const HParser *init_parser(void)
{
// CORE
H_RULE (digit, h_ch_range(0x30, 0x39));
H_RULE (alpha, h_choice(h_ch_range(0x41, 0x5a), h_ch_range(0x61, 0x7a), NULL));
H_RULE (space, h_in((uint8_t *)" \t\n\r\f\v", 6));
// AUX.
H_RULE (plus, h_ch('+'));
H_RULE (slash, h_ch('/'));
H_ARULE(equals, h_ch('='));
H_ARULE(bsfdig, h_choice(alpha, digit, plus, slash, NULL));
H_ARULE(bsfdig_4bit, h_in((uint8_t *)"AEIMQUYcgkosw048", 16));
H_ARULE(bsfdig_2bit, h_in((uint8_t *)"AQgw", 4));
H_ARULE(base64_3, h_repeat_n(bsfdig, 4));
H_ARULE(base64_2, h_sequence(bsfdig, bsfdig, bsfdig_4bit, equals, NULL));
H_ARULE(base64_1, h_sequence(bsfdig, bsfdig_2bit, equals, equals, NULL));
H_ARULE(base64, h_sequence(h_many(base64_3),
h_optional(h_choice(base64_2,
base64_1, NULL)),
NULL));
H_ARULE(ws, h_many(space));
H_ARULE(document, h_sequence(ws, base64, ws, h_end_p(), NULL));
// BUG sometimes inputs that should just don't parse.
// It *seemed* to happen mostly with things like "bbbbaaaaBA==".
// Using less actions seemed to make it less likely.
return document;
}
static void test_llk_zero_end(void) {
HParserBackend be = PB_LLk;
HParser *z = h_ch('\x00');
HParser *az = h_sequence(h_ch('a'), z, NULL);
HParser *ze = h_sequence(z, h_end_p(), NULL);
HParser *aze = h_sequence(h_ch('a'), z, h_end_p(), NULL);
// some cases surrounding the bug
g_check_parse_match (z, be, "\x00", 1, "u0");
g_check_parse_failed(z, be, "", 0);
g_check_parse_match (ze, be, "\x00", 1, "(u0)");
g_check_parse_failed(ze, be, "\x00b", 2);
g_check_parse_failed(ze, be, "", 0);
g_check_parse_match (az, be, "a\x00", 2, "(u0x61 u0)");
g_check_parse_match (aze, be, "a\x00", 2, "(u0x61 u0)");
g_check_parse_failed(aze, be, "a\x00b", 3);
// the following should not parse but did when the LL(k) backend failed to
// check for the end of input, mistaking it for a zero character.
g_check_parse_failed(az, be, "a", 1);
g_check_parse_failed(aze, be, "a", 1);
}
void init_parser(void)
{
// CORE
HParser *digit = h_ch_range(0x30, 0x39);
HParser *alpha = h_choice(h_ch_range(0x41, 0x5a), h_ch_range(0x61, 0x7a), NULL);
// AUX.
HParser *plus = h_ch('+');
HParser *slash = h_ch('/');
HParser *equals = h_ch('=');
HParser *bsfdig = h_choice(alpha, digit, plus, slash, NULL);
HParser *bsfdig_4bit = h_in((uint8_t *)"AEIMQUYcgkosw048", 16);
HParser *bsfdig_2bit = h_in((uint8_t *)"AQgw", 4);
HParser *base64_3 = h_repeat_n(bsfdig, 4);
HParser *base64_2 = h_sequence(bsfdig, bsfdig, bsfdig_4bit, equals, NULL);
HParser *base64_1 = h_sequence(bsfdig, bsfdig_2bit, equals, equals, NULL);
HParser *base64 = h_sequence(h_many(base64_3),
h_optional(h_choice(base64_2,
base64_1, NULL)),
NULL);
document = h_sequence(h_whitespace(base64), h_whitespace(h_end_p()), NULL);
}
const HParser* init_rdata(uint16_t type) {
static const HParser *parsers[RDATA_TYPE_MAX+1];
static int inited = 0;
if (type >= sizeof(parsers))
return NULL;
if (inited)
return parsers[type];
H_RULE (domain, init_domain());
H_ARULE(cstr, init_character_string());
H_RULE (a, h_uint32());
H_RULE (ns, domain);
H_RULE (md, domain);
H_RULE (mf, domain);
H_RULE (cname, domain);
H_ARULE(soa, h_sequence(domain, // MNAME
domain, // RNAME
h_uint32(), // SERIAL
h_uint32(), // REFRESH
h_uint32(), // RETRY
h_uint32(), // EXPIRE
h_uint32(), // MINIMUM
NULL));
H_RULE (mb, domain);
H_RULE (mg, domain);
H_RULE (mr, domain);
H_VRULE(null, h_many(h_uint8()));
H_RULE (wks, h_sequence(h_uint32(),
h_uint8(),
h_many(h_uint8()),
NULL));
H_RULE (ptr, domain);
H_RULE (hinfo, h_sequence(cstr, cstr, NULL));
H_RULE (minfo, h_sequence(domain, domain, NULL));
H_RULE (mx, h_sequence(h_uint16(), domain, NULL));
H_ARULE(txt, h_many1(cstr));
parsers[ 0] = NULL; // there is no type 0
parsers[ 1] = a;
parsers[ 2] = ns;
parsers[ 3] = md;
parsers[ 4] = mf;
parsers[ 5] = cname;
parsers[ 6] = soa;
parsers[ 7] = mb;
parsers[ 8] = mg;
parsers[ 9] = mr;
parsers[10] = null;
parsers[11] = wks;
parsers[12] = ptr;
parsers[13] = hinfo;
parsers[14] = minfo;
parsers[15] = mx;
parsers[16] = txt;
// All parsers must consume their input exactly.
for(uint16_t i; i<sizeof(parsers); i++) {
if(parsers[i]) {
parsers[i] = h_action(h_sequence(parsers[i], h_end_p(), NULL),
act_index0);
}
}
inited = 1;
return parsers[type];
}
