def create_ALL_BUT_NEWLINE_state_machine():
global Setup
result = StateMachine()
# NOTE: Buffer control characters are supposed to be filtered out by the code
# generator.
trigger_set = NumberSet(Interval(ord("\n")).inverse())
if Setup.get_character_value_limit() != sys.maxint:
trigger_set.intersect_with(Interval(0, Setup.get_character_value_limit()))
result.add_transition(result.init_state_index, trigger_set, AcceptanceF=True)
return result
def create_ALL_BUT_NEWLINE_state_machine():
global Setup
result = StateMachine()
# NOTE: Buffer control characters are supposed to be filtered out by the code
# generator.
trigger_set = NumberSet(Interval(ord("\n")).inverse())
if Setup.get_character_value_limit() != sys.maxint:
trigger_set.intersect_with(
Interval(0, Setup.get_character_value_limit()))
result.add_transition(result.init_state_index,
trigger_set,
AcceptanceF=True)
return result
def prune_range(TriggerMap):
"""Consider the 'useful range' starting from zero. Thus, the first
interval to be considered is the first that intersects with 0.
Then 'begin' must become '0' instead of a negative value.
"""
LowerLimit = 0
UpperLimit = Setup.get_character_value_limit()
# Iterate from 'low' to 'high'
for i, info in enumerate(TriggerMap):
interval, target = info
if interval.end <= LowerLimit: continue
# Found an interval that intersects with 'LowerLimit' line
interval.begin = LowerLimit
if i != 0: del TriggerMap[:i]
return
L = len(TriggerMap)
# Iterate from 'high' to 'low'
for ii, info in enumerate(reversed(TriggerMap)):
interval, target = info
if interval.begin > UpperLimit: continue
# Found an interval that intersects with 'UpperLimit' line
interval.end = UpperLimit
if ii != 0:
i = L - ii - 1 # index = Size - inverse index - 1
del TriggerMap[i+1:]
return
# The whole trigger map happens below 0. This is trash, no doubt!
assert False
def prune_range(TriggerMap):
"""Consider the 'useful range' starting from zero. Thus, the first
interval to be considered is the first that intersects with 0.
Then 'begin' must become '0' instead of a negative value.
"""
LowerLimit = 0
UpperLimit = Setup.get_character_value_limit()
# Iterate from 'low' to 'high'
for i, info in enumerate(TriggerMap):
interval, target = info
if interval.end <= LowerLimit: continue
# Found an interval that intersects with 'LowerLimit' line
interval.begin = LowerLimit
if i != 0: del TriggerMap[:i]
return
L = len(TriggerMap)
# Iterate from 'high' to 'low'
for ii, info in enumerate(reversed(TriggerMap)):
interval, target = info
if interval.begin > UpperLimit: continue
# Found an interval that intersects with 'UpperLimit' line
interval.end = UpperLimit
if ii != 0:
i = L - ii - 1 # index = Size - inverse index - 1
del TriggerMap[i+1:]
return
# The whole trigger map happens below 0. This is trash, no doubt!
assert False
def __delete_forbidden_ranges(sm, fh):
"""Unicode does define all code points >= 0. Thus there can be no code points
below zero as it might result from some number set operations.
NOTE: This operation might result in orphaned states that have to
be deleted.
"""
global Setup
character_value_limit = Setup.get_character_value_limit()
for state in sm.states.values():
for target_state_index, trigger_set in state.transitions().get_map().items():
# Make sure, all transitions lie inside the unicode code range
if trigger_set.minimum() < UnicodeInterval.begin or trigger_set.supremum() >= UnicodeInterval.end:
trigger_set.intersect_with(UnicodeInterval)
if trigger_set.supremum() > character_value_limit:
error_msg("Pattern contains character beyond the scope of the buffer element size (%s)\n" \
% Setup.get_character_value_limit_str() + \
"Please, cut the character range of the regular expression,\n"
"adapt \"--buffer-element-size\" or \"--buffer-element-type\",\n" + \
"or specify '--buffer-element-size-irrelevant' to ignore the issue.", fh)
if Setup.buffer_codec in ["utf16-le", "utf16-be"]:
# Delete the forbidden interval: D800-DFFF
if trigger_set.has_intersection(ForbiddenRange):
error_msg("Pattern contains characters in unicode range 0xD800-0xDFFF.\n"
"This range is not covered by UTF16. Cutting Interval.", fh, DontExitF=True)
trigger_set.cut_interval(ForbiddenRange)
# If the operation resulted in cutting the path to the target state, then delete it.
if trigger_set.is_empty():
state.transitions().delete_transitions_to_target(target_state_index)
def __delete_forbidden_ranges(sm, fh):
"""Unicode does define all code points >= 0. Thus there can be no code points
below zero as it might result from some number set operations.
NOTE: This operation might result in orphaned states that have to
be deleted.
"""
global Setup
character_value_limit = Setup.get_character_value_limit()
for state in sm.states.values():
for target_state_index, trigger_set in state.transitions().get_map(
).items():
# Make sure, all transitions lie inside the unicode code range
if trigger_set.minimum(
) < UnicodeInterval.begin or trigger_set.supremum(
) >= UnicodeInterval.end:
trigger_set.intersect_with(UnicodeInterval)
if trigger_set.supremum() > character_value_limit:
error_msg("Pattern contains character beyond the scope of the buffer element size (%s)\n" \
% Setup.get_character_value_limit_str() + \
"Please, cut the character range of the regular expression,\n"
"adapt \"--buffer-element-size\" or \"--buffer-element-type\",\n" + \
"or specify '--buffer-element-size-irrelevant' to ignore the issue.", fh)
if Setup.buffer_codec in ["utf16-le", "utf16-be"]:
# Delete the forbidden interval: D800-DFFF
if trigger_set.has_intersection(ForbiddenRange):
error_msg(
"Pattern contains characters in unicode range 0xD800-0xDFFF.\n"
"This range is not covered by UTF16. Cutting Interval.",
fh,
DontExitF=True)
trigger_set.cut_interval(ForbiddenRange)
# If the operation resulted in cutting the path to the target state, then delete it.
if trigger_set.is_empty():
state.transitions().delete_transitions_to_target(
target_state_index)
