package com.csray;import java.io.BufferedReader;
import java.io.File;
import java.io.FileNotFoundException;
import java.io.FileReader;
import java.io.IOException;public class LexicalAnalysis {
static String id;
public static void main(String[] args) throws IOException{
LexicalAnalysis la = new LexicalAnalysis();
String context = //la.BufferedReaderDemo
//("C:"+System.getProperty("file.separator")+"Users"+System.getProperty("file.separator")+"Administrator"+System.getProperty("file.separator")+"Desktop"+System.getProperty("file.separator")+"Lexical");
la.BufferedReaderDemo("C:\\Users\\Administrator\\Desktop\\Lexical\\lexical.c");
System.out.println(context);for(int i = 0; i < context.length();){
char nowc = context.charAt(i);
//System.out.println(nowc);
if(nowc == ' ' || nowc == '\n' || nowc == '\t'){
id = "illegalCharacter";
++i;
}else if(isAlpha(nowc)){
i = alphaProcess(context, nowc, i);
} else if(isDigit(nowc)){
i = digitProcess(context, nowc, i);
//i++;
} else {
i = otherProcess(context, nowc, i);
//i++;
}
}
}//read the lexical.c file to String
public String BufferedReaderDemo(String path) throws IOException{
File file = new File(path);
if(!file.exists() || file.isDirectory())
throw new FileNotFoundException();BufferedReader br = new BufferedReader(new FileReader(file));
String tmp = null;
StringBuffer sbuff = new StringBuffer();
tmp = br.readLine();
while(tmp != null){
sbuff.append(tmp+" ");
tmp = br.readLine();
}
return sbuff.toString();
}//alphaProcess
public static int alphaProcess(String context, char c, int i){
StringBuffer word = new StringBuffer();
while((isAlpha(c)) || (isDigit(c)) || c == '_'){
word.append(c);
c = context.charAt(++i);
}
//System.out.println(word.toString());//check this word if or not a keyword
if(checkKeyword(word.toString())){
id = "isKeyword";
System.out.println("( "+ word.toString() + ", " + id +" )");
} else {
id = "isCommonWord";
System.out.println("( "+ word.toString() + ", " + id +" )");
} return i;
}// checkKeyword
public static boolean checkKeyword(String word){
String keyword = "auto double int struct break else long switch case enum register typedef char "
+ "extern return union const float short unsigned continue for signed void default goto "
+ "sizeof volatile do if while static scnaf printf";
String[] keyWords = keyword.split(" ");
for(int i = 0; i < keyWords.length; ++i){
if(word.equals(keyWords[i])){
return true;
}
}
return false;
}// digitProcess
public static int digitProcess(String context, char c, int i){
StringBuffer digit = new StringBuffer();
while(isDigit(c)){
digit.append(c);
c = context.charAt(++i);
}
id = "isDigit";
System.out.println("( "+ digit.toString() + ", " + id + " )");
return i;
}// otherProcess
public static int otherProcess(String context, char c, int i){
//+ / ( ) { } += /= & " "
StringBuffer operator = new StringBuffer();
StringBuffer delimiter = new StringBuffer();
if(c == '+'){
operator.append(c);
id = "isOperator";
c = context.charAt(++i);
if(c == '='){
operator.append(c);
System.out.println("( "+ operator + ", " + id + " )");
return ++i;
}else if(c == '+'){
operator.append(c);
System.out.println("( "+ operator + ", " + id + " )");
return ++i;
}else {
System.out.println("( "+ operator + ", " + id + " )");
return --i;
}
} else if(c == '('){
delimiter.append(c);
id = "isDelimiter";
System.out.println("(" + delimiter + ", " + id + " )");
return ++i;
} else if(c == ')'){
delimiter.append(c);
id = "isDelimiter";
System.out.println("(" + delimiter + ", " + id + " )");
return ++i;
}else if(c == '{'){
delimiter.append(c);
id = "isDelimiter";
System.out.println("(" + delimiter + ", " + id + " )");
return ++i;
} else if(c == '}'){
delimiter.append(c);
id = "isDelimiter";
System.out.println("(" + delimiter + ", " + id + " )");
return ++i;
}else if(c == ';'){
delimiter.append(c);
id = "isDelimiter";
System.out.println("(" + delimiter + ", " + id + " )");
return ++i;
}else if(c == '='){
operator.append(c);
id = "isOperator";
c = context.charAt(++i);
if(c == '='){
operator.append(c);
System.out.println("( "+ operator + ", " + id + " )");
return ++i;
}else {
System.out.println("( "+ operator + ", " + id + " )");
return i;
}
} else if(c == '/'){
operator.append(c);
c = context.charAt(++i);
if(c == '='){
operator.append(c);
id = "isOperator";
System.out.println("( "+ operator + ", " + id + " )");
return ++i;
}else if(c == '/'){
operator.append(c);
id = "isAnnotation";
c = context.charAt(++i);
while(c != ' '){
operator.append(c);
c = context.charAt(++i);
}
System.out.println("( "+ operator + ", " + id + " )");
return i;
}
}else if(c == '&'){
operator.append(c);
id = "isOperator";
System.out.println("(" + operator + ", " + id + " )");
return ++i;
}else if(c == '"'){
delimiter.append(c);
id = "isDelimiter";
System.out.println("(" + delimiter + ", " + id + " )");
return ++i;
}else if(c == ','){
delimiter.append(c);
id = "isDelimiter";
System.out.println("(" + delimiter + ", " + id + " )");
return ++i;
}else if(c == '%'){
operator.append(c);
id = "isOperator";
c = context.charAt(++i);
if(c == 'd'){
operator.append(c);
System.out.println("(" + operator + ", " + id + " )");
return ++i;
}return --i;
}
return i;
}//isDigit
public static boolean isDigit(char c){
if(c >= '0' && c <= '9')
return true;
return false;
}//isAlpha
public static boolean isAlpha(char c){
if(c >= 'A' && c <= 'Z')
return true;
else if(c >= 'a' && c <= 'z')
return true;
return false;
}
}
实例：
//test
int main()
{float nu0_m;
int num = 100;
num++;
scanf("%d", &num);
printf("%d", num);
return 0;
}
结果：
//test int main() { float nu0_m; int num = 100; num++; scanf("%d", &num); printf("%d", num); return 0; }
( //test, isAnnotation )
( int, isKeyword )
( main, isCommonWord )
((, isDelimiter )
(), isDelimiter )
({, isDelimiter )
( float, isKeyword )
( nu0_m, isCommonWord )
(;, isDelimiter )
( int, isKeyword )
( num, isCommonWord )
( =, isOperator )
( 100, isDigit )
(;, isDelimiter )
( num, isCommonWord )
( ++, isOperator )
(;, isDelimiter )
( scanf, isCommonWord )
((, isDelimiter )
(", isDelimiter )
(%d, isOperator )
(", isDelimiter )
(,, isDelimiter )
(&, isOperator )
( num, isCommonWord )
(), isDelimiter )
(;, isDelimiter )
( printf, isKeyword )
((, isDelimiter )
(", isDelimiter )
(%d, isOperator )
(", isDelimiter )
(,, isDelimiter )
( num, isCommonWord )
(), isDelimiter )
(;, isDelimiter )
( return, isKeyword )
( 0, isDigit )
(;, isDelimiter )
(}, isDelimiter )
