generates strings of balanced parenthesis: S (S )S |

LL(1) parsing

LL(1) parsing - Example 1

We start by considering the following grammar that generates strings of balanced parenthesis: S (S )S | We will assume that $ marks the bottom of a stack and the end of input.

LL(1) parsing

LL(1) parsing - Example 1

We start by considering the following grammar that generates strings of balanced parenthesis: S (S )S | We will assume that $ marks the bottom of a stack and the end of input.
Parsing action of an LL(1) parser: 1 2 3 4 5 6 Parsing Stack $S $ S )S ( $ S )S $S ) $S $ Input ()$ ()$ )$ )$ $ $ Action S (S )S match S match S accept

LL(1) parsing

LL(1) parsing - Example 1

We start by considering the following grammar that generates strings of balanced parenthesis: S (S )S | We will assume that $ marks the bottom of a stack and the end of input.
Parsing action of an LL(1) parser: 1 2 3 4 5 6 Parsing Stack $S $ S )S ( $ S )S $S ) $S $ Input ()$ ()$ )$ )$ $ $ Action S (S )S match S match S accept $ S

The LL(1) parsing table M [N , T ]:

M [N , T ] S

( S (S )S

) S

LL(1) parsing

LL(1) parsing tables

We use the parsing table to decide which decision should be made if a given nonterminal N is at the top of the parsing stack, based on the current input symbol T .

LL(1) parsing

LL(1) parsing tables

We use the parsing table to decide which decision should be made if a given nonterminal N is at the top of the parsing stack, based on the current input symbol T . We add production choices to the LL(1) parsing table as follows:

LL(1) parsing

LL(1) parsing tables

We use the parsing table to decide which decision should be made if a given nonterminal N is at the top of the parsing stack, based on the current input symbol T . We add production choices to the LL(1) parsing table as follows:
1. If A is a production choice, and there is a derivation a , where a is a token, then we add A to the table entry M [A, a].

LL(1) parsing

LL(1) parsing tables

We use the parsing table to decide which decision should be made if a given nonterminal N is at the top of the parsing stack, based on the current input symbol T . We add production choices to the LL(1) parsing table as follows:
1. If A is a production choice, and there is a derivation a , where a is a token, then we add A to the table entry M [A, a]. 2. If A is a production choice, and there are derivations and S $ Aa , where S is the start symbol and a is a token (or $), then we add A to the table entry M [A, a].

LL(1) parsing

LL(1) parsing tables

The idea behind these rules are as follows:

LL(1) parsing

LL(1) parsing tables

The idea behind these rules are as follows:

1. In rule 1, given a token a in the input, we wish to select a rule A if can produce an a for matching.

LL(1) parsing

LL(1) parsing tables

The idea behind these rules are as follows:

1. In rule 1, given a token a in the input, we wish to select a rule A if can produce an a for matching. 2. In rule 2, if A derives the empty string (via A ), and if a is a token that can legally come after A in a derivation, then we want to select A to make A disappear.

LL(1) parsing

LL(1) parsing tables

The idea behind these rules are as follows:

1. In rule 1, given a token a in the input, we wish to select a rule A if can produce an a for matching. 2. In rule 2, if A derives the empty string (via A ), and if a is a token that can legally come after A in a derivation, then we want to select A to make A disappear.

These rules are dicult to implement directly, so we will develop algorithms involving rst and follow sets.

LL(1) parsing

LL(1) parsing tables

The idea behind these rules are as follows:

1. In rule 1, given a token a in the input, we wish to select a rule A if can produce an a for matching. 2. In rule 2, if A derives the empty string (via A ), and if a is a token that can legally come after A in a derivation, then we want to select A to make A disappear.

These rules are dicult to implement directly, so we will develop algorithms involving rst and follow sets. A grammar is an LL(1) grammar if the associated LL(1) parsing table has at most one production in each table entry.

LL(1) parsing

First and Follow sets

Before we give the precise denitions of First and Follow Sets, we show how to use it in the construction of LL(1) parsing tables.

LL(1) parsing

First and Follow sets

Before we give the precise denitions of First and Follow Sets, we show how to use it in the construction of LL(1) parsing tables. Repeat the following two steps for each nonterminal A and each production A :

LL(1) parsing

First and Follow sets

Before we give the precise denitions of First and Follow Sets, we show how to use it in the construction of LL(1) parsing tables. Repeat the following two steps for each nonterminal A and each production A :
1. For each token a in First(), add A to the entry M [A, a].

LL(1) parsing

First and Follow sets

Before we give the precise denitions of First and Follow Sets, we show how to use it in the construction of LL(1) parsing tables. Repeat the following two steps for each nonterminal A and each production A :
1. For each token a in First(), add A to the entry M [A, a]. 2. If is in First(), for each element a of Follow(A) (where a is a token or a is $), add A to M [A, a].

