def parse_character_string(Txt_or_File, PatternStringF=False):
sh, sh_ref, start_position = __prepare_text_or_file_stream(Txt_or_File)
try:
# -- parse regular expression, build state machine
state_machine = snap_character_string.do(sh)
state_machine = regex.__clean_and_validate(state_machine, AllowNothingIsNecessaryF=False, fh=sh)
if state_machine is None:
error_msg("No valid regular character string expression detected.", sh_ref)
except RegularExpressionException, x:
error_msg("Regular expression parsing:\n" + x.message, sh_ref)
def parse_character_string(Txt_or_File, PatternStringF=False):
sh, sh_ref, start_position = __prepare_text_or_file_stream(Txt_or_File)
try:
# -- parse regular expression, build state machine
state_machine = snap_character_string.do(sh)
state_machine = regex.__clean_and_validate(
state_machine, AllowNothingIsNecessaryF=False, fh=sh)
if state_machine is None:
error_msg("No valid regular character string expression detected.",
sh_ref)
except RegularExpressionException, x:
error_msg("Regular expression parsing:\n" + x.message, sh_ref)
def snap_primary(stream, PatternDict):
"""primary: " non_double_quote * " = character string
[ non_rect_bracket_close ] = set of characters
{ identifier } = pattern replacement
( expression )
non_control_character+ = lonely characters
primary repetition_cmd
"""
__debug_entry("primary", stream)
x = stream.read(1)
lookahead = stream.read(1)
if x != "" and lookahead != "": stream.seek(-1, 1)
if x == "": return __debug_exit(None, stream)
# -- 'primary' primary
if x == "\"": result = snap_character_string.do(stream)
elif x == "[":
stream.seek(-1, 1)
result = character_set_expression.do(stream, PatternDict)
elif x == "{":
result = snap_replacement(stream, PatternDict)
elif x == ".":
result = create_ALL_BUT_NEWLINE_state_machine()
elif x == "(":
result = snap_bracketed_expression(stream, PatternDict)
elif x.isspace():
# a lonestanding space ends the regular expression
stream.seek(-1, 1)
return __debug_exit(None, stream)
elif x in ["*", "+", "?"]:
raise RegularExpressionException(
"lonely operator '%s' without expression proceeding." % x)
elif x == "\\":
if lookahead == "C":
stream.read(1)
result = snap_case_folded_pattern(stream, PatternDict)
elif lookahead == "R":
result = get_expression_in_brackets(stream, PatternDict,
"reverse operator",
"R").get_inverse()
elif lookahead == "A":
result = get_expression_in_brackets(stream, PatternDict,
"anti-pattern operator", "A")
result.transform_to_anti_pattern()
else:
stream.seek(-1, 1)
trigger_set = character_set_expression.snap_property_set(stream)
if trigger_set is None:
stream.seek(
1, 1) # snap_property_set() leaves tream right before '\\'
char_code = snap_backslashed_character.do(stream)
if char_code is None:
raise RegularExpressionException(
"Backslash followed by unrecognized character code.")
trigger_set = char_code
result = StateMachine()
result.add_transition(result.init_state_index,
trigger_set,
AcceptanceF=True)
elif x not in CONTROL_CHARACTERS:
# NOTE: The '\' is not inside the control characters---for a reason.
# It is used to define for example character codes using '\x' etc.
stream.seek(-1, 1)
result = snap_non_control_character(stream, PatternDict)
else:
# NOTE: This includes the '$' sign which means 'end of line'
# because the '$' sign is in CONTROL_CHARACTERS, but is not checked
# against. Thus, it it good to leave here on '$' because the
# '$' sign is handled on the very top level.
# this is not a valid primary
stream.seek(-1, 1)
return __debug_exit(None, stream)
