411 lines
14 KiB
Python
Executable File
411 lines
14 KiB
Python
Executable File
import re
|
|
|
|
valid_name_re = re.compile('^[a-zA-Z_][a-zA-Z0-9_]*$')
|
|
reserved_names = ['start', 'null', 'end']
|
|
|
|
class Token(object):
|
|
def __init__(self, name, y, x, s, **vargs):
|
|
self.name = name
|
|
self.y = y
|
|
self.x = x
|
|
self.string = s
|
|
self.vargs = vargs
|
|
def add_to_string(self, s):
|
|
self.string += s
|
|
def __repr__(self):
|
|
if len(self.string) < 10:
|
|
s = self.string
|
|
else:
|
|
s = self.string[:10] + '...'
|
|
return "<Token(%r, %d, %d, %r)>" % (self.name, self.y, self.x, s)
|
|
def render(self):
|
|
return (self,)
|
|
|
|
class Rule:
|
|
name = 'abstract'
|
|
def match(self, lexer, context=[]):
|
|
raise Exception, "%s rule cannot match!" % self.name
|
|
def make_token(self, lexer, s, name, **vargs):
|
|
return Token(name, lexer.y, lexer.x, s, **vargs)
|
|
|
|
class ConstantRule(Rule):
|
|
def __init__(self, name, constant):
|
|
assert valid_name_re.match(name), 'invalid name %r' % name
|
|
assert name not in reserved_names, "reserved rule name: %r" % name
|
|
self.name = name
|
|
self.constant = constant
|
|
def match(self, lexer, context=[]):
|
|
if lexer.lines[lexer.y][lexer.x:].startswith(self.constant):
|
|
lexer.add_token(self.make_token(lexer, self.constant, self.name))
|
|
lexer.x += len(self.constant)
|
|
return True
|
|
else:
|
|
return False
|
|
|
|
class PatternRule(Rule):
|
|
def __init__(self, name, pattern):
|
|
assert valid_name_re.match(name), 'invalid name %r' % name
|
|
assert name not in reserved_names, "reserved rule name: %r" % name
|
|
self.name = name
|
|
self.pattern = pattern
|
|
self.re = re.compile(pattern)
|
|
def match(self, lexer, context=[]):
|
|
m = self.re.match(lexer.lines[lexer.y], lexer.x)
|
|
if m:
|
|
lexer.add_token(self.make_token(lexer, m.group(0), self.name))
|
|
lexer.x += len(m.group(0))
|
|
return True
|
|
else:
|
|
return False
|
|
|
|
class RegionRule(Rule):
|
|
def __init__(self, name, start, grammar, end):
|
|
assert valid_name_re.match(name), 'invalid name %r' % name
|
|
assert name not in reserved_names, "reserved rule name: %r" % name
|
|
self.name = name
|
|
self.start = start
|
|
self.grammar = grammar
|
|
self.end = end
|
|
self.start_re = re.compile(start)
|
|
def _add_from_regex(self, context, name, lexer, m):
|
|
t_name = '.'.join(context + [self.name, name])
|
|
t = self.make_token(lexer, m.group(0), t_name)
|
|
lexer.add_token(t)
|
|
lexer.x += len(m.group(0))
|
|
def match(self, lexer, context=[]):
|
|
m = self.start_re.match(lexer.lines[lexer.y], lexer.x)
|
|
if m:
|
|
self._add_from_regex(context, 'start', lexer, m)
|
|
|
|
null_t_name = '.'.join(context + [self.name, 'null'])
|
|
null_t = None
|
|
|
|
if self.end:
|
|
end_re = re.compile(self.end % m.groupdict())
|
|
|
|
done = False
|
|
while not done and lexer.y < len(lexer.lines):
|
|
line = lexer.lines[lexer.y]
|
|
if len(line) == 0:
|
|
