RECURSIVE DESCENT PARSING COP 4620 Programming Language Translators

GRAMMARS (SO FAR) Our “model” PL grammar. S → → SL → → E

PARSE TABLE Can parse using this table. Tedious (Both parsing and PT construction)

RECURSIVE DESCENT PARSING • Top-down parsing strategy, for LL(1) grammars. • One procedure per

RECURSIVE DESCENT PARSER proc S; case Next_Token of T_begin : Read(T_begin); SL(); Read(T_end); T_id

RECURSIVE DESCENT PARSER proc SL; S(); Z(); end; proc Z; case Next Token of

RECURSIVE DESCENT PARSER proc E; T(); Y(); end; proc Y; if Next Token =

RECURSIVE DESCENT PARSER proc T; P(); X(); end; T: Could have checked for T_(

RECURSIVE DESCENT PARSER proc P; case Next Token of T_(: Read(T_(); E(); Read(T_)); T_id:

TRACING THE PARSER Not so easy (as table-driven) In table-driven parser, stack keeps all

SUMMARY Recursive Descent Parsing Manually coded parser. PT information built into the code. Suitable

Slides: 11

Download presentation

RECURSIVE DESCENT PARSING COP 4620 – Programming Language. Translators Dr. Manuel E. Bermudez

GRAMMARS (SO FAR) Our “model” PL grammar. S → → SL → → E → → T → → P → → begin SL end {begin} id : = E; {id} SL S {begin, id} E+T {(, id} P*T {(, id} P {(, id} (E) {(} id {id} Modified, LL(1)) grammar. S SL Z E Y T X P → → → → begin SL end id : = E ; S Z T Y + T Y P X * T (E) id {begin} {id} {begin, id} {end} {(, id} {+} {; , )} {(, id} {*} {; , +, )} {(} {id}

PARSE TABLE Can parse using this table. Tedious (Both parsing and PT construction)

RECURSIVE DESCENT PARSING • Top-down parsing strategy, for LL(1) grammars. • One procedure per nonterminal. • Stack contents embedded in recursive call sequence. • Each procedure “commits” to one production, based on the next input symbol, and the select sets. • Good technique for hand-written parsers.

RECURSIVE DESCENT PARSER proc S; case Next_Token of T_begin : Read(T_begin); SL(); Read(T_end); T_id : Read(T_id); Read(T_: =); E(); Read(T_; ); otherwise Error; end; “Next_Token” is the upcoming token. Assume it is initialized. “Read (T_X)” verifies that the upcoming token is X, and consumes it.

RECURSIVE DESCENT PARSER proc SL; S(); Z(); end; proc Z; case Next Token of T_begin, T_id: S(); Z(); T_end: ; otherwise Error; end; SL: Technically, should insist Next Token ∊ { T_begin, T_id}, but S will do that anyway. Checking earlier or later ? Aid error recovery.

RECURSIVE DESCENT PARSER proc E; T(); Y(); end; proc Y; if Next Token = T_+ then Read(T_+); T(); Y(); end; E: Technically, should insist Next Token ∊ { T_id, T_( }, but T will do that anyway. Y: Could have used a case statement

RECURSIVE DESCENT PARSER proc T; P(); X(); end; T: Could have checked for T_( and T_id. . proc X; if Next Token = T_* then Read(T_*); T(); X: Could have used a case statement. end;

RECURSIVE DESCENT PARSER proc P; case Next Token of T_(: Read(T_(); E(); Read(T_)); T_id: Read(T_id); otherwise Error; end;

TRACING THE PARSER Not so easy (as table-driven) In table-driven parser, stack keeps all future (predicted) symbols. Stack (towards the end): Next item is obvious. In RD parser, stack keeps recursive calling sequence.

SUMMARY Recursive Descent Parsing Manually coded parser. PT information built into the code. Suitable for LL(1) grammars.