def perform_regex_tests(self, sublex):
testResult = 0
for rtKey in sorted(self.regexTests):
result = lexeme.check_for_regex(sublex, self.regexTests[rtKey][0],
self.errorHandler)
testResult = testResult * 2 + int(result)
return testResult
def apply(self, wf):
if wf.stem != self.stem and wf.lemma != self.lemma:
return None
for rxTest in self.searchFields:
if not lexeme.check_for_regex(wf, rxTest, errorHandler=self.errorHandler,
checkWordform=True):
return None
wfNew = copy.deepcopy(wf)
wfNew.otherData += self.addFields
return wfNew
def fork_regex(self, sublex):
"""
Write a regex-free version of self to the grammar if needed.
Return the name of the paradigm.
"""
if self.regexTests is None:
self.build_regex_tests()
if len(self.regexTests) == 0:
if self.name not in grammar.Grammar.paradigms:
grammar.Grammar.paradigms[self.name] = copy.deepcopy(self)
return self.name
testResult = self.perform_regex_tests(sublex)
newParaName = self.name + '=' + str(testResult)
if newParaName in grammar.Grammar.paradigms:
return newParaName
# If there is no such paradigm, make it:
newPara = copy.deepcopy(self)
testResult = 0
flex2remove = set()
for rtKey in sorted(newPara.regexTests):
result = lexeme.check_for_regex(sublex,
newPara.regexTests[rtKey][0],
self.errorHandler)
if not result:
flex2remove |= newPara.regexTests[rtKey][1]
testResult = testResult * 2 + int(result)
for iFlex in sorted(flex2remove, reverse=True):
newPara.flex.pop(iFlex)
newPara.name = newParaName
newPara.regexTests = {}
for flex in newPara.flex:
flex.regexTests = []
grammar.Grammar.paradigms[newParaName] = newPara
return newParaName