null_t = Token(null_t_name, lexer.y, lexer.x, '')
|
|
lexer.add_token(null_t)
|
|
while not done and lexer.x < len(line):
|
|
if self.end:
|
|
m = end_re.match(line, lexer.x)
|
|
if m:
|
|
self._add_from_regex(context, 'end', lexer, m)
|
|
done = True
|
|
continue
|
|
|
|
found = False
|
|
for rule in self.grammar.rules:
|
|
if rule.match(lexer, context + [self.name]):
|
|
found = True
|
|
null_t = None
|
|
break
|
|
if not found:
|
|
if null_t is None:
|
|
null_t = Token(null_t_name, lexer.y, lexer.x, '')
|
|
lexer.add_token(null_t)
|
|
null_t.add_to_string(line[lexer.x])
|
|
lexer.x += 1
|
|
|
|
null_t = None
|
|
lexer.y += 1
|
|
lexer.x = 0
|
|
return True
|
|
else:
|
|
return False
|
|
|
|
class Grammar:
|
|
rules = []
|
|
|
|
class NullGrammar(Grammar):
|
|
pass
|
|
|
|
class PodGrammar(Grammar):
|
|
pass
|
|
|
|
class StringGrammar(Grammar):
|
|
rules = [
|
|
PatternRule('escaped', '\\.'),
|
|
]
|
|
|
|
class TestGrammar(Grammar):
|
|
rules = [
|
|
RegionRule(
|
|
name='heredoc',
|
|
start="<< *(?P<heredoc>[a-zA-Z0-9_]+) *;",
|
|
grammar=StringGrammar(),
|
|
end='^%(heredoc)s$',
|
|
),
|
|
RegionRule(
|
|
name='string1',
|
|
start='"',
|
|
grammar=StringGrammar(),
|
|
end='"',
|
|
),
|
|
RegionRule(
|
|
name='string2',
|
|
start="'",
|
|
grammar=StringGrammar(),
|
|
end="'",
|
|
),
|
|
PatternRule(
|
|
name='word',
|
|
pattern='[^ \t\n]+',
|
|
),
|
|
]
|
|
|
|
class PerlGrammar(Grammar):
|
|
rules = [
|
|
RegionRule(
|
|
name='heredoc',
|
|
start="<< *(?P<heredoc>[a-zA-Z0-9_]+) *;",
|
|
grammar=StringGrammar(),
|
|
end='^%(heredoc)s$',
|
|
),
|
|
RegionRule(
|
|
name='endblock',
|
|
start="^__END__|__DATA__ *$",
|
|
grammar=NullGrammar(),
|
|
end='',
|
|
),
|
|
RegionRule(
|
|
name='pod',
|
|
start='^=[a-zA-Z0-9_]+',
|
|
grammar=PodGrammar(),
|
|
end='^=cut',
|
|
),
|
|
PatternRule(
|
|
name='comment',
|
|
pattern='#.*$',
|
|
),
|
|
RegionRule(
|
|
name='string1',
|
|
start='"',
|
|
grammar=StringGrammar(),
|
|
end='"',
|
|
),
|
|
RegionRule(
|
|
name='string2',
|
|
start="'",
|
|
grammar=StringGrammar(),
|
|
end="'",
|
|
),
|
|
RegionRule(
|
|
name='evalstring',
|
|
start="`",
|
|
grammar=StringGrammar(),
|
|
end="`",
|
|
),
|
|
PatternRule(
|
|
name='number',
|
|
pattern='0?\.[0-9]+|[0-9]+(?:\.[0-9]+)?',
|
|
),
|
|
PatternRule(
|
|
name='keyword',
|
|
pattern="(?<!->)(?:STDIN|STDERR|STDOUT|and|cmp|continue|do|else|elsif|eq|eval|foreach|for|if|last|my|next|ne|not|or|our|package|require|return|sub|undef|unless|until|use|while)(?![a-zA-Z0-9_])",
|
|
),
|
|
PatternRule(
|
|
name='hash_bareword_index',
|
|
pattern='(?<={) *[A-Za-z0-9_]+(?=})',
|
|
),
|
|
PatternRule(
|
|
name='literal_hash_bareword_index',
|
|
pattern='[A-Za-z0-9_]+(?= *=>)',
|
|
),
|
|
PatternRule(
|
|
name='length_scalar',
|
|
pattern=r"\$#[A-Za-z0-9_](?:[A-Za-z0-9_]|::)*",
|
|
),
|
|
PatternRule(
|
|
name='system_scalar',
|
|
pattern=r"\$[][><ab/'\"_@\?#\$!%^|&*()](?![A-Za-z0-9_])",
|
|
),
|
|
PatternRule(
|
|
name='system_array',
|
|
pattern="@_",
