Answer is b) A syntax analyzer or parser takes the input from a lexical analyzer in the form of token streams. The parser analyzes the source code (token stream) against the production rules to detect any errors in the code.
So, we'r concerned with the entities not the production or any for that case 'rule'.