Regex in Java (Regular Expression and Grammar)

Question

That is my problem: I got an assignment due tomorrow (I have been working on it for the past days) and I think a got half of it, but the other half I just got a slight clue. Basically, I have to write a code where the user gives me some words and their categories (like a dictionary) and a grammar eg:

A: article adjective A  
A: adjective B  
B: noun  

then the user will input a sentence and I will check if the words are in my dictionary, if so I will output their category and "Accept" if it matches the grammar and "Refuse" if it doesn't match.

Now, I know that I probably will have to use regex but I didn't find anything that really helped so far.
I will give a example of input and output:

Input:  
Dictionary:  
word: The  
category: article  
word: big  
category: adjective  
word: blue  
category: adjective  
word: car   
category: noun

Grammar:  
A: article adjective A  
A: adjective B  
B: noun  

Sentences:   
The big blue car  
The car big blue  

Output:  
(first sentence)  
article adjective adjective noun  
Accept  
(second sentence)  
article noun adjective adjective   
Refuse  

So, my program is already outputting the correct categories but I don't know how to implement the grammar part (I already have a class with a arrayList for the grammar). As I said before, I think I will have to use Regex but didn't find that very helpful so far.

Thank you in advance!

Answer 1

Grammars are not usually implemented as regex since they are complex and self referencing, regex are more useful for parsing the incoming text. What you need is a better data structure than a list to represent the grammars and help you verify if a sentence is grammatically correct. Look into abstract syntax trees.

Answer 2

You can't generally implement a grammar with a regular expression, unless it is a grammar for a regular language, for example doesn't contain nested syntax such as

 expression: '(' expression ')'

You need a pushdown automaton of some kind, eg a recursive descent or LALR(1) parser.

See Chomsky hierarchy .