def LispLexer(text, compiler):
pos = tokens.Position("", 1, 1, 1)
yield tokens.Token(Tag.LP, pos, pos)
yield tokens.Token(Tag.atom, pos, pos)
yield tokens.Token(Tag.atom, pos, pos)
yield tokens.Token(Tag.RP, pos, pos)
yield tokens.Token(Tag.END_OF_PROGRAM, pos, pos)
def main(output=False):
phonecmps = []
timecmps = []
doclist = CreateDoclist()
for pair in PAIRS_:
token1 = tokens.Token(pair[0])
TestPronunciations(token1)
token2 = tokens.Token(pair[1])
TestPronunciations(token2)
comparator = token_comp.OldPhoneticDistanceComparator(token1, token2)
comparator.ComputeDistance()
phonecmps.append(comparator)
if output:
p = open(GOLDEN_FILE_, 'w') ## clear golden file
p.close()
for pc in phonecmps:
pc.ComparisonResult().Print(GOLDEN_FILE_, 'a')
TestCorrelations(doclist, GOLDEN_FILE_)
TestSnowActivations(doclist, GOLDEN_FILE_)
else:
p = open(TEST_FILE_, 'w') ## clear test file
p.close()
for pc in phonecmps:
pc.ComparisonResult().Print(TEST_FILE_, 'a')
TestCorrelations(doclist, TEST_FILE_)
TestSnowActivations(doclist, TEST_FILE_)
unittest.TestUnitOutputs(sys.argv[0] + ' (main test & perceptron test)', \
GOLDEN_FILE_, TEST_FILE_)
TestAuxiliaryComparators(sys.argv[0])
def test_operators(self):
toks = list(lexer.Lexer("+-*/").make_token())
self.assertEqual(toks, [
tokens.Token(tokens.TokenType.PLUS),
tokens.Token(tokens.TokenType.MINUS),
tokens.Token(tokens.TokenType.MULTIPLY),
tokens.Token(tokens.TokenType.DIVIDE)
])
def test_numbers(self):
toks = list(lexer.Lexer("1245.4 .234 1234. .").make_token())
self.assertEqual(toks, [
tokens.Token(tokens.TokenType.NUMBER, 1245.4),
tokens.Token(tokens.TokenType.NUMBER, .234),
tokens.Token(tokens.TokenType.NUMBER, 1234.),
tokens.Token(tokens.TokenType.NUMBER, 000.000),
])
def driver(f, line):
this_state = 0
next_state = 0
tk = tokens.Token()
literal = ""
while this_state < 1000 and this_state > -1:
fpos = f.tell()
datum = f.read(1)
if datum == '#':
while True:
datum = f.read(1)
if datum == '\n':
break
fsa_state = get_column(datum)
next_state = fsa_table[this_state][fsa_state]
if next_state >= 1000 or next_state < 0:
if next_state >= 1000:
tk = get_tokens(next_state, literal, line)
tk.location = line
f.seek(fpos, os.SEEK_SET)
return tk, line
if next_state == -1:
tk.identity = tokens.token_ids.token_names[35]
tk.instance = 'EOF'
tk.location = line
return tk, line
if next_state == -2:
print "SCANNER ERROR: Illegal character '%s' on line %d" % (
datum, line)
tk.identity = tokens.token_ids.token_names[36]
tk.instance = 'bad token'
tk.location = line
return tk, line
else:
unit = datum
if unit in specials and symbols.has_key(unit) == False:
print "SCANNER ERROR: Illegal keyword character '%s' on line %d" % (
datum, line)
tk.identity = tokens.token_ids.token_names[36]
tk.instance = unit
tk.location = line
return tk, line
if unit.isspace() == False:
literal += unit
if len(literal) > 7:
print "SCANNER ERROR: Illegal keyword '%s' on line %d" % (
datum, line)
return tokens.Token(tokens.token_ids.token_names[36],
'illegal size', line)
if datum == '\n':
line = line + 1
this_state = next_state
return tokens.Token(tokens.token_ids.token_names[36], 'bad token', line)
def TestString(self):
program = ast.Program()
program.Statements = [
ast.LetStatement(
tkn.Token(tkn.LET, "let"),
ast.Identifier(tkn.Token(tkn.IDENT, "myVar"), "myVar"),
ast.Identifier(tkn.Token(tkn.IDENT, "anotherVar"),
"anotherVar"))
]
if program.String() != "let myVar = anotherVar":
raise Exception(f'program.String() wrong got={program.String()}')
def test_string():
program = ast.Program(statements=[
ast.LetStatement(token=tokens.Token(typ=tokens.LET, literal="let"),
name=ast.Identifier(token=tokens.Token(
typ=tokens.IDENT, literal="my_var"),
value="my_var"),
value=ast.Identifier(token=tokens.Token(
