//Returns the final state in with maximum prefix match
//i.e which has maximum forward pointer.
STATE_AND_FORWARD simulateNFA(char input[], STATE STARTSTATE, int forward)
{
char currentChar = input[forward];
if(isFinalState(STARTSTATE))//Final state has no output transition
{
STATE_AND_FORWARD t = {STARTSTATE, forward-1};
return t;
}
/*flag checks if there is any transition available for this state,
otherwise returns -1*/
int i, flag = 0;
STATE_AND_FORWARD max = {-1, -1}, temp = {-1, -1};
for(i=1;i<nextAvailableState;++i)
{
if(STT[STARTSTATE][i][currentChar])//Move forward ahead
{
flag = 1;
temp = simulateNFA(input, i, forward+1);
}
else if(STT[STARTSTATE][i]['$'])//EPSILON transition.Dont move forward ahead
{
flag = 1;
temp = simulateNFA(input, i, forward);
}
if(temp.forward>max.forward)
max = temp;
}
if(flag == 0)//No transition found
{
STATE_AND_FORWARD t = {-1, -1};
return t;
}
return max;
}
int main()
{
int i, LB;
FILE * fp1,* fp2;
char PATTERN[100];
char REGEX[100];
char * POSTFIX;
fp1 = fopen("regexp.in", "r");
if(fp1==NULL)
{
printf("Could not open regexp.in!!");
exit(0);
}
while(!feof(fp1))
{
fscanf(fp1, "%s %s",PATTERN, REGEX);
if(strlen(REGEX)==1 || (strlen(REGEX)==2 && REGEX[1]=='*'))
{
buildIndividualNFA(PATTERN, REGEX);
continue;
}
else if(strlen(REGEX)==2 && REGEX[1]!='*')
{
REGEX[2]='&';
REGEX[3]='\0';
buildIndividualNFA(PATTERN, REGEX);
continue;
}
POSTFIX = conversion(REGEX);
buildIndividualNFA(PATTERN, POSTFIX);
}
fclose(fp1);
buildCombinedNFA();
char input[1000];
char tokenread[100];
initialize_Symbol_Table();
remove("a4_2.out");
while(gets(input))
{
int lexemeBegin = 0;
while(input[lexemeBegin]!='\0'&&input[lexemeBegin]!='\n' &&input[lexemeBegin]!='\r') //while end of line is not found
{
while(input[lexemeBegin]==' ' || input[lexemeBegin]=='\t')
lexemeBegin++;
lexemeBegin = comment(input, lexemeBegin);
if(input[lexemeBegin] == '\0' || input[lexemeBegin]=='\n'||input[lexemeBegin]=='\r')
break;
STATE_AND_FORWARD final = simulateNFA(input, HEAD, lexemeBegin);
i = final.forward - lexemeBegin+1;
memcpy(tokenread, input+lexemeBegin, i);
tokenread[i] = '\0';
token t;//Token
attrType a;//attribute field
a.string = (char *)malloc((i+1)*sizeof(char));
int tokenID = getTokenID(final.state);
if(tokenID == IDNTIFIER)
{
strcpy(a.string, tokenread);
int flag = 0;
int j;
for(j=0;j<24;j++)
{
if(strcmp(a.string,keywordArr[j].keyword)==0) //checking if string detected is a keyword
{
t.tokenID = keywordArr[j].tokenID;
t.attribute.string = '\0';
flag = 1;
break;
}
}
if(flag==0) //not a keyword; is an identifier
{
t.tokenID = IDNTIFIER;
fp1 = fopen("symbol_table_2.out","r");
int nflag = 0;
char identifier[100];
int marker;
while(!feof(fp1))
{
fscanf(fp1,"%s %d",identifier,&marker);
if(strcmp(identifier,a.string)==0) //if identifer is already present in symbol table
{
nflag = 1;
break;
}
}
fclose(fp1);
if(nflag==0) //identifier not present in symbol table
{
fp1 = fopen("symbol_table_2.out","a");
fprintf(fp1,"%s 1\n",a.string);
fclose(fp1);
}
t.attribute = a;
}
}
else if(tokenID == INT_CONST)
{
a.integer = atoi(tokenread);
t.tokenID = INT_CONST;
t.attribute = a;
}
else if(tokenID == FLO_CONST)
{
a.integer = atof(tokenread);
t.tokenID = FLO_CONST;
t.attribute = a;
}
else if(tokenID!=NOTOK)
{
t.tokenID = tokenID;
t.attribute.string = '\0';
}
if(tokenID != NOTOK)
write_token(t);
else
{
lexemeBegin++;
continue;
}
lexemeBegin = final.forward+1;
}
}
}
int NFA::FindFirstMatchedString(string strToExtract)
{
return simulateNFA(strToExtract);
}
