int arg_count_of_string( char* s, int len ) {
char c = *s;
if ( !is_id_alpha(c) ) return 1;
if ( s[ len - 1 ] != ':' ) return 0; // an optimization
fint argc = 1;
for ( const char* ss = s + len - 3; // last is :, next-to-last is alpha
ss > s; // do not need to look at first one
--ss )
if (*ss == ':') ++argc;
return argc;
}
fint str_arg_count(const char* s, fint len) {
assert(len > 0, "should have a positive length");
char c = *s;
if (!is_id_alpha(c)) {
return 1;
}
if ( s[ len-1 ] != ':' ) return 0; // an optimization
fint argc = 1;
for (const char* ss = s + len - 3; // last is :, next-to-last is alpha
ss > s; // do not need to look at first one
--ss ) {
if (*ss == ':') {
argc ++;
}
}
return argc;
}
fint str_arg_count(const char* s) {
char c = *s++;
assert(c != '\0', "shouldn't be testing the empty string");
if (!is_id_alpha(c)) {
return 1;
}
fint argc = 0;
for (;;) {
c = *s++;
if (c == '\0') {
break;
} else if (c == ':') {
argc ++;
}
}
return argc;
}
Token* Scanner::get_token() {
Token* t;
if (tokens) {
t = tokens->token;
tokens = tokens->prev;
} else {
t = NULL;
while (t == NULL) {
fint c = get_char();
switch (c) {
case EOF:
depth = 0;
t = new Token(Token::ACCEPT, line, column, sourceAddr());
break;
case '\n':
case '\r':
if (depth <= 0 && !is_string_scanner()) {
t = new Token(Token::ACCEPT, line, column - 1, sourceAddr() - 1);
depth = 0;
}
break;
case ' ':
case '\t':
case '\v':
case '\b':
case '\f':
break;
case '"':
t = skip_comment();
break;
case '(':
depth ++;
t = new Token(as_TokenType('('), line, column - 1, sourceAddr() - 1);
break;
case ')':
depth --;
t = new Token(as_TokenType(')'), line, column - 1, sourceAddr() - 1);
break;
case '[':
depth ++;
t = new Token(as_TokenType('['), line, column - 1, sourceAddr() - 1);
break;
case ']':
depth --;
t = new Token(as_TokenType(']'), line, column - 1, sourceAddr() - 1);
break;
case '.':
t = read_dot();
break;
case '\'':
t = read_string();
break;
case '\\':
c = get_char();
if (c == '\n' || c == '\r') {
// an escaped newline; ignore
} else {
push_char(c);
c = '\\';
t = read_op(c);
}
break;
case '{':
t = new Token(Token::ANNOTATION_START, line, column - 1, sourceAddr() - 1);
break;
case '}':
t = new Token(Token::ANNOTATION_END, line, column - 1, sourceAddr() - 1);
break;
default:
if (is_digit(c) || c == '-') t = read_number(c);
else if (is_id_alpha(c) || c == ':') t = read_name(c);
else if (is_punct(c)) t = read_op(c);
else t = TokenizingError("unknown character in input");
}
}
}
if (t && PrintTokens) t->print();
return t;
}
Token* Scanner::read_name(fint c) {
Token::TokenType t;
fint l = line;
fint col = column - 1;
const char* ss = sourceAddr() - 1;
fint len;
if (c == ':') {
t = Token::ARG;
len = 0;
} else {
t = c == '_' ? Token::PRIMNAME : Token::NAME;
len = 1;
buffer[0] = char(c);
}
while (c = get_char(), is_id_char(c)) {
buffer[len++] = char(c);
}
if (c == ':' && (t == Token::NAME || t == Token::PRIMNAME)) {
buffer[len++] = char(c);
if (is_upper((fint)*buffer)) t = Token::CAPKEYWORD;
else
t = c == '_' ? Token::PRIMKEYWORD : Token::KEYWORD;
}
else {
push_char(c);
}
buffer[len] = '\0';
if (t == Token::ARG && len == 0) {
t = as_TokenType(':');
} else if (t == Token::NAME || t == Token::PRIMNAME) {
c = get_char();
if (c == '.') {
c = get_char();
push_char(c);
if (is_id_alpha(c) || is_punct(c)) {
t = Token::DELEGATE;
} else {
push_char('.');
}
} else {
push_char(c);
}
}
if (strcmp(buffer, "self") == 0) {
if (t == Token::NAME) {
t = Token::SELF_TOKEN;
} else {
return TokenizingError(
"using \"self\" as a parent name for a directed resend");
}
} else if (strcmp(buffer, "resend") == 0) {
if (t == Token::DELEGATE) {
t = Token::RESEND_TOKEN;
} else {
return TokenizingError("not using \"resend\" in a resend");
}
}
String* s;
if (t == Token::NAME || t == Token::PRIMNAME || t == Token::ARG || t == Token::DELEGATE ||
t == Token::KEYWORD || t == Token::PRIMKEYWORD || t == Token::CAPKEYWORD) {
s = new String(copy_string(buffer));
} else {
s = NULL;
}
return new Token(t, s, l, col, ss);
}