|
|
),
|
|
PatternRule(
|
|
name='scalar',
|
|
pattern="\$\$*[A-Za-z0-9_](?:[A-Za-z0-9_]|::)*",
|
|
),
|
|
PatternRule(
|
|
name='array',
|
|
pattern="@\$*[A-Za-z_](?:[A-Za-z0-9_]|::)*",
|
|
),
|
|
PatternRule(
|
|
name='hash',
|
|
pattern="%\$*[A-Za-z_](?:[A-Za-z0-9_]|::)*",
|
|
),
|
|
PatternRule(
|
|
name='dereference',
|
|
pattern="[@%\$&\*](?={)",
|
|
),
|
|
RegionRule(
|
|
name='quoted_region1',
|
|
start='q[rqwx]? *(?P<delim>[^ #])',
|
|
grammar=Grammar(),
|
|
end='%(delim)s',
|
|
),
|
|
RegionRule(
|
|
name='quoted_region2',
|
|
start='q[rqwx]?#',
|
|
grammar=Grammar(),
|
|
end='#',
|
|
),
|
|
RegionRule(
|
|
name='bracket_quoted_region1',
|
|
start='q[rqwx]? *\(',
|
|
grammar=Grammar(),
|
|
end='\)',
|
|
),
|
|
RegionRule(
|
|
name='bracket_quoted_region2',
|
|
start='q[rqwx]? *{',
|
|
grammar=Grammar(),
|
|
end='}',
|
|
),
|
|
RegionRule(
|
|
name='bracket_quoted_region3',
|
|
start='q[rqwx]? *<',
|
|
grammar=Grammar(),
|
|
end='>',
|
|
),
|
|
RegionRule(
|
|
name='bracket_quoted_region4',
|
|
start='q[rqwx]? *\[',
|
|
grammar=Grammar(),
|
|
end='\]',
|
|
),
|
|
RegionRule(
|
|
name='implicit_match_regex',
|
|
start='(?:(?<==~)|(?<=!~)|(?<=\()) */',
|
|
grammar=StringGrammar(),
|
|
end='/',
|
|
),
|
|
RegionRule(
|
|
name='explicit_match_regex1',
|
|
start='m *(?P<delim>[^ #])',
|
|
grammar=StringGrammar(),
|
|
end='%(delim)s',
|
|
),
|
|
RegionRule(
|
|
name='explicit_match_regex1',
|
|
start='m#',
|
|
grammar=StringGrammar(),
|
|
end='#',
|
|
),
|
|
# we officially don't support the bullshit s{a}{b} thing perl has going.
|
|
# those guys are on crack. we only support things like s#a#b# or s/a/b/.
|
|
# same comments as above apply
|
|
#{'name': 'replace regex',
|
|
# 'expr': r"""(?:y|tr|s)([^<[{(A-Za-z0-9 \t\n])(?:\\.|[^\\])*?\1(?:\\.|[^\\])*?\1[a-z]*""",
|
|
# 'action': lex.make_token},
|
|
|
|
PatternRule(
|
|
name='package',
|
|
pattern="(?<=package )(?:[a-zA-Z_][a-zA-Z_0-9]*::)*[a-zA-Z_][a-zA-Z_0-9]*",
|
|
),
|
|
PatternRule(
|
|
name='sub',
|
|
pattern="(?<=sub )[a-zA-Z_][a-zA-Z_0-9]*(?=[ \n]*{)",
|
|
),
|
|
PatternRule(
|
|
name='use',
|
|
pattern="(?<=use )(?:[a-zA-Z_][a-zA-Z_0-9]*::)*[a-zA-Z_][a-zA-Z_0-9]*",
|
|
),
|
|
PatternRule(
|
|
name='label',
|
|
pattern='[a-zA-Z_][a-zA-Z0-9_]*:',
|
|
),
|
|
PatternRule(
|
|
name='instance_method',
|
|
pattern="(?<=->)[a-zA-Z_][a-zA-Z_0-9]*",
|
|
),
|
|
PatternRule(
|
|
name='static_method',
|
|
pattern="&?(?:[a-zA-Z_][a-zA-Z_0-9]*::)+[a-zA-Z_][a-zA-Z_0-9]*",
|
|
),
|
|
PatternRule(
|
|
name='builtin_method',
|
|