typ=tokens.IDENT, literal="another_var"),
value="another_var"))
])
expected = "let my_var = another_var;"
assert str(
program
) == expected, f"str(program) wrong. got '{str(program)}'' but expected '{expected}'"
def assem_conv(op, words, line):
logging.debug("Conv %s words:%s" % (op, hl_parser.format_word_list(words)))
token = tokens.Token("conv", err, line)
if (op == "cmptime"):
if (len(words) != 1):
err.report_error("Cmptime needs one arguement")
return
else:
token.add_bits(0,6, 3, 2)
token.add_bits(0, 3, 1, 0)
token.add_bits(0, 0, 7, 7)
if (words[0].type() == "arg"):
token.add_byte(1, words[0].num())
elif (words[0].type() == "var"):
token.add_byte(1, 0)
token.add_vname(1, 0, words[0].val())
else:
err.report_error("Cmptime takes a variable as it's argument")
return
else:
if (len(words) != 0):
err.report_error("Conversions don't that arguments")
return
else:
token.add_bits(0,6, 3, 2)
token.add_bits(0, 0, 7, 3)
if (op == "convm"):
token.add_bits(0, 3, 1, 1)
token.finish(token_stream)
def parse_tokens(text, size, source_id):
"""Parse sentence and return tokens"""
text = re.sub(r'\s+', ' ', text).strip() #remove multiple spaces
string_start = STRING_START_TEXT * (size - 1)
text = string_start + text #append to text some special start tokens
words = split_into_words(text)
if not words: #empty sentence
return []
length = len(words)
lists = []
for i in range(0, length):
wordlist = []
for j in range(0, size):
if i + j < length: #we can append more words
wordlist.append(words[i + j])
lists.append(wordlist)
result = []
length = len(lists)
for i in range(0, length):
is_begin = 1 if i <= size - 1 else 0
is_end = 1 if i + size >= length - 1 else 0
start = ' '.join(lists[i]) if i < length else ''
end = lists[i + size][0] if i + size < length else ''
token = tokens.Token(start, end, source_id, is_begin, is_end)
result.append(token)
return result
def assem_spec_insert(words, line):
logging.debug("Spec_insert words:%s" % (hl_parser.format_word_list(words)))
if (len(words) != 2):
err.report_error("INSERT needs type and filename arguments")
return
type = words[0].astr()
f_name = words[1].astr()
if (type.lower() not in ['tokens', 'binary']):
err.report_error("INSERT type must be one of: 'tokens', 'binary'")
return
if (type.lower() == 'tokens'):
err.report_error("INSERT TOKENS should have been consumed higher up! Eek!");
return
# we have an insert binary special to deal with
if (not os.path.isfile(f_name) or not os.access(f_name, os.R_OK)):
err.report_error("INSERT BINARY file:%s doesn't exist or isn't readable")
return
token = tokens.Token("binary", err, line)
token.add_binary_file(f_name)
token.finish(token_stream)
def LineSegment(self, line):
try:
utext = unicode(line.strip(), 'utf-8')
except TypeError:
utext = line.strip()
for i in range(len(utext)):
for k in [4, 3, 2]:
sub = utext[i:i + k]
if len(sub) != k: continue
if k > 2 and sub[:2].encode('utf-8') in FAMILY_NAMES_:
if not (Utils.script.HasDigit(sub)
or Utils.script.HasPunctuation(sub)):
self.tokens_.append(tokens.Token(sub))
elif k < 4 and sub[:1].encode('utf-8') in FAMILY_NAMES_:
if not Utils.script.HasDigit(sub):
self.tokens_.append(tokens.Token(sub))
def assem_uni_math(op, size, words, line):
logging.debug("Mathu %s size:%d, words:%s" % (op, size, hl_parser.format_word_list(words)))
if (len(words) > 1):
err.report_error("Unary Math has at most one argument")
return
token = tokens.Token("uni-math", err, line)
token.add_bits(0, 6, 3, 2)
token.add_bits(0, 3, 1, size)
token.add_bits(0, 0, 3, uni_assem_map[op])
if (not words or
(words[0].type() == "modreg" and words[0].val() == hl_parser.modreg_names["acc"])):
token.add_bits(0, 2, 1, 0)
token.finish(token_stream)
elif (words[0].type() in ["var", "arg"]):
token.add_bits(0, 2, 1, 1)
if (words[0].type() == "arg"):
