static
int scanner_symbol(mvc * c, int cur)
{
struct scanner *lc = &c->scanner;
int next = 0;
int started = lc->started;
switch (cur) {
case '/':
lc->started = 1;
next = scanner_getc(lc);
if (next == '*') {
lc->started = started;
cur = skip_c_comment(lc);
if (cur < 0)
return EOF;
return tokenize(c, cur);
} else {
utf8_putchar(lc, next);
return scanner_token(lc, cur);
}
case '0':
case '1':
case '2':
case '3':
case '4':
case '5':
case '6':
case '7':
case '8':
case '9':
return number(c, cur);
case '#':
if ((cur = skip_sql_comment(lc)) == EOF)
return cur;
return tokenize(c, cur);
case '\'':
case '"':
return scanner_string(c, cur);
case '{':
return scanner_body(c);
case '-':
lc->started = 1;
next = scanner_getc(lc);
if (next == '-') {
lc->started = started;
if ((cur = skip_sql_comment(lc)) == EOF)
return cur;
return tokenize(c, cur);
}
lc->started = 1;
utf8_putchar(lc, next);
return scanner_token(lc, cur);
case '~': /* binary not */
case '^': /* binary xor */
case '*':
case '?':
case '%':
case '+':
case '(':
case ')':
case ',':
case '=':
case '[':
case ']':
lc->started = 1;
return scanner_token(lc, cur);
case '&':
lc->started = 1;
cur = scanner_getc(lc);
if(cur == '<') {
next = scanner_getc(lc);
if(next == '|') {
return scanner_token(lc, GEOM_OVERLAP_OR_BELOW);
} else {
utf8_putchar(lc, next); //put the char back
return scanner_token(lc, GEOM_OVERLAP_OR_LEFT);
}
} else if(cur == '>')
return scanner_token(lc, GEOM_OVERLAP_OR_RIGHT);
else if(cur == '&')
return scanner_token(lc, GEOM_OVERLAP);
else {/* binary and */
utf8_putchar(lc, cur); //put the char back
return scanner_token(lc, '&');
}
case '@':
lc->started = 1;
return scanner_token(lc, AT);
case ';':
lc->started = 0;
return scanner_token(lc, SCOLON);
case '<':
lc->started = 1;
cur = scanner_getc(lc);
if (cur == '=') {
return scanner_token( lc, COMPARISON);
} else if (cur == '>') {
return scanner_token( lc, COMPARISON);
} else if (cur == '<') {
next = scanner_getc(lc);
if (next == '=') {
return scanner_token( lc, LEFT_SHIFT_ASSIGN);
} else if (next == '|') {
return scanner_token(lc, GEOM_BELOW);
} else {
utf8_putchar(lc, next); //put the char back
return scanner_token( lc, LEFT_SHIFT);
}
} else if(cur == '-') {
next = scanner_getc(lc);
if(next == '>') {
return scanner_token(lc, GEOM_DIST);
} else {
//put the characters back and fall in the next possible case
utf8_putchar(lc, next);
utf8_putchar(lc, cur);
return scanner_token( lc, COMPARISON);
}
} else {
utf8_putchar(lc, cur);
return scanner_token( lc, COMPARISON);
}
case '>':
lc->started = 1;
cur = scanner_getc(lc);
if (cur == '>') {
cur = scanner_getc(lc);
if (cur == '=')
return scanner_token( lc, RIGHT_SHIFT_ASSIGN);
utf8_putchar(lc, cur);
return scanner_token( lc, RIGHT_SHIFT);
} else if (cur != '=') {
utf8_putchar(lc, cur);
return scanner_token( lc, COMPARISON);
} else {
return scanner_token( lc, COMPARISON);
}
case '.':
lc->started = 1;
cur = scanner_getc(lc);
if (!iswdigit(cur)) {
utf8_putchar(lc, cur);
return scanner_token( lc, '.');
} else {
utf8_putchar(lc, cur);
cur = '.';
return number(c, cur);
}
case '|': /* binary or or string concat */
lc->started = 1;
cur = scanner_getc(lc);
if (cur == '|') {
return scanner_token(lc, CONCATSTRING);
} else if (cur == '&') {
next = scanner_getc(lc);
if(next == '>') {
return scanner_token(lc, GEOM_OVERLAP_OR_ABOVE);
} else {
utf8_putchar(lc, next); //put the char back
utf8_putchar(lc, cur); //put the char back
return scanner_token(lc, '|');
}
} else if (cur == '>') {
next = scanner_getc(lc);
if(next == '>') {
return scanner_token(lc, GEOM_ABOVE);
} else {
utf8_putchar(lc, next); //put the char back
utf8_putchar(lc, cur); //put the char back
return scanner_token(lc, '|');
}
} else {
utf8_putchar(lc, cur);
return scanner_token(lc, '|');
}
}
(void)sql_error( c, 3, "unexpected symbol (%lc)", (wint_t) cur);
return LEX_ERROR;
}
bool tokz_get_token(Tokenizer *tokz, Token *tok)
{
int c, c2, e;
if (!(tokz->flags&TOKZ_READ_FROM_BUFFER))
assert(tokz->file!=NULL);
tok_free(tok);
if(!TOK_IS_INVALID(&(tokz->ungettok))){
*tok=tokz->ungettok;
tokz->ungettok.type=TOK_INVALID;
return TRUE;
}
while(1){
e=0;
do{
c=GETCH();
}while(c!='\n' && c!=EOF && isspace(c));
tok->line=tokz->line;
switch(c){
case EOF:
TOK_SET_OP(tok, OP_EOF);
return TRUE;
case '\n':
INC_LINE();
if(tokz->flags&TOKZ_IGNORE_NEXTLINE)
continue;
TOK_SET_OP(tok, OP_NEXTLINE);
return TRUE;
case '\\':
do{
c=GETCH();
if(c==EOF){
TOK_SET_OP(tok, OP_EOF);
return FALSE;
}
if(!isspace(c) && e==0){
e=E_TOKZ_EOL_EXPECTED;
tokz_warn_error(tokz, tokz->line, e);
if(!(tokz->flags&TOKZ_ERROR_TOLERANT))
return FALSE;
}
}while(c!='\n');
INC_LINE();
continue;
case '#':
if(tokz->flags&TOKZ_READ_COMMENTS){
e=scan_line_comment(tok, tokz);
break;
}else if((e=skip_line_comment(tokz))){
break;
}
continue;
case '/':
c2=GETCH();
if(c2=='='){
TOK_SET_OP(tok, OP_AS_DIV);
return TRUE;
}
if(c2!='*'){
UNGETCH(c2);
TOK_SET_OP(tok, OP_DIV);
return TRUE;
}
if(tokz->flags&TOKZ_READ_COMMENTS){
e=scan_c_comment(tok, tokz);
break;
}else if((e=skip_c_comment(tokz))){
break;
}
continue;
case '\"':
e=scan_string(tok, tokz, TRUE);
break;
case '\'':
e=scan_char(tok, tokz);
break;
default:
if(('0'<=c && c<='9') || c=='-' || c=='+'){
e=scan_number(tok, tokz, c);
break;
}
if(START_IDENT(c))
e=scan_identifier(tok, tokz, c);
else
e=scan_op(tok, tokz, c);
}
if(!e)
return TRUE;
tokz_warn_error(tokz, tokz->line, e);
return FALSE;
}
}