LL(1) parsing

LL(1) parsing table construction with First and Follow sets

Now consider the grammar S ( S ) S |

LL(1) parsing

LL(1) parsing table construction with First and Follow sets

Now consider the grammar S ( S ) S | In this case we have that First( (S )S ) = { ( } First() = {} Follow(S ) = { ), $ }

LL(1) parsing

LL(1) parsing table construction with First and Follow sets

Now consider the grammar S ( S ) S | In this case we have that First( (S )S ) = { ( } First() = {} Follow(S ) = { ), $ } Thus we get the following LL(1) parsing table: M [N , T ] ( ) $ S S (S )S S S

LL(1) parsing

First sets - denition

Let X be a grammar symbol (a terminal or nonterminal) or . Then the set First(X ) consisting of terminals, and possibly , is dened as follows:

LL(1) parsing

First sets - denition

Let X be a grammar symbol (a terminal or nonterminal) or . Then the set First(X ) consisting of terminals, and possibly , is dened as follows:
1. If X is a terminal or , First(X ) = {X }.

LL(1) parsing

First sets - denition

Let X be a grammar symbol (a terminal or nonterminal) or . Then the set First(X ) consisting of terminals, and possibly , is dened as follows:
1. If X is a terminal or , First(X ) = {X }. 2. If X is a n nonterminal, then for each production choice X X1 X2 ...Xn , First(X ) contains First(X1 ){}. If also for some i < n, all the sets First(X1 ),...,First(Xi ) contains , then First(X ) contains First(Xi +1 ){}. If all the sets First(X1 ),...,First(Xn ) contains , then First(X ) also contains .

LL(1) parsing

First sets - denition continue

We now dene First() for any string = X1 X2 ...Xn (a string of terminals and nonterminals) as follows:

LL(1) parsing

First sets - denition continue

We now dene First() for any string = X1 X2 ...Xn (a string of terminals and nonterminals) as follows:
1. First() contains First(X1 ){}.

LL(1) parsing

First sets - denition continue

We now dene First() for any string = X1 X2 ...Xn (a string of terminals and nonterminals) as follows:
1. First() contains First(X1 ){}. 2. For each i = 2, ..., n, if First(Xk ) contains for all k = 1, ..., i 1, then First() contains First(Xi ){}.

LL(1) parsing

First sets - denition continue

We now dene First() for any string = X1 X2 ...Xn (a string of terminals and nonterminals) as follows:
1. First() contains First(X1 ){}. 2. For each i = 2, ..., n, if First(Xk ) contains for all k = 1, ..., i 1, then First() contains First(Xi ){}. 3. Finally, if for all i = 1, ..., n, First(Xi ) contains , then First() contains .

LL(1) parsing

Algorithm for computing First(A) for nonterminals A

for all nonterminals A do First(A):={}; while there are changes to any First(A) do for each production choice A X1 ...Xn do k := 1; Continue:=true; while Continue = true and k <= n do add First(Xk ){} to First(A); if not in First(Xk ) then Continue:=false; k := k + 1; if Continue = true then add to First(A);

LL(1) parsing

Algorithm for computing First(A) for nonterminals A

for all nonterminals A do First(A):={}; while there are changes to any First(A) do for each production choice A X1 ...Xn do k := 1; Continue:=true; while Continue = true and k <= n do add First(Xk ){} to First(A); if not in First(Xk ) then Continue:=false; k := k + 1; if Continue = true then add to First(A); Simplied algorithm for First Sets in the absence of -productions:

LL(1) parsing

Algorithm for computing First(A) for nonterminals A

for all nonterminals A do First(A):={}; while there are changes to any First(A) do for each production choice A X1 ...Xn do k := 1; Continue:=true; while Continue = true and k <= n do add First(Xk ){} to First(A); if not in First(Xk ) then Continue:=false; k := k + 1; if Continue = true then add to First(A); Simplied algorithm for First Sets in the absence of -productions: for all nonterminals A do First(A):={}; while there are changes to any First(A) do for each production choice A X1 ...Xn do add First(X1 ) to First(A);

LL(1) parsing

First sets example 1

Consider the simple integer expression grammar:
exp exp addop term | term addop + | term term mulop factor | factor mulop factor ( exp ) | number

LL(1) parsing

First sets example 1

Consider the simple integer expression grammar:
exp exp addop term | term addop + | term term mulop factor | factor mulop factor ( exp ) | number Grammar rule exp exp addop term exp term addop + addop term term mulop factor term factor mulop factor ( exp ) factor number First(mulop ) = {} First(factor ) ={(} First(factor ) = {(, number} Pass 1 Pass 2 Pass 3

