Alternatives to de Marcken:

 

(…various compression schemes, related to Lempel-Ziv…and)

 

1. Nevill-Manning (1996) Inferring Sequential Structure (student of Ian Witten)

 

Basic idea: scan the string, never permitting a repeat of a string of symbols. Whenever a repeat is found, replace it.

 

Input

Grammar

 

a

S -> a

 

ab

S -> ab

 

abc

S -> abc

 

abcd

S -> abcd

 

abcdb

S -> abcdb

 

abcdbc

S -> abcdbc

2 bc’s

 

S-> aAdA;  A->bc

 

abcdbca

S-> aAaAa; A->bc

 

abcdbcab

S-> aAdAab; A->bc

 

abcdbcabc

S-> aAdAabc; A->bc

2 bc’s

 

S-> aAdAaA; A->bc

2 aA’s

 

S-> BdAB; A->bc; B->aA

 

abcdbcabcd

S-> BdABd; A->bc; B->aA

2 Bd’s

 

S-> CAC; A->bc; B->aA;

C->Bd

B occurs only once on the rhs

 

S-> CAC; A->bc; C->aAd