def test3():
from langscape.langlets.ls_grammar.grammar_object import GrammarObject
class NFA2GrammarObjTranslator(NFA2GrammarTranslator):
def __init__(self, langlet):
self.langlet = langlet
nfas = self.langlet.nfagenerator.nfas
self.names = {}
for r in nfas:
name = nfas[r][0].split(":")[0]
self.names[r] = name
super(NFA2GrammarObjTranslator, self).__init__(langlet)
def get_constants(self):
return {} #self.langlet.nfagenerator.nfadata.constants
def get_node_name(self, nid):
return self.names[nid]
rules = []
rules = open("tests/rulefile.txt").readlines()
for i, rule in enumerate(rules):
print "RULE", rule
go = GrammarObject.grammar_from_rule(rule)
nfa = go.nfagenerator.nfas[go.get_start_symbol()]
nfa2gra = NFA2GrammarObjTranslator(go)
if not nfa2gra.check_translation(nfa, False):
print (i, rule)
if i == 10:
break
def test3():
from langscape.langlets.ls_grammar.grammar_object import GrammarObject
class NFA2GrammarObjTranslator(NFA2GrammarTranslator):
def __init__(self, langlet):
self.langlet = langlet
nfas = self.langlet.nfagenerator.nfas
self.names = {}
for r in nfas:
name = nfas[r][0].split(":")[0]
self.names[r] = name
super(NFA2GrammarObjTranslator, self).__init__(langlet)
def get_constants(self):
return {} # self.langlet.nfagenerator.nfadata.constants
def get_node_name(self, nid):
return self.names[nid]
rules = []
rules = open("tests/rulefile.txt").readlines()
for i, rule in enumerate(rules):
print "RULE", rule
go = GrammarObject.grammar_from_rule(rule)
nfa = go.nfagenerator.nfas[go.get_start_symbol()]
nfa2gra = NFA2GrammarObjTranslator(go)
if not nfa2gra.check_translation(nfa, False):
print (i, rule)
if i == 10:
break
def check_refinement(self, parser_rules):
rules = []
templates = []
for name, R in parser_rules:
if isinstance(R, RuleTemplate):
rules.append(ls_grammar.tokenize(R.get_subgrammar_rule()))
templates.append(R)
else:
rules.append(ls_grammar.tokenize(R))
go = GrammarObject(rules)
go.set_langlet_id(self.langlet.langlet_id)
go.langlet.lex_nfa = self.langlet.lex_nfa
go.langlet.token = go.langlet.parse_token = self.langlet.token
go.create_grammar(expansion=False)
go.langlet._load_unparser()
nfas = go.get_nfas()
# check that grammars are specialized
for rt in templates:
nid = rt.get_nid()
nfa = nfas[rt.get_nid()]
traces = compute_span_traces(nfa)
for tr in traces:
tro = TraceObject(tr[:-1])
if not self.trchecker.checktrace(tro, start_symbol=nid):
raise GrammarError("Not a refinement: '%s'" % rt.rule_name)
return go
def check_refinement(self, parser_rules):
rules = []
templates = []
for name, R in parser_rules:
if isinstance(R, RuleTemplate):
rules.append(ls_grammar.tokenize(R.get_subgrammar_rule()))
templates.append(R)
else:
rules.append(ls_grammar.tokenize(R))
go = GrammarObject(rules)
go.set_langlet_id(self.langlet.langlet_id)
go.langlet.lex_nfa = self.langlet.lex_nfa
go.langlet.token = go.langlet.parse_token = self.langlet.token
go.create_grammar(expansion = False)
go.langlet._load_unparser()
nfas = go.get_nfas()
# check that grammars are specialized
for rt in templates:
nid = rt.get_nid()
nfa = nfas[rt.get_nid()]
traces = compute_span_traces(nfa)
for tr in traces:
tro = TraceObject(tr[:-1])
if not self.trchecker.checktrace(tro, start_symbol = nid):
raise GrammarError("Not a refinement: '%s'"%rt.rule_name)
return go
def check_translation(self, nfa, dbg=True):
global __DEBUG__
__DEBUG__ = dbg
if __DEBUG__:
print "----------------------------------------------"
pprint.pprint(nfa)
print "----------------------------------------------"
R = self.translate(nfa)
if __DEBUG__:
print "[check-translation : Rule]", R
go = GrammarObject.grammar_from_rule(R)
nfa_R = go.get_nfas()[go.get_start_symbol()]
__DEBUG__ = False
return R, nfa_comp(nfa, nfa_R)
def check_translation(self, nfa, dbg=True):
global __DEBUG__
__DEBUG__ = dbg
if __DEBUG__:
print "----------------------------------------------"
pprint.pprint(nfa)
print "----------------------------------------------"
R = self.translate(nfa)
if __DEBUG__:
print "[check-translation : Rule]", R
go = GrammarObject.grammar_from_rule(R)
nfa_R = go.get_nfas()[go.get_start_symbol()]
__DEBUG__ = False
return R, nfa_comp(nfa, nfa_R)
def create_refined_grammar(self):
parser_rules = self.get_parse_gen()
variables = {}
n = 0
for name in dir(self):
m = getattr(self, name)
if hasattr(m, "refined"):
doc = m.__doc__.strip()
if doc.startswith(name) and doc[len(name):len(name)+1] in " :":
rt = RuleTemplate(self.langlet, doc)
for i, (nm, _) in enumerate(parser_rules):
if name == nm:
parser_rules[i] = (name, rt)
#variables[names] = m.__call__() or []
go = self.check_refinement(parser_rules)
go.variables = variables
return go
if n == 0:
return go
else:
rules = []
for name, R in parser_rules:
if isinstance(R, RuleTemplate):
g_rule = R.get_grammar_rule()
rules.append(ls_grammar.tokenize(g_rule))
for varname in R.variables:
rules.append(ls_grammar.tokenize(varname + ": '$%s$'"%varname))
else:
rules.append(ls_grammar.tokenize(R))
go = GrammarObject(rules)
go.set_langlet_id(self.langlet.langlet_id)
go.langlet.lex_nfa = self.langlet.lex_nfa
go.langlet.token = go.langlet.parse_token = self.langlet.token
go.create_grammar(expansion = False)
go.langlet._load_unparser()
return go
def create_refined_grammar(self):
parser_rules = self.get_parse_gen()
variables = {}
n = 0
for name in dir(self):
m = getattr(self, name)
if hasattr(m, "refined"):
doc = m.__doc__.strip()
if doc.startswith(name) and doc[len(name):len(name) +
1] in " :":
rt = RuleTemplate(self.langlet, doc)
for i, (nm, _) in enumerate(parser_rules):
if name == nm:
parser_rules[i] = (name, rt)
#variables[names] = m.__call__() or []
go = self.check_refinement(parser_rules)
go.variables = variables
return go
if n == 0:
return go
else:
rules = []
for name, R in parser_rules:
if isinstance(R, RuleTemplate):
g_rule = R.get_grammar_rule()
rules.append(ls_grammar.tokenize(g_rule))
for varname in R.variables:
rules.append(
ls_grammar.tokenize(varname +
": '$%s$'" % varname))
else:
rules.append(ls_grammar.tokenize(R))
go = GrammarObject(rules)
go.set_langlet_id(self.langlet.langlet_id)
go.langlet.lex_nfa = self.langlet.lex_nfa
go.langlet.token = go.langlet.parse_token = self.langlet.token
go.create_grammar(expansion=False)
go.langlet._load_unparser()
return go