def __unichr (self, code_point):
rv = six.unichr(code_point)
if 0 == code_point:
rv = six.u('x00')
if code_point in self.__XMLtoPythonREEscapedCodepoints:
rv = six.unichr(0x5c) + rv
return rv
def __unichr(self, code_point):
rv = six.unichr(code_point)
if 0 == code_point:
rv = six.u('x00')
if code_point in self.__XMLtoPythonREEscapedCodepoints:
rv = six.unichr(0x5c) + rv
return rv
def asSingleCharacter(self):
"""If this set represents a single character, return it as its
unicode string value. Otherwise return C{None}."""
if (2 != len(self.__codepoints)) or (
1 < (self.__codepoints[1] - self.__codepoints[0])):
return None
return six.unichr(self.__codepoints[0])
def XMLToPython(pattern):
"""Convert the given pattern to the format required for Python
regular expressions.
@param pattern: A Unicode string defining a pattern consistent
with U{XML regular
expressions<http://www.w3.org/TR/xmlschema-2/index.html#regexs>}.
@return: A Unicode string specifying a Python regular expression
that matches the same language as C{pattern}."""
assert isinstance(pattern, six.text_type)
new_pattern_elts = []
new_pattern_elts.append('^(')
position = 0
while position < len(pattern):
cg = MaybeMatchCharacterClass(pattern, position)
if cg is None:
ch = pattern[position]
if ch == six.u('^') or ch == six.u('$'):
# These characters have no special meaning in XSD. But they
# match start and end of string in Python, so they have to
# be escaped.
new_pattern_elts.append(six.unichr(0x5c) + ch)
else:
new_pattern_elts.append(ch)
position += 1
else:
(cps, position) = cg
new_pattern_elts.append(cps.asPattern())
new_pattern_elts.append(')$')
return ''.join(new_pattern_elts)
def XMLToPython (pattern):
"""Convert the given pattern to the format required for Python
regular expressions.
@param pattern: A Unicode string defining a pattern consistent
with U{XML regular
expressions<http://www.w3.org/TR/xmlschema-2/index.html#regexs>}.
@return: A Unicode string specifying a Python regular expression
that matches the same language as C{pattern}."""
assert isinstance(pattern, six.text_type)
new_pattern_elts = []
new_pattern_elts.append('^(')
position = 0
while position < len(pattern):
cg = MaybeMatchCharacterClass(pattern, position)
if cg is None:
ch = pattern[position]
if ch == six.u('^') or ch == six.u('$'):
# These characters have no special meaning in XSD. But they
# match start and end of string in Python, so they have to
# be escaped.
new_pattern_elts.append(six.unichr(0x5c) + ch)
else:
new_pattern_elts.append(ch)
position += 1
else:
(cps, position) = cg
new_pattern_elts.append(cps.asPattern())
new_pattern_elts.append(')$')
return ''.join(new_pattern_elts)
def _InitializeAllEsc():
"""Set the values in _AllEsc without introducing C{k} and C{v} into
the module."""
_AllEsc.update({six.u('.'): pyxb.utils.unicode.WildcardEsc})
bs = six.unichr(0x5c)
for k, v in six.iteritems(pyxb.utils.unicode.SingleCharEsc):
_AllEsc[bs + six.text_type(k)] = v
for k, v in six.iteritems(pyxb.utils.unicode.MultiCharEsc):
_AllEsc[bs + six.text_type(k)] = v
for k, v in six.iteritems(pyxb.utils.unicode.catEsc):
_AllEsc[bs + six.text_type(k)] = v
for k, v in six.iteritems(pyxb.utils.unicode.complEsc):
_AllEsc[bs + six.text_type(k)] = v
for k, v in six.iteritems(pyxb.utils.unicode.IsBlockEsc):
_AllEsc[bs + six.text_type(k)] = v
def _InitializeAllEsc ():
"""Set the values in _AllEsc without introducing C{k} and C{v} into
the module."""
