Compilers: Lexical Error

Question

This screenshot is token from the book Ullman,

How can following be a lexical error? because "elipseSize" should have a token recorded as an identifier.

Comments

@Arjun

---

@manu00x 

author is clearly saying "the use of identifier elipseSize" that means it detected as identifier already.

But I also don't have a good reason why it is a lexical error, but Lexical analyzer also use symbol table while processing and I think some DFA also, so maybe by the help of DFA it will recognize that something is a valid token or not and so misspell of id, keyword will become a lexical error.[not so sure]

---

Lexical Analyzer is not worried about the usage of tokens, its job is only to read source program character by character and if it matches a lexeme, then it creates a token and stores it in the symbol table.

---

You must find "How it matches lexeme to a pattern(description) to form a valid token?"

---

yes it reads the token charecter by charecter it doesnt care about anything else even if u have recorded it once as ellipsesize and next time there is a spelling mistake it makes a new entry for this new one so no lexical error

---

this is clearly a wrong definition. elipseSize and ellipseSize both are valid.

Answer 1

Compiler is all about impementation.

If an automata is already defined for your statement, then it will be counted as a valid token ,Otherwise it will show lexical error.

Here, mis-spelling of keywords or identifier means error while identifying lexemes(longest common sequence).