def init_parser(): """Set up the parser with the grammar to be recognized. """ # CORE digit = h.ch_range(0x30, 0x39) alpha = h.choice(h.ch_range(0x41, 0x5a), h.ch_range(0x61, 0x7a)) space = h.in_(b" \t\n\r\f\v") # AUX. plus = h.ch(b'+') slash = h.ch(b'/') equals = h.ch(b'=') bsfdig = h.choice(alpha, digit, plus, slash) bsfdig_4bit = h.in_(b"AEIMQUYcgkosw048") bsfdig_2bit = h.in_(b"AQgw") base64_3 = h.repeat_n(bsfdig, 4) base64_2 = h.sequence(bsfdig, bsfdig, bsfdig_4bit, equals) base64_1 = h.sequence(bsfdig, bsfdig_2bit, equals, equals) base64 = h.action( h.sequence(h.many(base64_3), h.optional(h.choice(base64_2, base64_1))), act_base64) # TODO This is not quite the same as the C example, with uses act_ignore. # But I can't get hammer to filter any value returned by act_ignore. ws = h.ignore(h.many(space)) document = h.action(h.sequence(ws, base64, ws, h.end_p()), act_document) # 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
def init_parser(): """Set up the parser with the grammar to be recognized. """ # CORE digit = h.ch_range(0x30, 0x39) alpha = h.choice(h.ch_range(0x41, 0x5a), h.ch_range(0x61, 0x7a)) space = h.in_(" \t\n\r\f\v") # AUX. plus = h.ch('+') slash = h.ch('/') equals = h.ch('=') bsfdig = h.choice(alpha, digit, plus, slash) bsfdig_4bit = h.in_("AEIMQUYcgkosw048") bsfdig_2bit = h.in_("AQgw") base64_3 = h.repeat_n(bsfdig, 4) base64_2 = h.sequence(bsfdig, bsfdig, bsfdig_4bit, equals) base64_1 = h.sequence(bsfdig, bsfdig_2bit, equals, equals) base64 = h.action(h.sequence(h.many(base64_3), h.optional(h.choice(base64_2, base64_1))), act_base64) # TODO This is not quite the same as the C example, with uses act_ignore. # But I can't get hammer to filter any value returned by act_ignore. ws = h.ignore(h.many(space)) document = h.action(h.sequence(ws, base64, ws, h.end_p()), act_document) # 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
def init_parser(): """Return a parser with the grammar to be recognized. """ # CORE # This is a direct translation of the C example. In C the literal 0x30 # is interchangable with the char literal '0' (note the single quotes). # This is not the case in Python. # TODO In the interests of being more Pythonic settle on either string # literals, or integers digit = h.ch_range(0x30, 0x39) alpha = h.choice(h.ch_range(0x41, 0x5a), h.ch_range(0x61, 0x7a)) space = h.in_(" \t\n\r\f\v") # AUX. plus = h.ch('+') slash = h.ch('/') equals = h.action(h.ch('='), act_equals) bsfdig = h.action(h.choice(alpha, digit, plus, slash), act_bsfdig) bsfdig_4bit = h.action(h.in_("AEIMQUYcgkosw048"), act_bsfdig_4bit) bsfdig_2bit = h.action(h.in_("AQgw"), act_bsfdig_2bit) base64_3 = h.action(h.repeat_n(bsfdig, 4), act_base64_3) base64_2 = h.action(h.sequence(bsfdig, bsfdig, bsfdig_4bit, equals), act_base64_2) base64_1 = h.action(h.sequence(bsfdig, bsfdig_2bit, equals, equals), act_base64_1) base64 = h.action(h.sequence(h.many(base64_3), h.optional(h.choice(base64_2, base64_1))), act_base64) # TODO This is not quite the same as the C example, with uses act_ignore. # But I can't get hammer to filter any value returned by act_ignore. ws = h.ignore(h.many(space)) document = h.action(h.sequence(ws, base64, ws, h.end_p()), act_document) # 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
def setUpClass(cls): cls.parser = h.sequence(h.ch("a"), h.ignore(h.ch("b")), h.ch("c"))
def setUpClass(cls): cls.parser = h.sequence(h.ch(b"a"), h.ignore(h.ch(b"b")), h.ch(b"c"))