Many researcher have tried to move run-time computation to compile-time computation. In the progress of LR parsing, the sequences of states in the parsing stack determines the expected vocabulary strings. We have investigated the methods of computing ...
Many researcher have tried to move run-time computation to compile-time computation. In the progress of LR parsing, the sequences of states in the parsing stack determines the expected vocabulary strings. We have investigated the methods of computing expected vocabulary strings. Recently, we find that most of expected vocabulary string are not dependent on the history of parsing. In other words, expected vocabulary strings are pre-computable at compile-time not at run-time, and it enable us to get more run-time efficiency. We will show that most of expected vocabulary strings are not dependent on left contexts through the experimentation on several programming languages.