First(exp ) = {(, number} First(addop ) = {+} First(addop ) = {+, }

First(term)= {(, number}

LL(1) parsing

First sets example 1 continue

Thus: First(exp ) = {(, number} First(term) = {(, number} First(factor ) = {(, number} First(addop ) = {+, } First(mulop ) = {}

LL(1) parsing

First sets example 2

Consider the grammar:

statement if -stmt | other if -stmt if ( exp ) statement else -part else -part else statement | exp 0 | 1

LL(1) parsing

First sets example 2

Consider the grammar:

statement if -stmt | other if -stmt if ( exp ) statement else -part else -part else statement | exp 0 | 1 Grammar rule statement if -stmt statement other if -stmt if ( exp ) statement else -part else -part else statement else -part exp 0 exp 1 Pass 1 Pass 2 First(statement )= {if , other}

First(statement ) = {other} First(if -stmt )= = {if } First(else -part )= = {else} First(else -part )= = {else, } First(exp )={0} First(exp )={0, 1}

LL(1) parsing

First sets example 2 continue

Thus: First(statement ) = {if , other} First(if -stmt ) = {if } First(else -part ) = {else, } First(exp ) = {0, 1}

LL(1) parsing

Follow sets - denition

Given a nonterminal A, the follow set Follow(A), consisting of terminals, and possibly $, is dened as follows:

LL(1) parsing

Follow sets - denition

Given a nonterminal A, the follow set Follow(A), consisting of terminals, and possibly $, is dened as follows:
1. If A is the start symbol, then $ is in Follow(A).

LL(1) parsing

Follow sets - denition

Given a nonterminal A, the follow set Follow(A), consisting of terminals, and possibly $, is dened as follows:
1. If A is the start symbol, then $ is in Follow(A). 2. If there is a production B A , then First( ){} is in Follow(A).

LL(1) parsing

Follow sets - denition

Given a nonterminal A, the follow set Follow(A), consisting of terminals, and possibly $, is dened as follows:
1. If A is the start symbol, then $ is in Follow(A). 2. If there is a production B A , then First( ){} is in Follow(A). 3. If there is a production B A such that is in First( ), then Follow(A) contains Follow(B ).

LL(1) parsing

Algorithm for the computation of Follow Sets

Follow(start-symbol):= {$}; for all nonterminals A = start-symbol do Follow(A):={}; while there are changes to any Follow sets do for each production A X1 ...Xn do for each Xi that is a nonterminal do add First(Xi +1 ...Xn ){} to Follow(Xi ) (* Note: if i = n, then Xi +1 ...Xn = *) if is in First(Xi +1 ...Xn ) then add Follow(A) to Follow(Xi )

LL(1) parsing

Follow sets example 1

We consider again the grammar: exp exp addop term | term addop + | term term mulop factor | factor mulop factor ( exp ) | number

LL(1) parsing

Follow sets example 1

We consider again the grammar: exp exp addop term | term addop + | term term mulop factor | factor mulop factor ( exp ) | number Recall that: First(exp ) = {(, number} First(term) = {(, number} First(factor ) = {(, number} First(addop ) = {+, } First(mulop ) = {}

LL(1) parsing

Follow sets example 1 continue

In the computation of the Follow sets for the grammar we omit the four grammar rule choices that have no possibility of aecting the computation.

LL(1) parsing

Follow sets example 1 continue

In the computation of the Follow sets for the grammar we omit the four grammar rule choices that have no possibility of aecting the computation.
Grammar rule exp exp addop term Pass 1 Follow(exp )= {$, +, } Follow(addop )= {(, number} Follow(term)= {$, +, } Follow(term)= {$, +, , } Follow(mulop )= {(, number} Follow(factor ) = {$, +, , } Follow(exp ) = { $, +, , )} Pass 2 Follow(term)= {$, +, , , )}

exp term term term mulop factor

Follow(factor )= {$, +, , , )}

term factor factor ( exp )

LL(1) parsing

Follow sets example 1 continue

Thus: Follow(exp ) = {$, +, , )} Follow(term) = {$, +, , , )} Follow(factor ) = {$, +, , , )} Follow(addop ) = {(, number} Follow(mulop ) = {(, number}

LL(1) parsing

LL(1) parsing example

statement if -stmt | other if -stmt if ( exp ) statement else -part else -part else statement | exp 0 | 1

LL(1) parsing

LL(1) parsing example

statement if -stmt | other if -stmt if ( exp ) statement else -part else -part else statement | exp 0 | 1 Recall that: First(statement ) = {if , other} First(if -stmt ) = {if } First(else -part ) = {else, } First(exp ) = {0, 1}