# -- optional repetition command?
result_repeated = __snap_repetition_range(result, stream)
if result_repeated is not None: result = result_repeated
return __debug_exit(beautifier.do(result), stream)
def snap_primary(stream, PatternDict):
"""primary: " non_double_quote * " = character string
[ non_rect_bracket_close ] = set of characters
{ identifier } = pattern replacement
( expression )
non_control_character+ = lonely characters
primary repetition_cmd
"""
__debug_entry("primary", stream)
x = stream.read(1); lookahead = stream.read(1);
if x != "" and lookahead != "": stream.seek(-1, 1)
if x == "": return __debug_exit(None, stream)
# -- 'primary' primary
if x == "\"": result = snap_character_string.do(stream)
elif x == "[":
stream.seek(-1, 1);
result = character_set_expression.do(stream, PatternDict)
elif x == "{": result = snap_replacement(stream, PatternDict)
elif x == ".": result = create_ALL_BUT_NEWLINE_state_machine()
elif x == "(": result = snap_bracketed_expression(stream, PatternDict)
elif x.isspace():
# a lonestanding space ends the regular expression
stream.seek(-1, 1)
return __debug_exit(None, stream)
elif x in ["*", "+", "?"]:
raise RegularExpressionException("lonely operator '%s' without expression proceeding." % x)
elif x == "\\":
if lookahead == "C":
stream.read(1)
result = snap_case_folded_pattern(stream, PatternDict)
elif lookahead == "R":
result = get_expression_in_brackets(stream, PatternDict, "reverse operator", "R").get_inverse()
elif lookahead == "A":
result = get_expression_in_brackets(stream, PatternDict, "anti-pattern operator", "A")
result.transform_to_anti_pattern()
else:
stream.seek(-1, 1)
trigger_set = character_set_expression.snap_property_set(stream)
if trigger_set is None:
stream.seek(1, 1) # snap_property_set() leaves tream right before '\\'
char_code = snap_backslashed_character.do(stream)
if char_code is None:
raise RegularExpressionException("Backslash followed by unrecognized character code.")
trigger_set = char_code
result = StateMachine()
result.add_transition(result.init_state_index, trigger_set, AcceptanceF=True)
elif x not in CONTROL_CHARACTERS:
# NOTE: The '\' is not inside the control characters---for a reason.
# It is used to define for example character codes using '\x' etc.
stream.seek(-1, 1)
result = snap_non_control_character(stream, PatternDict)
else:
# NOTE: This includes the '$' sign which means 'end of line'
# because the '$' sign is in CONTROL_CHARACTERS, but is not checked
# against. Thus, it it good to leave here on '$' because the
# '$' sign is handled on the very top level.
# this is not a valid primary
stream.seek(-1, 1)
return __debug_exit(None, stream)
# -- optional repetition command?
result_repeated = __snap_repetition_range(result, stream)
if result_repeated is not None: result = result_repeated
return __debug_exit(beautifier.do(result), stream)
def snap_primary(stream, PatternDict):
"""primary: " non_double_quote * " = character string
[ non_rect_bracket_close ] = set of characters
{ identifier } = pattern replacement
( expression )
non_control_character+ = lonely characters
primary repetition_cmd
"""
global SPECIAL_TERMINATOR
__debug_entry("primary", stream)
x = stream.read(1)
if x == "": return __debug_exit(None, stream)
# -- 'primary' primary
if x == "\"": result = snap_character_string.do(stream)
elif x == "[":
stream.seek(-1, 1);
result = snap_character_set_expression(stream, PatternDict)
elif x == "{": result = snap_replacement(stream, PatternDict)
elif x == ".": result = create_ALL_BUT_NEWLINE_state_machine(stream)
elif x == "(": result = snap_bracketed_expression(stream, PatternDict)
elif x.isspace():
# a lonestanding space ends the regular expression
stream.seek(-1, 1)
return __debug_exit(None, stream)
elif x in ["*", "+", "?"]:
raise RegularExpressionException("lonely operator '%s' without expression proceeding." % x)
elif x == "\\":
result = snap_command(stream, PatternDict)
if result is None:
stream.seek(-1, 1)
trigger_set = snap_property_set(stream)
if trigger_set is None:
# snap the '\'
stream.read(1)
char_code = snap_backslashed_character.do(stream)
if char_code is None:
raise RegularExpressionException("Backslash followed by unrecognized character code.")
trigger_set = char_code
result = StateMachine()
result.add_transition(result.init_state_index, trigger_set, AcceptanceF=True)
elif x not in CONTROL_CHARACTERS and x != SPECIAL_TERMINATOR:
# NOTE: The '\' is not inside the control characters---for a reason.
# It is used to define for example character codes using '\x' etc.
stream.seek(-1, 1)
result = snap_non_control_character(stream, PatternDict)
else:
# NOTE: This includes the '$' sign which means 'end of line'
# because the '$' sign is in CONTROL_CHARACTERS, but is not checked
# against. Thus, it it good to leave here on '$' because the
# '$' sign is handled on the very top level.
# this is not a valid primary
stream.seek(-1, 1)
return __debug_exit(None, stream)
# -- optional repetition command?
result_repeated = __snap_repetition_range(result, stream)
if result_repeated is not None: result = result_repeated
# There's something going wrong with pseudo-ambigous post context
# if we do not clean-up here. TODO: Investigate why?
# See tests in generator/TEST directory.
return __debug_exit(beautifier.do(result), stream)
def test(TestString):
print "expression = \"" + TestString + "\""
print "state machine\n", snap_character_string.do(
StringIO(TestString + '"'))
def test(TestString):
print "expression = \"" + TestString + "\""
print "state machine\n", snap_character_string.do(StringIO(TestString + '"'))
