int
yylex(
void
)
{
yylast_g = yylex0();
return(yylast_g);
}
int yylex()
{
int token = yylex0();
int i;
#ifdef LEXER_SAVE_TOKENCODE
((struct Token *)yylval)->code = token;
#endif
yyCommentBuffer[yyCommentLength] = 0; /* ensure NUL termination */
#ifdef DEBUG
/* for debugging */
switch (token)
{
case IDENT:
fprintf(stderr, "[IDENT %s]", ((struct Token *)yylval)->ident);
return token;
case T_YIELDS:
fprintf(stderr, "[=>]");
return token;
case T_HBAR:
fprintf(stderr, "[---]");
return token;
default:
if (token > 0 && token < 256)
{
fprintf(stderr, "['%c']", token);
return token;
}
for(i = 0; keywords[i].name; ++i)
if (keywords[i].code == token)
{
fprintf(stderr, "[KW %s]", keywords[i].name);
return token;
}
fprintf(stderr, "[TOKEN %d]", token);
}
#endif
return token;
}
static int yylex0()
{
int i, c, d, e, token;
struct Token *val;
static char ident[LEXER_IDENT_MAXLENGTH];
static char string[LEXER_STRING_MAXLENGTH];
if(creg >= MAX_COMMENTINFO-1)
yyerror("!commentinfo-buffer overflow");
if (tmp != yylineno)
{
tmp = yylineno;
//printf("Curren row:%d \n",yylineno);
}
if(creg == 0)
{ //initial bound
struct CommentInfo cib;
cib.bound = 1;
cib.lastline = 1;
cib.lastcol = 1;
commentInfo[creg++] = cib;
}
else if(yylineno == specboundline + 2 && columnno == 0)
{ //bound after result
spreg = creg;
createBound(1);
}
val = (struct Token *) malloc(sizeof(struct Token));
if (!val)
yyerror("!out of memory");
memset(val, 0, sizeof(struct Token));
yylval = (rml_t) val;
c = get();
#ifdef LEXER_TOKEN_POSITION
val->file = yyThisFileName;
val->firstline = yylineno;
val->firstcol = columnno;
val->lastline = yylineno;
val->lastcol = columnno;
#endif
switch (c)
{
case '\n': case ' ': case '\t':
return yylex0();
case EOF: case 0:
return EOF;
case '=': /* '=' or '=>' or == or ==. */
d = get();
if (d == '>')
{ token = YIELDS; break; }
if (d == '=')
{
e = get();
if (e == '.') { token = EQEQ_REAL; break; }
unget();
token = EQEQ_INT; break;
}
unget();
token = EQUALS;
break;
case '<': /* < or <= or <=. or <. or <> or <>. */
d = get();
if (d == '.') { token = LT_REAL; break; }
if (d == '=')
{
e = get();
if (e = '.') { token = LE_REAL; break; }
unget();
token = LE_INT; break;
}
if (d == '>')
{
e = get();
if (e = '.') { token = NE_REAL; break; }
unget();
token = NE_INT; break;
}
unget();
token = LT_INT;
break;
case '>': /* > or >= or >=. or >. */
d = get();
if (d == '.') { token = GT_REAL; break; }
if (d == '=')
{
e = get();
if (e = '.') { token = GE_REAL; break; }
unget();
token = GE_INT; break;
}
unget();
token = GT_INT;
break;
case '!': /* != or !=. */
d = get();
if (d == '=')
{
e = get();
if (e = '.') { token = NE_REAL; break; }
unget();
token = NE_INT; break;
}
yyerror("invalid comparison operator!");
break;
case ':': /* ':' or '::' */
d = get();
if (d == ':')
{ token = COLONCOLON; break; }
unget();
token = COLON;
break;
case '(': /* '(' or '(*...comment...*)' */
d = get();
if (d == '*')
{
struct CommentInfo ci;//added
ci.firstline = yylineno;//added
ci.firstcol = columnno;
yyCommentLength = 0; //added
int innercomments = 1;
while (d != EOF)
{
d = get();
if (d != '*' && d != '(')
{
if (yyCommentLength < LEXER_COMMENT_MAXLENGTH)
ci.buffer[yyCommentLength++] = d;
continue;
}
if (d == '(')
{
d = get();
if (d == '*')
{
if (yyCommentLength < LEXER_COMMENT_MAXLENGTH)
ci.buffer[yyCommentLength++] = '(';
if (yyCommentLength < LEXER_COMMENT_MAXLENGTH)
ci.buffer[yyCommentLength++] = '*';
innercomments++;
}
else
{
if (yyCommentLength < LEXER_COMMENT_MAXLENGTH)
ci.buffer[yyCommentLength++] = '(';
}
unget();
continue;
}
if (d == '*')
{
d = get();
if (d == ')')
{
if (yyCommentLength < LEXER_COMMENT_MAXLENGTH)
ci.buffer[yyCommentLength++] = '*';