token.add_byte(1, words[0].num())
else:
token.add_byte(1, 0)
token.add_vname(1, size, words[0].val())
token.finish(token_stream)
else:
err.report_error("Unary Math - invalid argument type")
return
def assem_basic_math(op, size, words, line):
logging.debug("Mathb %s size:%s, words:%s" % (op, size, hl_parser.format_word_list(words)))
if (len(words) != 1):
err.report_error("Basic Math needs one argument")
return
token = tokens.Token("basic-math", err, line)
token.add_bits(0, 6, 3, 2)
token.add_bits(0, 4, 3, 1)
token.add_bits(0, 3, 1, size)
token.add_bits(0, 0, 3, basic_assem_map[op])
if (words[0].type() in ["var", "arg"]):
# math with a variable
token.add_bits(0, 2, 1, 0)
if (words[0].type() == "arg"):
token.add_byte(1, words[0].num())
else:
token.add_byte(1, 0)
token.add_vname(1, size, words[0].val())
token.finish(token_stream)
elif (words[0].type() == "const"):
token.add_bits(0, 2, 1, 1)
if (size == 0):
token.add_byte(1, words[0].val())
else:
token.add_word(1, words[0].val())
token.finish(token_stream)
else:
err.report_error("Basic Math - invalid argument type")
return
def push_macro(self, optimizer):
macro = optimizer.rewind()
end = tokens.Token(tokens.Token.END)
optimizer.push_node(end)
optimizer.push_node(macro.statements)
optimizer.open_scope()
return True
def assem_other_math(op, size, words, line):
logging.debug("Matho %s size:%d, words:%s" % (op, size, hl_parser.format_word_list(words)))
if (len(words) != 1):
err.report_error("Logic Math needs one arguement")
return
token = tokens.Token("log-math", err, line)
token.add_bits(0, 6, 3, 2)
token.add_bits(0, 3, 1, size)
if (op in other1_assem_map.keys()):
token.add_bits(0,4,3,2)
token.add_bits(0,0,3, other1_assem_map[op])
else:
token.add_bits(0,4,3,3)
token.add_bits(0,0,3, other2_assem_map[op])
if (words[0].type() in ["var", "arg"]):
# math with a variable
token.add_bits(0, 2, 1, 0)
if (words[0].type() == "arg"):
token.add_byte(1, words[0].val())
else:
token.add_byte(1, 0)
token.add_vname(1, size, words[0].val())
token.finish(token_stream)
elif (words[0].type() == "const"):
token.add_bits(0, 2, 1, 1)
token.add_byte(1, words[0].val())
token.finish(token_stream)
else:
err.report_error("Logic Math - invalid argument type")
return
def assem_jump(op, cond, words, line):
logging.debug("Jump %s cond:%s, words:%s" %
(op, cond, hl_parser.format_word_list(words)))
# ops is one of: branch, sub, ret, dbnz, dsnz
# cond is one of: a, e, ne, g, l, le, lg or empty for ret, d?nz
token = tokens.Token("jump", err, line)
token.add_bits(0, 6, 3, 3)
if (op == "ret"):
if (len(words) != 0):
err.report_error("Ret don't take an argument")
return
else:
token.add_bits(0, 0, 0x3f, 0x28)
elif (len(words) != 1):
err.report_error("Jumps need a target to jump to")
return
else:
if (op[:2] in ['su', 'ds']):
# a call to a subroutine - push a frame
token.add_bits(0, 3, 1, 1)
else:
token.add_bits(0, 3, 1, 0)
if (not cond):
# one of dbnz or dsnz
token.add_bits(0, 0, 7, 7)
else:
token.add_bits(0, 0, 7, jcond_assem_map[cond])
# now the target
if (words[0].type() == "const"):
offset = words[0].val()
if (offset >= tokens.MIN_SBYTE and offset <= tokens.MAX_SBYTE):
if (offset < 0):
# convert to a signed offset
offset += 256
token.add_byte(1, offset)
elif (offset >= tokens.MIN_WORD and offsec <= tokens.MAX_WORD):
token.add_bits(0, 4, 1, 1)
token.add_word(1, offset)
elif (words[0].type() == "label"):
if (words[0].val().startswith(':')):
# globals are always long jumps
token.add_bits(0, 4, 1, 1)
token.set_jump_label(1, words[0].val(), True)
token.add_word(1, 0) # A placeholder
else:
token.set_jump_label(1, words[0].val())
token.add_byte(1, 0) # A placeholder
else:
err.report_error(
"Jumps need either a constant or a label as argument, not a: "
+ words[0].type())
return
token.finish(token_stream)
