def test__discover_closure_states(self):
"""_discover_closure_states(NFA)"""
act = nfa_data()
act.states[0] = b_State(0,set())
act.states[1] = b_State(1,set())
act.states[2] = b_State(2,set())
act.states[3] = b_State(3,set([0]))
act.states[4] = b_State(4,set())
act.states[5] = b_State(5,set())
act.states[6] = b_State(6,set([1]))
act.alphabet[0] = b_Sym_char("a", "a", 0)
act.alphabet[1] = b_Sym_char("b", "b", 1)
act.alphabet[2] = b_Sym_char("c", "c", 2)
act.alphabet[3] = b_Sym_char("d", "d", 3)
act.alphabet[4] = b_Sym_char("e", "e", 4)
act.alphabet[5] = b_Sym_char("f", "f", 5)
star = set()
for ord_char in range(0, 256):
star.add(chr(ord_char))
act.alphabet[6] = b_Sym_char_class("*", star, 6)
mimo_a = set()
for ord_char in range(0, 256):
mimo_a.add(chr(ord_char))
mimo_a.remove('a')
act.alphabet[7] = b_Sym_char_class("^a", mimo_a, 7)
act.start = 0
act.final.add(3)
act.final.add(6)
act.transitions.add( (0, 6, 0) )
act.transitions.add( (0, 0, 1) )
act.transitions.add( (1, 1, 2) )
act.transitions.add( (2, 7, 2) )
act.transitions.add( (2, 2, 3) )
act.transitions.add( (0, 3, 4) )
act.transitions.add( (4, 4, 5) )
act.transitions.add( (5, 5, 6) )
history = HistoryFA()
history._automaton = act
history.remove_epsilons()
NFA = history.get_automaton(True)
self.assertTrue(history._discover_closure_states(NFA) == [2])
def get_hfa(ruleset):
"""
Generate number of states, transitions and consumed memory for \
History FA.
"""
# Create parser - use default parser
po = parser.parser()
# Parse input file
po.load_file(ruleset)
# Create History FA object
history = HistoryFA()
# Make automaton from RE which was in input file
history.create_by_parser(po)
# Remove epsilons
history.remove_epsilons()
# Store the NFA
NFA = history.get_automaton(True)
# Determinise the automaton
history.determinise(create_table = True)
# Create History FA
history.compute(NFA)
# Return experimental results
return ["History FA", history.get_state_num(), history.get_trans_num(), history.report_memory_optimal(), history.report_memory_naive()]
def get_hfa(ruleset):
"""
Generate number of states, transitions and consumed memory for \
History FA.
"""
# Create parser - use default parser
po = parser.parser()
# Parse input file
po.load_file(ruleset)
# Create History FA object
history = HistoryFA()
# Make automaton from RE which was in input file
history.create_by_parser(po)
# Remove epsilons
history.remove_epsilons()
# Store the NFA
NFA = history.get_automaton(True)
# Determinise the automaton
history.determinise(create_table=True)
# Create History FA
history.compute(NFA)
# Return experimental results
return [
"History FA",
history.get_state_num(),
history.get_trans_num(),
history.report_memory_optimal(),
history.report_memory_naive()
]
def test__identify_fading_states(self):
"""_identify_fading_states(nfa_closure_states)"""
history = HistoryFA()
history._state_representation = [ set([0]),
set([0,1]),
set([0,2]),
set([0,3]),
set([0,4]),
set([0,5]),
set([0,6]),
set([0,2,4]),
set([0,2,5]),
set([0,2,6])
]
self.assertTrue(history._identify_fading_states([2]) == [2, 7, 8, 9])
act = nfa_data()
act.states[0] = b_State(0,set())
act.states[1] = b_State(1,set())
act.states[2] = b_State(2,set())
act.states[3] = b_State(3,set([0]))
act.states[4] = b_State(4,set())
act.states[5] = b_State(5,set())
act.states[6] = b_State(6,set([1]))
act.alphabet[0] = b_Sym_char("a", "a", 0)
act.alphabet[1] = b_Sym_char("b", "b", 1)
act.alphabet[2] = b_Sym_char("c", "c", 2)
act.alphabet[3] = b_Sym_char("d", "d", 3)
act.alphabet[4] = b_Sym_char("e", "e", 4)
act.alphabet[5] = b_Sym_char("f", "f", 5)
star = set()
for ord_char in range(0, 256):
star.add(chr(ord_char))
act.alphabet[6] = b_Sym_char_class("*", star, 6)
mimo_a = set()
for ord_char in range(0, 256):
mimo_a.add(chr(ord_char))
mimo_a.remove('a')
act.alphabet[7] = b_Sym_char_class("^a", mimo_a, 7)
act.start = 0
act.final.add(3)
act.final.add(6)
act.transitions.add( (0, 6, 0) )
act.transitions.add( (0, 0, 1) )
act.transitions.add( (1, 1, 2) )
act.transitions.add( (2, 7, 2) )
act.transitions.add( (2, 2, 3) )
act.transitions.add( (0, 3, 4) )
act.transitions.add( (4, 4, 5) )
act.transitions.add( (5, 5, 6) )
history = HistoryFA()
history._automaton = act
history.remove_epsilons()
NFA = history.get_automaton(True)
history.determinise(create_table = True)
nfa_closure_states = history._discover_closure_states(NFA)
self.assertTrue(history._identify_fading_states(nfa_closure_states) ==
[5, 7, 8, 9])
def test_report_memory_naive(self):
"""report_memory_naive()"""
# /ab[^1234]*cd|efg/; test with an expression containing one
# alternation [^1234]*, the second is not
par = parser("pcre_parser")
par.set_text("/ab[^1234]*cd|efg/")
history = HistoryFA()
history.create_by_parser(par)
history.remove_epsilons()
NFA = history.get_automaton(True)
history.determinise(create_table = True)
history.compute(NFA)
self.assertTrue(history.report_memory_naive() == 187)
def test_search(self):
"""search()"""
