flex-help

[Mahendra L. <lm...@ya...>]

Hello,

       Here is my simplified flex file

%%
"keyword1"                                         {  return KEYWORD1;   };
"keyword2"                                         {  return KEYWORD2;   };
"unit1"                                                {  return UNIT1;      };
"unit2"                                                {  return UNIT2;      };
"activate"                                            {  return ACTIVATE;   };
"deactivate"                                         {  return DEACTIVATE; };
"."                                                      {  return DOT;        };
[A-Za-z][A-Za-z0-9_-]*([.][A-Za-z][A-Za-z0-9_-]*)+     {  return OBJECT_ID;  };
%%

I'm having a bison file also to generate the parser.

If the input to the tokenizer is
xyz.pqr                : I expect it to return OBJECT_ID
unit1xyz.activatepqr   : should return OBJECT_ID
unit1.activate         : should return 3 tokens, namely, UNIT1, DOT, ACTIVATE
keyword1.keyword2      : should return 3 tokens, namely, KEYWORD1, DOT, KEYWORD2

But I'm seeing that the last 2 inputs result in OBJECT_ID token.
This happens since Flex generated tokenizers always return the longest token.
How can I get the expected behaviour as described above ?

Thank you,
Mahendra