def make_number(self):
num_str = ''
dot_count = 0
pos_start = self.pos.copy()
while self.current_char != None and self.current_char in constants.DIGITS + '.':
if self.current_char == '.':
if dot_count == 1:
break
dot_count += 1
num_str += self.current_char
self.advance()
if dot_count == 0:
return t.Token(t.TT_INT, int(num_str), pos_start, self.pos)
else:
return t.Token(t.TT_FLOAT, float(num_str), pos_start, self.pos)
def tester(fn):
t = tokens.Token()
line = 1
with open(fn) as f:
while True:
t, line = scanner.driver(f, line)
print "%s '%s' on line %d" % (t.identity, t.instance, t.location)
if t.identity == tokens.token_ids.token_names[36]: break
if t.identity == tokens.token_ids.token_names[35]: break
def _tokenize_line(cls, _line: str) -> typing.List[typing.Any]:
_result = []
for i in re.findall(cls.grammer, _line):
print(i)
_r = [[*c, i] for c in cls.token_list if re.findall(c[-1], i)][0]
print(_r)
_result.append(tokens.Token(*_r))
return _result
def make_identifier(self):
id_str = ''
pos_start = self.pos.copy()
while self.current_char != None and self.current_char in constants.LETTERS_DIGITS + '_':
id_str += self.current_char
self.advance()
tok_type = t.TT_KEYWORD if id_str in t.KEYWORDS else t.TT_IDENTIFIER
return t.Token(tok_type, id_str, pos_start, self.pos)
def make_greater_than(self):
tok_type = t.TT_GT
pos_start = self.pos.copy()
self.advance()
if self.current_char == '=':
self.advance()
tok_type = t.TT_GTE
return t.Token(tok_type, pos_start=pos_start, pos_end=self.pos)
def make_minus_or_arrow(self):
tok_type = t.TT_MINUS
pos_start = self.pos.copy()
self.advance()
if self.current_char == '>':
self.advance()
tok_type = t.TT_ARROW
return t.Token(tok_type, pos_start=pos_start, pos_end=self.pos)
def make_not_equals(self):
pos_start = self.pos.copy()
self.advance()
if self.current_char == '=':
self.advance()
return t.Token(t.TT_NE, pos_start=pos_start, pos_end=self.pos), None
self.advance()
return None, e.ExpectedCharError(pos_start, self.pos, "'=' (after '!')")
def make_equals(self):
tok_type = t.TT_EQ
pos_start = self.pos.copy()
self.advance()
if self.current_char == '=':
self.advance()
tok_type = t.TT_EE
return t.Token(tok_type, pos_start=pos_start, pos_end=self.pos)
def assem_stack(op, size, words, line):
logging.debug("Stack %s size:%d, words:%s" % (op, size, hl_parser.format_word_list(words)))
# op in ['push', 'pop'], size in [0, 1], one word
if (len(words) != 1):
err.report_error("Stack ops need 1 argument")
return
token = tokens.Token("stack", err, line)
token.add_bits(0, 6, 3, 3)
token.add_bits(0, 5, 1, 1)
token.add_bits(0, 4, 1, size)
if (op == "push"):
if (words[0].type() == 'const'):
if (size == 0):
token.add_byte(1, words[0].val())
else:
token.add_word(1, words[0].val())
elif (words[0].type() == 'modreg'):
if (words[0].val() == hl_parser.modreg_names["acc"]):
token.add_bits(0, 0, 0xf, 0x3)
else:
token.add_bits(0, 0, 0xf, 2)
token.add_byte(1, words[0].val())
elif (words[0].type() in ["var", "arg"]):
token.add_bits(0, 0, 0xf, 1)
if (words[0].type() == "arg"):
token.add_byte(1, words[0].num())
else:
token.add_byte(1, 0)
token.add_vname(1, size, words[0].val())
else:
err.report_error("Push - invalid operand: %s" % (words[0].type()))
return
else: # pop
if (words[0].type() == 'modreg'):
if (words[0].val() == hl_parser.modreg_names["acc"]):
token.add_bits(0, 0, 0xf, 0x4)
else:
token.add_bits(0, 0, 0xf, 6)
token.add_byte(1, words[0].val())
elif (words[0].type() in ["var", "arg"]):
token.add_bits(0, 0, 0xf, 5)
if (words[0].type() == "arg"):
token.add_byte(1, words[0].num())
else:
token.add_byte(1, 0)
token.add_vname(1, size, words[0].val())
else:
err.report_error("Pop - invalid operand: %s" % (words[0].type()))
return
token.finish(token_stream)
def LineSegment(self, line):
