Parsing

Pushdown Automata (PDAs) and Parsing for CFLs
Pushdown Automata: A pushdown automaton (PDA) M is a generalization of an NFA that also includes
a stack. That is, M = (Q, Σ, Γ, δ, q0 , F ) where Γ, the alphabet of symbols allowed on the stack and δ is a
finite collection of transition rules that say how the stack and state can change based on the current scanned
symbol and what is on the top of the stack. (PDAs are the machine model that exactly recognizes CFLs
and they need to be nondeterministic to do so. The deterministic version only recognizes a limited subset of
CFLs.)
While the formal definition as given in Chapter 2 of Sipser’s text only allows the PDA to examine the top
symbol on the stack, we allow PDA rules to depend on whether longer strings are on the stack; as noted in
Chapter 2, we can simulate such rules with a PDA having extra states. Therefore our PDA diagrams have
transitions of the following form for any a ∈ Σ ∪ {ε}, and any u, v ∈ Γ∗ ,
which means that if PDA is in state p, the next scanned symbol is a, and the top portion of the stack matches
the string u, then the PDA can read a, replace u on the top of the stack with v and go to state q.
We will show that, via two different methods called Top-Down Parsing and Bottom-Up Parsing, given any
CFG G there is a PDA M such that L(M ) = L(G). The other direction, showing that for every PDA there
is a CFG G such that L(G) = L(M ) is much trickier. The textbook gives a very nice proof of that fact, but
since PDAs are not a main focus of this course we will only discuss the parsing side since it is related to
practical application.
Let G be a context-free grammar with start symbol S and rules A1 → w1 , A2 → w2 , . . ., Ar → wr . (Of

course, S will typically be at least one of the Ai .)
Top-Down Parsing In top-down parsing, the algorithm marks the bottom of the stack with a special $
symbol (which is not a symbol in G) and puts S on the stack. The stack is used to store (the unmatched
portion) of a string generated by the grammar that it matches with the input. The top-down parser nonde-
terministically guesses which rules to apply without even looking at the input. When the top of the stack is
an input symbol a, it must match with the next symbol of the input. When it is a variable A, the algorithm
ignores the input and replaces A with the RHS of some rule it is associated with. The parser accepts if the
input is fully matched when the stack only has the bottom-of-stack symbol on it. Here it is in PDA diagram
form:
1
Bottom-Up Parsing In bottom-up parsing, the algorithm marks the bottom of the stack with $ and works
from the input string w, pushing symbols of w onto the stack and applying rules backwards (from right to
left) on the stack, eventually trying to produce a stack that only has the start symbol S above the bottom-
of-stack symbol. Since pushing the symbols of w onto the stack one-by-one means that when we pop them,
they will come out in reverse order. Therefore, when we apply rules from right to left, we need to begin with
the reverse of the right-side. Bottom-up parsing is also nondeterministic since the reverses of more than one
rule may be possible at a given step and one may also choose to transfer an input symbol to the stack. In
practice, programming language grammars are designed so that in bottom-up parsing, it is clear what to do
next. Here it is in PDA diagram form:
An Example: Consider the grammar that generates matched parentheses: S → (S), S → SS, S → ε.
We won’t draw the diagrams, but note that the reverse of (S) is )S( since the individual symbols in the
reverse are not flipped.
Here the parses for the string (()()):
Top-Down Parse Bottom-Up Parse
State Input Stack State Input Stack
q0 (()()) ε q0 (()()) ε
q1 (()()) S$ q1 (()()) $
q1 (()()) (S)$ q1 ()()) ($
q1 ()()) S)$ q1 )()) (($
q1 ()()) SS)$ q1 )()) S(($
q1 ()()) (S)S)$ q1 ()) )S(($
q1 )()) S)S)$ q1 ()) S($
q1 )()) )S)$ q1 )) (S($
q1 ()) S)$ q1 )) S(S($
q1 ()) (S))$ q1 ) )S(S($
q1 )) S))$ q1 ) SS($
q1 )) ))$ q1 ) S($
q1 ) )$ q1 ε )S($
q1 ε $ q1 ε S$
q2 ε ε q2 ε ε

Parsing

Uploaded by

Copyright:

Available Formats

Parsing

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Parsing

Uploaded by

Copyright:

Available Formats

Pushdown Automata (PDAs) and Parsing for CFLs

Let G be a context-free grammar with start symbol S and rules A1 → w1 , A2 → w2 , . . ., Ar → wr . (Of

You might also like