Swift Playground词法分析器DEMO

正在看极客时间宫文学老师的编译原理之美，用swift playground写了一个第二课“int age >= 45”的词法解析DEMO

为了保持原课程代码，DEMO用了顺序结构，看起来有点散乱，后面我再抽时间优化一下

//识别：“int age >= 45”

import Foundation

enum DfaState: String {
    case Initial = "Initial"
    case Id = "Id"
    case IntLiteral = "IntLiteral"
    case GT = "GT"
    case GE = "GE"
}

enum TokenType: String {
    case Identifier = "Identifier"
    case IntLiteral = "IntLiteral"
    case GT = "GT"
    case GE = "GE"
}

func isAlpha(_ ch:Character) -> Bool {
    return ch.isLetter
}

func isDigit(_ ch:Character) -> Bool {
    return ch.isNumber
}


class Token {
    var type: TokenType?
    var tokenText: String = ""
}

var newState = DfaState.Initial
var token = Token()
var tokenText = token.tokenText
//FIXME:这里赋值识别的语句
let str = "int age >= 45"


func initToken() -> DfaState {
    token = Token()
    tokenText = token.tokenText
    return .Initial
}

func read(_ string: String) {
    let strArr = string.components(separatedBy: CharacterSet(charactersIn: " "))
    for string in strArr {
        readToken(string)
        print("token is: \n" + (tokenText) + "\ntype is:\n" + ((token.type)?.rawValue ?? "undefined"))
        initToken()
    }
}

func readToken(_ string:String) {
    checkType(string.first)
    let str = String(string.dropFirst())
    if (str.count == 0) {
        return
    }
    for ch in str {
        readLeft(ch)
    }
}

func checkType(_ ch:Character?) {
    if (ch == nil) {
        return
    }
    let ch = ch!
    if (isAlpha(ch)) {              // 第一个字符是字母
        newState = DfaState.Id; // 进入 Id 状态
        token.type = TokenType.Identifier;
        tokenText.append(ch);
    } else if (isDigit(ch)) {       // 第一个字符是数字
        newState = DfaState.IntLiteral;
        token.type = TokenType.IntLiteral;
        tokenText.append(ch);
    } else if (ch == ">") {         // 第一个字符是 >
        newState = DfaState.GT;
        token.type = TokenType.GT;
        tokenText.append(ch);
    }
}

func readLeft(_ ch: Character) {
    var state = newState
    switch state {
    case .Initial:
        state = initToken();          // 重新确定后续状态
        break;
    case .Id:
        if (isAlpha(ch) || isDigit(ch)) {
            tokenText.append(ch);       // 保持标识符状态
        } else {
            state = initToken(); // 退出标识符状态，并保存 Token
        }
        break;
    case .GT:
        if (ch == "=") {
            token.type = TokenType.GE;  // 转换成 GE
            state = DfaState.GE;
            tokenText.append(ch);
        } else {
            state = initToken();      // 退出 GT 状态，并保存 Token
        }
        break;
    case .GE:
        state = initToken();        // 退出当前状态，并保存 Token
        break;
    case .IntLiteral:
        if (isDigit(ch)) {
            tokenText.append(ch);    // 继续保持在数字字面量状态
        } else {
            state = initToken();    // 退出当前状态，并保存 Token
        }
        break;
    }
}


read(str)

输出结果

token is: 
int
type is:
Identifier
token is: 
age
type is:
Identifier
token is: 
>=
type is:
GE
token is: 
45
type is:
IntLiteral

---

END