Difference in LALR Grammar Parsing

[Crayon112]

Description

I am new to using LALR and I'm attempting to implement a grammar as follows:

A: "a"
B: "b"
a: b A
b: (A | B)+

However, when I try to parse the input string aaaba, I encounter the following error:

lark.exceptions.UnexpectedToken: Unexpected token Token('$END', '') at line 1, column 5.
Expected one of:
        * A
        * B

It seems that rule b is greedy in this case, despite my attempts to set a higher priority for the grammar rule a.

Proposed Solution

After making a simple modification to the grammar, I was able to get it working as follows:

A: "a"
B: "b"
a: b+ A
b: A | B

This adjustment resolved the error for the aforementioned input string.

Questions

What is the key difference between the two grammar definitions:

• a: b+ A and b: A | B
• a: b A and b: (A | B)+

How can I avoid the greediness in the first grammar configuration, specifically when defining rules for b? Any insights or guidance on this issue would be greatly appreciated. Thank you!