static const char *
lex_parse_macro(const char *p, struct lex_token *token)
{
p++;
if (!lex_is_id1(*p)) {
lex_error(token, "`$' must be followed by a valid identifier.");
return p;
}
return lex_parse_id(p, LEX_T_MACRO, token);
}
/* Initializes 'token' and parses the first token from the beginning of
* null-terminated string 'p' into 'token'. Stores a pointer to the start of
* the token (after skipping white space and comments, if any) into '*startp'.
* Returns the character position at which to begin parsing the next token. */
const char *
lex_token_parse(struct lex_token *token, const char *p, const char **startp)
{
lex_token_init(token);
next:
*startp = p;
switch (*p) {
case '\0':
token->type = LEX_T_END;
return p;
case ' ': case '\t': case '\n': case '\r':
p++;
goto next;
case '/':
p++;
if (*p == '/') {
do {
p++;
} while (*p != '\0' && *p != '\n');
goto next;
} else if (*p == '*') {
p++;
for (;;) {
if (*p == '*' && p[1] == '/') {
p += 2;
goto next;
} else if (*p == '\0' || *p == '\n') {
lex_error(token, "`/*' without matching `*/'.");
return p;
} else {
p++;
}
}
goto next;
} else {
lex_error(token,
"`/' is only valid as part of `//' or `/*'.");
}
break;
case '(':
token->type = LEX_T_LPAREN;
p++;
break;
case ')':
token->type = LEX_T_RPAREN;
p++;
break;
case '{':
token->type = LEX_T_LCURLY;
p++;
break;
case '}':
token->type = LEX_T_RCURLY;
p++;
break;
case '[':
token->type = LEX_T_LSQUARE;
p++;
break;
case ']':
token->type = LEX_T_RSQUARE;
p++;
break;
case '=':
p++;
if (*p == '=') {
token->type = LEX_T_EQ;
p++;
} else {
token->type = LEX_T_EQUALS;
}
break;
case '!':
p++;
if (*p == '=') {
token->type = LEX_T_NE;
p++;
} else {
token->type = LEX_T_LOG_NOT;
}
break;
case '&':
p++;
if (*p == '&') {
token->type = LEX_T_LOG_AND;
p++;
} else {
lex_error(token, "`&' is only valid as part of `&&'.");
}
break;
case '|':
p++;
if (*p == '|') {
token->type = LEX_T_LOG_OR;
p++;
} else {
lex_error(token, "`|' is only valid as part of `||'.");
}
break;
case '<':
p++;
if (*p == '=') {
token->type = LEX_T_LE;
p++;
} else if (*p == '-' && p[1] == '>') {
token->type = LEX_T_EXCHANGE;
p += 2;
} else {
token->type = LEX_T_LT;
}
break;
case '>':
p++;
if (*p == '=') {
token->type = LEX_T_GE;
p++;
} else {
token->type = LEX_T_GT;
}
break;
case '.':
p++;
if (*p == '.') {
token->type = LEX_T_ELLIPSIS;
p++;
} else {
lex_error(token, "`.' is only valid as part of `..' or a number.");
}
break;
case ',':
p++;
token->type = LEX_T_COMMA;
break;
case ';':
p++;
token->type = LEX_T_SEMICOLON;
break;
case '-':
p++;
if (*p == '-') {
token->type = LEX_T_DECREMENT;
p++;
} else {
lex_error(token, "`-' is only valid as part of `--'.");
}
break;
case '0': case '1': case '2': case '3': case '4':
case '5': case '6': case '7': case '8': case '9':
case ':':
p = lex_parse_integer(p, token);
break;
case '"':
p = lex_parse_string(p, token);
break;
case 'a': case 'b': case 'c': case 'd': case 'e': case 'f':
case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
/* We need to distinguish an Ethernet address or IPv6 address from an
* identifier. Fortunately, Ethernet addresses and IPv6 addresses that
* are ambiguous based on the first character, always start with hex
* digits followed by a colon, but identifiers never do. */
p = (p[strspn(p, "0123456789abcdefABCDEF")] == ':'
? lex_parse_integer(p, token)
: lex_parse_id(p, token));
break;
default:
if (lex_is_id1(*p)) {
p = lex_parse_id(p, token);
} else {
if (isprint((unsigned char) *p)) {
lex_error(token, "Invalid character `%c' in input.", *p);
} else {
lex_error(token, "Invalid byte 0x%d in input.", *p);
}
p++;
}
break;
}
return p;
}
static bool
lex_is_idn(unsigned char c)
{
return lex_is_id1(c) || (c >= '0' && c <= '9');
}