_AllEsc.update({ six.u('.'): pyxb.utils.unicode.WildcardEsc })
bs = six.unichr(0x5c)
for k, v in six.iteritems(pyxb.utils.unicode.SingleCharEsc):
_AllEsc[bs + six.text_type(k)] = v
for k, v in six.iteritems(pyxb.utils.unicode.MultiCharEsc):
_AllEsc[bs + six.text_type(k)] = v
for k, v in six.iteritems(pyxb.utils.unicode.catEsc):
_AllEsc[bs + six.text_type(k)] = v
for k, v in six.iteritems(pyxb.utils.unicode.complEsc):
_AllEsc[bs + six.text_type(k)] = v
for k, v in six.iteritems(pyxb.utils.unicode.IsBlockEsc):
_AllEsc[bs + six.text_type(k)] = v
def _MatchPosCharGroup(text, position):
'''Parse a U{posCharGroup<http://www.w3.org/TR/xmlschema-2/#nt-posCharGroup>} term.
@return: A tuple C{(cps, fs, p)} where:
- C{cps} is a L{pyxb.utils.unicode.CodePointSet} containing the code points associated with the group;
- C{fs} is a C{bool} that is C{True} if the next character is the C{-} in a U{charClassSub<http://www.w3.org/TR/xmlschema-2/#nt-charClassSub>} and C{False} if the group is not part of a charClassSub;
- C{p} is the text offset immediately following the closing brace.
@raise RegularExpressionError: if the expression is syntactically
invalid.
'''
start_position = position
# DASH is just some unique object, used as a marker.
# It can't be unicode or a CodePointSet.
class DashClass:
pass
DASH = DashClass()
# We tokenize first, then go back and stick the ranges together.
tokens = []
has_following_subtraction = False
while True:
if position >= len(text):
raise RegularExpressionError(
position,
"Incomplete character class expression, missing closing ']'")
ch = text[position]
if ch == six.u('['):
# Only allowed if this is a subtraction
if not tokens or tokens[-1] is not DASH:
raise RegularExpressionError(
position, "'[' character not allowed in character class")
has_following_subtraction = True
# For a character class subtraction, the "-[" are not part of the
# posCharGroup, so undo reading the dash
tokens.pop()
position = position - 1
break
elif ch == six.u(']'):
# End
break
elif ch == six.unichr(0x5c): # backslash
cps, position = _MatchCharClassEsc(text, position)
single_char = cps.asSingleCharacter()
if single_char is not None:
tokens.append(single_char)
else:
tokens.append(cps)
elif ch == six.u('-'):
# We need to distinguish between "-" and "\-". So we use
# DASH for a plain "-", and u"-" for a "\-".
tokens.append(DASH)
position = position + 1
else:
tokens.append(ch)
position = position + 1
if not tokens:
raise RegularExpressionError(position,
"Empty character class not allowed")
# At the start or end of the character group, a dash has to be a literal
if tokens[0] is DASH:
tokens[0] = six.u('-')
if tokens[-1] is DASH:
tokens[-1] = six.u('-')
result_cps = pyxb.utils.unicode.CodePointSet()
cur_token = 0
while cur_token < len(tokens):
start = tokens[cur_token]
if cur_token + 2 < len(tokens) and tokens[cur_token + 1] is DASH:
end = tokens[cur_token + 2]
if not isinstance(start, six.text_type) or not isinstance(
end, six.text_type):
if start is DASH or end is DASH:
raise RegularExpressionError(
start_position,
'Two dashes in a row is not allowed in the middle of a character class.'