# For RE upstairs do tests for this method.
# /abcd/; test with an expression that does not use properties
# of HistoryFA
par = parser("pcre_parser")
par.set_text("/abcd/")
history = HistoryFA()
history.create_by_parser(par)
history.remove_epsilons()
NFA = history.get_automaton(True)
history.determinise(create_table = True)
history.compute(NFA)
self.assertTrue(history.search("efg") == [0])
self.assertTrue(history.search("efabcg") == [0])
self.assertTrue(history.search("efabcd") == [1])
self.assertTrue(history.search("abcefabcdabcx") == [1])
self.assertTrue(history.search("abcd") == [1])
self.assertTrue(history.search("123abcd456") == [1])
self.assertTrue(history.search("123ab0cd456") == [0])
# /ab.*cd /; test with an expression that contains .*
par = parser("pcre_parser")
par.set_text("/ab.*cd/")
history = HistoryFA()
history.create_by_parser(par)
history.remove_epsilons()
NFA = history.get_automaton(True)
history.determinise(create_table = True)
history.compute(NFA)
self.assertTrue(history.search("efg") == [0])
self.assertTrue(history.search("abnecoefg") == [0])
self.assertTrue(history.search("abnecocx") == [0])
self.assertTrue(history.search("necoabnecocdneco") == [1])
self.assertTrue(history.search("abcd") == [1])
self.assertTrue(history.search("123abcd456") == [1])
self.assertTrue(history.search("123ab0cd456") == [1])
# /ab[^1234]*cd|efg/; test with an expression containing one
# alternation [^1234]*, the second is not
par = parser("pcre_parser")
par.set_text("/ab[^1234]*cd|efg/")
history = HistoryFA()
history.create_by_parser(par)
history.remove_epsilons()
NFA = history.get_automaton(True)
history.determinise(create_table = True)
history.compute(NFA)
self.assertTrue(history.search("exfg") == [0])
self.assertTrue(history.search("abnecocx") == [0])
self.assertTrue(history.search("efg") == [1])
self.assertTrue(history.search("textefgtext") == [1])
self.assertTrue(history.search("abnecoefg") == [1])
self.assertTrue(history.search("necoabnecocdneco") == [1])
self.assertTrue(history.search("abcd") == [1])
self.assertTrue(history.search("123abcd456") == [1])
self.assertTrue(history.search("123ab0cd456") == [1])
self.assertTrue(history.search("ab1cd") == [0])
self.assertTrue(history.search("ab4cd") == [0])
self.assertTrue(history.search("abcefg") == [1])
self.assertTrue(history.search("123abblekekcd456") == [1])
self.assertTrue(history.search("123ab0cd456") == [1])
def test_compute(self):
"""compute()"""
# Check the correctness of the logical machine output over
# self.assertTrue on individual items + focus on the properties
# of HistoryFA (transitions, flags, counters)
# /abcd/; test with an expression that does not use properties
# of HistoryFA
par = parser("pcre_parser")
par.set_text("/abcd/")
history = HistoryFA()
history.create_by_parser(par)
history.remove_epsilons()
NFA = history.get_automaton(True)
history.determinise(create_table = True)
history.compute(NFA)
copy = history.get_automaton()
result = nfa_data().load_from_file(aux_func.getPatternMatchDir() + "/algorithms/history_fa/test_data/test_data_1.nfa_data")
self.assertTrue(sorted(copy.states.keys()) ==
sorted(result.states.keys()))
self.assertTrue(copy.alphabet == result.alphabet)
self.assertTrue(copy.start == result.start)
self.assertTrue(copy.final == result.final)
self.assertTrue(copy.transitions == result.transitions)
# /ab.*cd /; test with an expression that contains .*
par = parser("pcre_parser")
par.set_text("/ab.*cd/")
history = HistoryFA()
history.create_by_parser(par)
history.remove_epsilons()
NFA = history.get_automaton(True)
history.determinise(create_table = True)
history.compute(NFA)
copy = history.get_automaton()
result = nfa_data().load_from_file(aux_func.getPatternMatchDir() + "/algorithms/history_fa/test_data/test_data_2.nfa_data")
self.assertTrue(sorted(copy.states.keys()) ==
sorted(result.states.keys()))
self.assertTrue(copy.alphabet == result.alphabet)
self.assertTrue(copy.start == result.start)
self.assertTrue(copy.final == result.final)
self.assertTrue(copy.transitions == result.transitions)
self.assertTrue(copy.Flags == result.Flags)
# /ab[^1234]*cd|efg/; test with an expression containing one
# alternation [^1234]*, the second is not
par = parser("pcre_parser")
par.set_text("/ab[^1234]*cd|efg/")
history = HistoryFA()
history.create_by_parser(par)
history.remove_epsilons()
NFA = history.get_automaton(True)
history.determinise(create_table = True)
history.compute(NFA)
copy = history.get_automaton()
result = nfa_data().load_from_file(aux_func.getPatternMatchDir() + "/algorithms/history_fa/test_data/test_data_3.nfa_data")
self.assertTrue(sorted(copy.states.keys()) ==
sorted(result.states.keys()))
self.assertTrue(copy.alphabet == result.alphabet)
self.assertTrue(copy.start == result.start)
self.assertTrue(copy.final == result.final)
self.assertTrue(copy.transitions == result.transitions)
self.assertTrue(copy.Flags == result.Flags)