pattern="(?<!->)&?(?:write|warn|wantarray|waitpid|wait|vec|values|utime|use|untie|unshift|unpack|unlink|undef|umask|ucfirst|uc|truncate|times|time|tied|tie|telldir|tell|syswrite|system|sysseek|sysread|sysopen|syscall|symlink|substr|sub|study|stat|srand|sqrt|sprintf|split|splice|sort|socketpair|socket|sleep|sin|shutdown|shmwrite|shmread|shmget|shmctl|shift|setsockopt|setservent|setpwent|setprotoent|setpriority|setpgrp|setnetent|sethostent|setgrent|send|semop|semget|semctl|select|seekdir|seek|scalar|rmdir|rindex|rewinddir|reverse|return|reset|require|rename|ref|redo|recv|readpipe|readlink|readline|readdir|read|rand|quotemeta|push|prototype|printf|print|pos|pop|pipe|package|pack|our|ord|opendir|open|oct|no|next|my|msgsnd|msgrcv|msgget|msgctl|mkdir|map|lstat|log|lock|localtime|local|listen|link|length|lcfirst|lc|last|kill|keys|join|ioctl|int|index|import|hex|grep|goto|gmtime|glob|getsockopt|getsockname|getservent|getservbyport|getservbyname|getpwuid|getpwnam|getpwent|getprotoent|getprotobynumber|getprotobyname|getpriority|getppid|getpgrp|getpeername|getnetent|getnetbyname|getnetbyaddr|getlogin|gethostent|gethostbyname|gethostbyaddr|getgrnam|getgrgid|getgrent|getc|formline|format|fork|flock|fileno|fcntl|exp|exit|exists|exec|eval|eof|endservent|endpwent|endprotoent|endnetent|endhostent|endgrent|each|dump|do|die|delete|defined|dbmopen|dbmclose|crypt|cos|continue|connect|closedir|close|chroot|chr|chown|chop|chomp|chmod|chdir|caller|bless|binmode|bind|atan2|alarm|accept|abs)(?![a-zA-Z0-9_])",
|
|
),
|
|
PatternRule(
|
|
name='method',
|
|
pattern="&(?:[a-zA-Z_][a-zA-Z_0-9]*::)*[a-zA-Z_][a-zA-Z_0-9]*",
|
|
),
|
|
PatternRule(
|
|
name='ref_method',
|
|
pattern="&\$(?:[a-zA-Z_][a-zA-Z_0-9]*::)*[a-zA-Z_][a-zA-Z_0-9]*",
|
|
),
|
|
PatternRule(
|
|
name='bareword_method',
|
|
pattern="(?:[a-zA-Z_][a-zA-Z_0-9]*::)*[a-zA-Z_][a-zA-Z_0-9]* *\(",
|
|
),
|
|
PatternRule(
|
|
name='delimiter',
|
|
pattern="\(|\)|\[|\]|{|}|,|;|->|=>|=|\?|(?<!:):(?!=:)",
|
|
),
|
|
PatternRule(
|
|
name='unary_operator',
|
|
pattern="\+=|-=|\*=|/=|//=|%=|&=\|\^=|>>=|<<=|\*\*=",
|
|
),
|
|
PatternRule(
|
|
name='operator',
|
|
pattern="\+|<=>|<>|<<|<=|<|-|>>|>=|>|\*\*|&|\*|\||/|\^|==|//|~|=~|!~|!=|%|!|\.",
|
|
),
|
|
PatternRule(
|
|
name='bareword',
|
|
pattern='(?:[a-zA-Z_][a-zA-Z_0-9]*::)*[a-zA-Z_][a-zA-Z_0-9]*',
|
|
),
|
|
]
|
|
|
|
class Lexer:
|
|
def __init__(self, name, grammar):
|
|
self.name = name
|
|
self.grammar = grammar
|
|
self.y = 0
|
|
self.x = 0
|
|
self.lines = None
|
|
self.tokens = []
|
|
|
|
def add_token(self, t):
|
|
self.tokens.append(t)
|
|
|
|
def lex(self, lines, y=0, x=0):
|
|
self.y = y
|
|
self.x = x
|
|
self.lines = lines
|
|
self.tokens = []
|
|
|
|
def __iter__(self):
|
|
if self.lines is None:
|
|
raise Exception, "no lines to lex"
|
|
return self
|
|
|
|
def next(self):
|
|
null_t_name = 'null'
|
|
null_t = None
|
|
|
|
while self.y < len(self.lines):
|
|
line = self.lines[self.y]
|
|
while self.x < len(line):
|
|
for rule in self.grammar.rules:
|
|
if rule.match(self):
|
|
assert self.tokens, "AAAAA %s" % repr(self.tokens)
|
|
return self.tokens.pop(0)
|
|
if null_t is None:
|
|
null_t = Token(null_t_name, self.y, self.x, '')
|
|
self.add_token(null_t)
|
|
null_t.add_to_string(line[self.x])
|
|
self.x += 1
|
|
self.y += 1
|
|
self.x = 0
|
|
|
|
if self.tokens:
|
|
return self.tokens.pop(0)
|
|
else:
|
|
raise StopIteration
|