if (yyCommentLength < LEXER_COMMENT_MAXLENGTH)
ci.buffer[yyCommentLength++] = ')';
innercomments--;
if (innercomments == 0)
{
yyCommentLength--; /* don't put the last ')' */
break;
}
}
else
{
if (yyCommentLength < LEXER_COMMENT_MAXLENGTH)
ci.buffer[yyCommentLength++] = '*';
}
unget();
continue;
}
}
ci.buffer[yyCommentLength] = 0;
ci.lastline = yylineno; //added
ci.lastcol = columnno; //added
ci.bound = 0;
commentInfo[creg++] = ci;//added
return yylex0(); /* comment: skip and return next token */
}
unget();
iinner = 1;
token = LPAR;
break;
case '-': /* '-' or horizontal bar */
d = get();
if (d == '-')
{
while (d == '-')
d = get();
unget();
createBound(1);
specboundline = yylineno;
token = DASHES;
break;
}
unget();
token = MINUS;
break;
case '"': /* string constant */
i = 0;
while (1)
{
c = get();
if (c == '"')
break;
if (c == '\\')
c = esc_char(0);
if (i < LEXER_STRING_MAXLENGTH-1)
string[i++] = c;
else
yyerror("string buffer overflow");
}
string[i] = 0;
token = SCON;
val->u.string = strdup(string);
//printf("String: %s", val->u.string);
break;
case '#':
c = get();
if (c != '"')
yyerror("invalid character constant");
c = get();
if (c == '\\')
c = esc_char(1);
d = get();
if (d != '"')
yyerror("invalid character constant");
val->u.number = c;
token = CCON;
break;
case ')':
iinner = 0;
token = RPAR;
break;
case '.':
token = DOT;
break;
case ',':
token = COMMA;
break;
case '/':
token = DIV;
break;
case '+':
token = PLUS;
break;
case '*':
specboundline = yylineno; //can be problem if an expression...
if(!iinner)
createBound(3);
token = STAR;
break;
case '|':
token = BAR;
createBound(1);
specboundline = yylineno;
break;
case '&':
token = AMPERSAND;
createBound(1);
break;
case '[':
token = LBRACK;
break;
case ']':
token = RBRACK;
break;
case '%':
token = MOD_INT;
break;
default:
if (isalpha(c) || c == '\'' || c == '_')
{
i = 0;
do
{
if (i < 50)
{
ident[i++] = c;
#ifdef LEXER_TOKEN_POSITION
val->lastline = yylineno;
val->lastcol = columnno;
#endif
}
c = get();
}
while (isalnum(c) || c == '\'' || c == '_');
unget();
ident[i] = 0;
if (!keywords_sorted)
{
init_sort_keywords();
keywords_sorted = 1;
}
for(i = 0; keywords[i].name; ++i) {
if (!strcmp(ident, keywords[i].name)) {
if((keywords[i].code == KW_RELATION) || (keywords[i].code == KW_RULE) ||
(keywords[i].code == KW_END) || (keywords[i].code == KW_MODULE) || (keywords[i].code == KW_TYPE)
|| (keywords[i].code == KW_VAL) || (keywords[i].code == KW_WITH)) {
if(keywords[i].code == KW_END) {
if(spreg != 0) {
commentInfo[spreg].bound = 2;
spreg = 0;
}
}
createBound(1);
}
else if((keywords[i].code == KW_AXIOM) || (keywords[i].code == KW_DATATYPE)
|| (keywords[i].code == KW_AND)){
createBound(1);
specboundline = yylineno;
}
return keywords[i].code;
}
}
val->u.ident = strdup(ident);
//printf("the string: %s \n\n",ident);
return ident[0] == '\'' ? TYVARIDENT : IDENT;
}
else if (isdigit(c))
{
int is_float = 0;
i = 0;
for(val->u.number = 0; isdigit(c); c = get())
{
val->u.number = val->u.number*10 + c-'0';
if (i < LEXER_IDENT_MAXLENGTH)
ident[i++] = c;
};
if (c == '.')
{
do
{
if (i < LEXER_IDENT_MAXLENGTH)
ident[i++] = c;
c = get();
} while (isdigit(c));
is_float = 1;
};
if (c == 'e')
{
do
{
if (i < LEXER_IDENT_MAXLENGTH)
ident[i++] = c;
c = get();
} while (isdigit(c) || c == '+' || c == '-');
is_float = 1;
};
unget();
if (is_float)
{
ident[i] = 0;
if (sscanf(ident, "%lf", &(val->u.realnumber)) == 1)
{
//printf("Real value %f",val->u.realnumber);
token = RCON;
break;
}
yyerror("invalid real number");
}
token = ICON;
break;
}
else
yyerror("invalid input character");
}
#ifdef LEXER_TOKEN_POSITION
val->lastline = yylineno;
val->lastcol = columnno;
#endif
return token;
}
