Sample code for finite state machine 1

[sedawk]

Hi,

Where can I get a neat sample code for finite state machine that will be used in a lexicon scanner? Did a goole but too much text, no immediate codes found. So seek help from you guys. Maybe you can put it on FAQ.

Thanks.

 

[xwb]

This is how we used to do it in the early days before lex existed

#include <stdio.h>
#include <string.h>
#include <stdlib.h>

/* First the events */
enum LexType
{
   eIll, /* Illegal char */
   eNum, /* Number */
   eAlf, /* Alphabetic */
   eSpc, /* Whitespace */
   eEof, /* End of file */
   eMax
};

/* State/Event lookup - assumes ASCII */
enum LexType lookup[] =
{
   eEof, eIll, eIll, eIll, eIll, eIll, eIll, eIll, /* 00-07 */
   eIll, eIll, eSpc, eIll, eIll, eSpc, eIll, eIll, /* 08-0F */
   eIll, eIll, eIll, eIll, eIll, eIll, eIll, eIll, /* 10-17 */
   eIll, eIll, eEof, eIll, eIll, eIll, eIll, eIll, /* 18-1F */
   eSpc, eIll, eIll, eIll, eIll, eIll, eIll, eIll, /* 20-27 */
   eIll, eIll, eIll, eIll, eIll, eIll, eIll, eIll, /* 28-2F */
   eNum, eNum, eNum, eNum, eNum, eNum, eNum, eNum, /* 30-37 */
   eNum, eNum, eIll, eIll, eIll, eIll, eIll, eIll, /* 38-3F */
   eIll, eAlf, eAlf, eAlf, eAlf, eAlf, eAlf, eAlf, /* 40-47 */
   eAlf, eAlf, eAlf, eAlf, eAlf, eAlf, eAlf, eAlf, /* 48-4F */
   eAlf, eAlf, eAlf, eAlf, eAlf, eAlf, eAlf, eAlf, /* 50-57 */
   eAlf, eAlf, eAlf, eIll, eIll, eIll, eIll, eIll, /* 58-5F */
   eIll, eAlf, eAlf, eAlf, eAlf, eAlf, eAlf, eAlf, /* 60-67 */
   eAlf, eAlf, eAlf, eAlf, eAlf, eAlf, eAlf, eAlf, /* 68-6F */
   eAlf, eAlf, eAlf, eAlf, eAlf, eAlf, eAlf, eAlf, /* 70-77 */
   eAlf, eAlf, eAlf, eIll, eIll, eIll, eIll, eIll  /* 78-7F */
};

FILE* srcFile;
char srcBuff[128];/* source buffer */
char* src;
int  srcIx;       /* position in buffer */
int  srcMax;      /* size of buffer */

/* Initialize */
void SrcInit (int argc, const char* argv[])
{
   if (argc < 2)
   {
      printf ("Input File required\n");
      exit (0);
   }
   srcFile = fopen (argv[1], "r");
   if (srcFile == NULL)
   {
      printf ("Unable to open %s\n", argv[1]);
      exit (0);
   }
   /* Set conditions for kicking something into the buffer */
   srcIx = 1;
   srcMax = 0;
}
/* Get the current character */
char SrcGet ()
{
   return src[srcIx];
}
/* Get the next character */
char SrcNext ()
{
   if (++srcIx >= srcMax)
   {
      src = fgets (srcBuff, 128, srcFile);
      if (!src)
      {
         /* Mark EOF */
         src = srcBuff;
         src[0] = 0;
      }
      srcIx = 0;
      srcMax = strlen (src);
   }

   /* (char) 0 used to denote EOF */
   return src[srcIx];
}

void SkipToNextWord ()
{
   char ch;

   ch = SrcGet ();
   while (lookup[ch] == eIll || lookup[ch] == eSpc)
      ch = SrcNext ();
}
/* Get the next word */
void SrcNextWord (char* word, enum LexType lextok)
{
   int wordLen;
   char ch;
   /* Assume the buffer is big enough */

   /* Get the word */
   wordLen = 0;
   ch = SrcGet ();
   while (lookup[ch] == lextok)
   {
      word[wordLen++] = ch;
      ch = SrcNext ();
   }
   word[wordLen] = '\0';
}
int main(int argc, const char* argv[])
{
   char theWord[128];
   int  resume;

   SrcInit (argc, argv);
   SrcNext ();
   resume = 1; /* true */
   do {
      SkipToNextWord ();
      switch (lookup[SrcGet()])
      {
      case eAlf:
         SrcNextWord (theWord, eAlf);
         printf ("Word %s\n", theWord);
         break;

      case eNum:
         SrcNextWord (theWord, eNum);
         printf ("Number %s\n", theWord);
         break;

      case eEof:
         resume = 0;
         break;
      }
   } while (resume);
	return 0;
}

 

[sedawk]

so this approach is not a strictly state machine. anyway, thanks a lot.