)
raise RegularExpressionError(
start_position,
'Dashes must be surrounded by characters, not character class escapes. %r %r'
% (start, end))
if start > end:
raise RegularExpressionError(
start_position,
'Character ranges must have the lowest character first')
result_cps.add((ord(start), ord(end)))
cur_token = cur_token + 3
else:
if start is DASH:
raise RegularExpressionError(
start_position, 'Dash without an initial character')
elif isinstance(start, six.text_type):
result_cps.add(ord(start))
else:
result_cps.extend(start)
cur_token = cur_token + 1
return result_cps, has_following_subtraction, position
def _MatchPosCharGroup(text, position):
'''Parse a U{posCharGroup<http://www.w3.org/TR/xmlschema-2/#nt-posCharGroup>} term.
@return: A tuple C{(cps, fs, p)} where:
- C{cps} is a L{pyxb.utils.unicode.CodePointSet} containing the code points associated with the group;
- C{fs} is a C{bool} that is C{True} if the next character is the C{-} in a U{charClassSub<http://www.w3.org/TR/xmlschema-2/#nt-charClassSub>} and C{False} if the group is not part of a charClassSub;
- C{p} is the text offset immediately following the closing brace.
@raise RegularExpressionError: if the expression is syntactically
invalid.
'''
start_position = position
# DASH is just some unique object, used as a marker.
# It can't be unicode or a CodePointSet.
class DashClass:
pass
DASH = DashClass()
# We tokenize first, then go back and stick the ranges together.
tokens = []
has_following_subtraction = False
while True:
if position >= len(text):
raise RegularExpressionError(position, "Incomplete character class expression, missing closing ']'")
ch = text[position]
if ch == six.u('['):
# Only allowed if this is a subtraction
if not tokens or tokens[-1] is not DASH:
raise RegularExpressionError(position, "'[' character not allowed in character class")
has_following_subtraction = True
# For a character class subtraction, the "-[" are not part of the
# posCharGroup, so undo reading the dash
tokens.pop()
position = position - 1
break
elif ch == six.u(']'):
# End
break
elif ch == six.unichr(0x5c): # backslash
cps, position = _MatchCharClassEsc(text, position)
single_char = cps.asSingleCharacter()
if single_char is not None:
tokens.append(single_char)
else:
tokens.append(cps)
elif ch == six.u('-'):
# We need to distinguish between "-" and "\-". So we use
# DASH for a plain "-", and u"-" for a "\-".
tokens.append(DASH)
position = position + 1
else:
tokens.append(ch)
position = position + 1
if not tokens:
raise RegularExpressionError(position, "Empty character class not allowed")
# At the start or end of the character group, a dash has to be a literal
if tokens[0] is DASH:
tokens[0] = six.u('-')
if tokens[-1] is DASH:
tokens[-1] = six.u('-')
result_cps = pyxb.utils.unicode.CodePointSet()
cur_token = 0
while cur_token < len(tokens):
start = tokens[cur_token]
if cur_token + 2 < len(tokens) and tokens[cur_token + 1] is DASH:
end = tokens[cur_token + 2]
if not isinstance(start, six.text_type) or not isinstance(end, six.text_type):
if start is DASH or end is DASH:
raise RegularExpressionError(start_position, 'Two dashes in a row is not allowed in the middle of a character class.')
raise RegularExpressionError(start_position, 'Dashes must be surrounded by characters, not character class escapes. %r %r' %(start, end))
if start > end:
raise RegularExpressionError(start_position, 'Character ranges must have the lowest character first')
result_cps.add((ord(start), ord(end)))
cur_token = cur_token + 3
else:
if start is DASH:
raise RegularExpressionError(start_position, 'Dash without an initial character')
elif isinstance(start, six.text_type):
result_cps.add(ord(start))
else:
result_cps.extend(start)
cur_token = cur_token + 1
return result_cps, has_following_subtraction, position
def asSingleCharacter (self):
"""If this set represents a single character, return it as its
unicode string value. Otherwise return C{None}."""
if (2 != len(self.__codepoints)) or (1 < (self.__codepoints[1] - self.__codepoints[0])):
return None
return six.unichr(self.__codepoints[0])