try: utext = unicode(line.strip(), 'utf-8')
except TypeError: utext = line.strip()
word = []
for u in utext:
if Utils.script.CharacterToScript(u) == 'Katakana':
word.append(u.encode('utf-8'))
else:
if word and word != ['・']:
self.tokens_.append(tokens.Token(''.join(word)))
word = []
def get_tokens(state, literal, line):
state_token = tokens.Token()
if literal in keywords:
state_token.identity = keywords.get(literal)
state_token.instance = literal
state_token.location = line
elif final_states.has_key(state):
state_token.identity = final_states.get(state)
state_token.instance = literal
state_token.location = line
return state_token
def TestAuxiliaryComparators(unitname):
## Added tests for Wade-Giles and Pinyin comparators
t1 = tokens.Token('毛泽东')
t2 = tokens.Token('周恩来')
t1py = tokens.Token('Mao Zedong')
t2py = tokens.Token('Zhou Enlai')
t1wg = tokens.Token('Mao Tse-tung')
t2wg = tokens.Token('Chou Enlai')
comparator = auxiliary_comp.PinyinComparator(t1, t1py)
comparator.ComputeDistance()
assert comparator.ComparisonResult().Cost() == auxiliary_comp.MATCH_, \
'%s should match %s' % (t1.String(), t1py.String())
comparator = auxiliary_comp.PinyinComparator(t2, t2py)
comparator.ComputeDistance()
assert comparator.ComparisonResult().Cost() == auxiliary_comp.MATCH_, \
'%s should match %s' % (t2.String(), t2py.String())
comparator = auxiliary_comp.WadeGilesComparator(t1, t1wg)
comparator.ComputeDistance()
assert comparator.ComparisonResult().Cost() == auxiliary_comp.MATCH_, \
'%s should match %s' % (t1.String(), t1wg.String())
comparator = auxiliary_comp.WadeGilesComparator(t2, t2wg)
comparator.ComputeDistance()
assert comparator.ComparisonResult().Cost() == auxiliary_comp.MATCH_, \
'%s should match %s' % (t2.String(), t2wg.String())
comparator = auxiliary_comp.WadeGilesComparator(t2, t2py)
comparator.ComputeDistance()
assert comparator.ComparisonResult().Cost() == auxiliary_comp.NO_MATCH_, \
'%s should not match %s' % (t2.String(), t2py.String())
print '%s (auxiliary tests) successful' % unitname
def assem_misc(op, words, line):
logging.debug("Misc op:%s words:%s" %
(op, hl_parser.format_word_list(words)))
if (op == "stop"):
if (len(words) != 0):
err.report_error("Stop doesn't take arguments")
return
token = tokens.Token("misc", err, line)
token.add_bits(0, 0, 0xff, 0xff)
token.finish(token_stream)
elif (op in ["bitset", "bitclr"]):
if (len(words) != 2):
err.report_error(
"Bitset/bitclr needs 2 arguments: bit and mod/reg")
return
bit = words[0].anum()
if (bit < 0 or bit > 7):
err.report_error(
"Bitset/bitclr bit must be between 0 and 7 (not %d)" % (bit))
return
modreg = words[1].amodreg()
token = tokens.Token("misc", err, line)
token.add_bits(0, 4, 0x0f, 0x00)
if (op == "bitset"):
token.add_bits(0, 3, 0x1, 0x1)
else:
token.add_bits(0, 3, 0x1, 0x0)
token.add_bits(0, 0, 0x7, bit)
token.add_byte(1, modreg)
token.finish(token_stream)
else:
err.report_error("Unknown misc operator: " + op)
return
def parseSimpleState(text):
# Parses a simple statement into a token list
# Assumes that the inputted text is a valid statement.
# But we check anyway
if text == None:
return None
tokenList = []
tokenlistpos = 0
pos = 0
# This loop handles most of the things
while (pos < len(text)):
cc = text[pos] # cc = current char
if cc.isalnum() == True:
# Just avoiding errors with the if statement below
if pos > 0:
if text[pos - 1].isalnum == True:
tokenList[tokenlistpos].value += cc
else:
tokenList.append(tokens.Token(tokens.OD_NAME, cc))
else:
tokenList.append(tokens.Token(tokens.OD_NAME, cc))
tokenlistpos = len(tokenList)
elif cc == '+' or cc == '-' or cc == '*' or cc == '-' or cc == '=':
tokenList.append(tokens.Token(tokens.OP, cc))
tokenlistpos = len(tokenList)
else:
tokenlistpos = len(tokenList)
pos += 1
return tokenList