Consider in contradiction, that C is not a deterministic CFG. The valid ahb is considered, that has unforced handle h . It has a possibility that some string has another handle. Thus, ahb' can be rewritten as where, b' is a terminal.
While, ah= , the reduce rule will be changed, because h and are different handles. Thus, it does not satisfy the DK-test since it has two completed rules.
While ah , assume that the proper prefix of is ah . Assume that s is an accepting state of the input in DK. The s accepts the states because h is the handle.
Hence, the transition must label b' because, doesnot has null rule. It does not satisfy the DK-test.
Therefore, the DK-test fails, In contradiction to the assumption, CFG is a deterministic context-free grammar.