原本是将常量这划分成了所有的类型，还特别标注了byte、boolean与String是要自己实现的
还想好了，用结构与联合实现存储不定长度内容的东西
突然想起来，这是词法分析
简单展示分类
/* 关键字定义 */
ABSTRACT abstract
DEFAULT default
GOTO goto
NULL null
SWITCH switch
BOOLEAN boolean
DO do
IF if
PACKAGE package
NCHRONZED nchronzed
BREAK break
DOUBLE double
IMPLEMENTS implements
PRIVATE private
THIS this
BYTE byte
ELSE else
IMPORT import
PROTECTED protected
THROW throw
THROWS throws
CASE case
EXTENDS extends
INSTANCEOF instanceof
PUBLIC public
TRANSIENT transient
CATCH catch
INT int
RETURN return
CHAR char
FINAL final
INTERFACE interface
SHORT short
TRY try
CLASS class
FINALLY finally
LONG long
STATIC static
VOID void
CONST const
FLOAT float
NATIVE native
STRICTFP strictfp
VOLATILE volatile
CONTINUE continue
FOR for
NEW new
SUPER super
WHILE while
ASSERT assert
ENUM enum
/* 关键字定义结束 */
/* 运算符定义 */
QUES \?
COLON [:]
ASSIGNOP [=]
RELOP [>]|[<]|[>][=]|[<][=]|[=][=]|[!][=]
PLUS [+]
SUB [-]
MUL [*]
DIV [/]
AND [&][&]
OR [|][|]
/* 运算符定义结束 */
/* 分界符定义 */
ATT \@
DOT [.]
COMMA [,]
NOT [!]
SEM [;]
LP \(
RP \)
LB \[
RB \]
LC \{
RC \}
/* 分界符定义结束 */
/* 常数定义 */
STA_INT (-?[1-9][0-9]*|0)
STA_FLOAT -?(0|[1-9]+)\.[0-9]*
STA_CHAR \'[a-zA-Z]\'
STA_STRING \".*\"
STA_BOOLEAN (true|false)
/* 常数定义结束 */
/* 标识符定义 */
ID [a-zA-Z_$][a-zA-Z_$]*
/* 标识符定义结束 */
LINECOMMENT \/\/.*
COMMENT ((\/\*((\*[^\/]+)|[^\*])+\*\/)|(\/\/))
%option noyywrap
%%
{ABSTRACT} |
{DEFAULT} |
{GOTO} |
{NULL} |
{SWITCH} |
{BOOLEAN} |
{DO} |
{IF} |
{PACKAGE} |
{NCHRONZED} |
{BREAK} |
{DOUBLE} |
{IMPLEMENTS} |
{PRIVATE} |
{THIS} |
{BYTE} |
{ELSE} |
{IMPORT} |
{PROTECTED} |
{THROW} |
{THROWS} |
{CASE} |
{EXTENDS} |
{INSTANCEOF} |
{PUBLIC} |
{TRANSIENT} |
{CATCH} |
{INT} |
{RETURN} |
{CHAR} |
{FINAL} |
{INTERFACE} |
{SHORT} |
{TRY} |
{CLASS} |
{FINALLY} |
{LONG} |
{STATIC} |
{VOID} |
{CONST} |
{FLOAT} |
{NATIVE} |
{STRICTFP} |
{VOLATILE} |
{CONTINUE} |
{FOR} |
{NEW} |
{SUPER} |
{WHILE} |
{ASSERT} |
{ENUM} {
printf("< 关键字 , %10s >\n", yytext);
}
{QUES} |
{COLON} |
{ASSIGNOP} |
{RELOP} |
{PLUS} |
{SUB} |
{MUL} |
{DIV} |
{AND} |
{OR} {
printf("< 运算符 , %10s >\n", yytext);
}
{ATT} |
{COMMA} |
{DOT} |
{NOT} |
{SEM} |
{LP} |
{RP} |
{LB} |
{RB} |
{LC} |
{RC} {
printf("< 分界符 , %10s >\n", yytext);
}
{STA_INT} |
{STA_FLOAT} |
{STA_CHAR} |
{STA_STRING} |
{STA_BOOLEAN} {
printf("< 常数 , %10s >\n", yytext);
}
{ID} {
printf("< 标识符 , %10s >\n", yytext);
}
{LINECOMMENT} |
{COMMENT} |
[ \n\t] //eat the white space注意yywrap()这个函数
具体可以参看flex manual手册
有三种方法：
- 在
definition部分，添加如下内容
%{
#define yywrap() 1
%}```
2. 在`user code`部分，添加函数
int yywrap(){
return 1;
}```
- 在
definition部分添加%option noyywrap