Lex

Lex is a Regex Engine for lexical analyzer

build

It's used cmake to build the project, and doxygen to generate the html doc.

1. cd to the project dir.
2. mkdir build
3. cd build
4. cmake ..
5. make or make install
6. make doc to build the doc

test

use ./lextest to test the project.

use Lex to Scan the source file

you may use the LexInterface to get the token

first include the .h files in the include dir,

then you should link the liblex.a library you built.

//it's a token for return
struct Token
{
	CHAR* pToken; // the string will return
	int type;     // type id for the token 
    int col_num;      // which column the token is
	int row_num;  // which rownum the token is
};

class LexInterface 
{
public:
    virtual Token* Read() = 0;  // get the next token

    virtual const wchar_t* getRule(int id) = 0; 

    virtual int getRuleSize() = 0;
};

Regex rules

if you want to scan the source file, you need to make some rule

like this:

int=[0-9]+
float=[0-9]*\.[0-9]*
id=[a-zA-Z_]\w*
opta= =|\+=|\-=
optc= [*/]
optd= [+\-]
opte= >|<|<=|>=|==|!=
other=[(){},;:]
ignore=\s+

It means matching the regex will return the type id,

from the first id is 1,

0 is the ignore rule.

so you can get the token id and the token data from the Token structure.