LL(1) parsing

LL(1) parsing example

statement if -stmt | other if -stmt if ( exp ) statement else -part else -part else statement | exp 0 | 1 Recall that: First(statement ) = {if , other} First(if -stmt ) = {if } First(else -part ) = {else, } First(exp ) = {0, 1} One can verify that: Follow(statement ) = {$, else} Follow(if -stmt ) = {$, else} Follow(else -part ) = {$, else} Follow(exp ) = { ) }
LL(1) parsing

LL(1) parsing example continue

Recall that the LL(1) parsing table M [N , T ] is constructed by repeating the following two steps for each nonterminal A and each production A :

LL(1) parsing

LL(1) parsing example continue

Recall that the LL(1) parsing table M [N , T ] is constructed by repeating the following two steps for each nonterminal A and each production A :
1. For each token a in First(), add A to the entry M [A, a].

LL(1) parsing

LL(1) parsing example continue

Recall that the LL(1) parsing table M [N , T ] is constructed by repeating the following two steps for each nonterminal A and each production A :
1. For each token a in First(), add A to the entry M [A, a]. 2. If is in First(), for each element a of Follow(A) (where a is a token or a is $), add A to M [A, a].

LL(1) parsing

LL(1) parsing example continue

Using the procedure on the previous slide, we obtain the following table:

LL(1) parsing

LL(1) parsing example continue

Using the procedure on the previous slide, we obtain the following table:
M [N , T ] statement if -stmt if statement if -stmt if -stmt if ( exp ) statement else -part other statement other else 0 1 $

else -part

else -part else statement else -part

else -part

exp

exp 0

exp 1

LL(1) parsing

LL(1) parsing example continue

We notice, that as expected, this grammar is not LL(1), since the entry M[else -part ,else] contains two entries, corresponding to the dangling else ambiguity.

LL(1) parsing

LL(1) parsing example continue

We notice, that as expected, this grammar is not LL(1), since the entry M[else -part ,else] contains two entries, corresponding to the dangling else ambiguity. We could apply the disambiguating rule that would always prefer the rule that generates the current lookahead token over any other (this corresponds to the most closely nested disambiguating rule), and thus the production else -part else statement

LL(1) parsing

LL(1) parsing example continue

We notice, that as expected, this grammar is not LL(1), since the entry M[else -part ,else] contains two entries, corresponding to the dangling else ambiguity. We could apply the disambiguating rule that would always prefer the rule that generates the current lookahead token over any other (this corresponds to the most closely nested disambiguating rule), and thus the production else -part else statement We now show the LL(1) parsing actions for the string if (0) if (1) other else other

LL(1) parsing

LL(1) parsing example continue

We notice, that as expected, this grammar is not LL(1), since the entry M[else -part ,else] contains two entries, corresponding to the dangling else ambiguity. We could apply the disambiguating rule that would always prefer the rule that generates the current lookahead token over any other (this corresponds to the most closely nested disambiguating rule), and thus the production else -part else statement We now show the LL(1) parsing actions for the string if (0) if (1) other else other We use the following abbreviations:
statement = S if -stmt = I else -part = L exp = E if = i else = e other = o

LL(1) parsing

LL(1) parsing example continue

Parsing stack $S $I $LS )E (i $LS )E ( $LS )E $LS )0 $LS ) $LS $LI $LLS )E (i $LLS )E ( $LLS )E $LLS )1 $LLS ) $LLS $LLo $LL $LS e $LS $Lo $L $

i (0) i (1) i (0) i (1) i (0) i (1) (0) i (1) 0) i (1) 0) i (1) ) i (1) i (1) i (1) i (1) (1) 1) 1) )

Input o e o$ o e o$ o e o$ o e o$ o e o$ o e o$ o e o$ o e o$ o e o$ o e o$ o e o$ o e o$ o e o$ o e o$ o e o$ o e o$ e o$ e o$ o$ o$ $ $

Action S I I i(E )S L match match E 0 match match S I I i(E )S L match match E 1 match match S o match LeS match S o match L accept

LL(1) parsing

LL(1) parsing in JFLAP

Finally we consider the LL(1) parsing example discussed in the JFLAP tutorial at http://www.jap.org/tutorial/

LL(1) parsing

LL(1) parsing in JFLAP

Finally we consider the LL(1) parsing example discussed in the JFLAP tutorial at http://www.jap.org/tutorial/ We use JFLAP to calculate First and Follow sets and the LL(1) parse table for the grammar: S aAB b AaAc A B bB B c

LL(1) parsing

LL(1) parsing in JFLAP

Finally we consider the LL(1) parsing example discussed in the JFLAP tutorial at http://www.jap.org/tutorial/ We use JFLAP to calculate First and Follow sets and the LL(1) parse table for the grammar: S aAB b AaAc A B bB B c Now we use JFLAP to parse aacbbcb

LL(1) parsing