def i_hyphenate_derived(self,aWord):
"""
You can use this method in classes derived from ExplicitHyphenator.
It will first split the word using BaseHyphenator,
then for each "subword" it will call ExplicitHyphenator,
and only call the derived classes hyph method for the still
unknown subwords.
TODO: The implementation does not match the docstring
test: "hohenlimburg.de", "hohenlimburg.de)"
"""
#print "ExplicitHyphenator.i_hyphenate_derived", aWord
assert isinstance(aWord, unicode_type)
# Helper function
sub_hwords = []
hword = BaseHyphenator.i_hyphenate(self,aWord)
#print "BaseHyphenator.i_hyphenate returned %r" % hword
if hword is None:
hword = HyphenatedWord(aWord,hyphenations=[])
base_hyph_points = hword.hyphenations
last_indx = 0
nr = 0
for hpnum, hp in enumerate(base_hyph_points):
if isinstance(hp, int):
hp = HyphenationPoint(hp, quality=5, sl=SHY)
subword = hword[last_indx+nr:hp.indx]
# handle subword
if SHY in subword:
sub_hword = self.stripper.apply_stripped(BaseHyphenator.hyph, self, subword)
else:
sub_hword = self.stripper.apply_stripped(ExplicitHyphenator.hyph, self, subword)
if sub_hword is None:
sub_hword = self.stripper.apply_stripped(self.hyph, self, subword)
if sub_hword is None:
sub_hword = HyphenatedWord(subword, hyphenations=[])
sub_hwords.append(sub_hword)
# end handle subword
last_indx = hp.indx
nr = hp.nr
# Now the last subword
subword = hword[last_indx:]
# handle subword
if SHY in subword:
sub_hword = self.stripper.apply_stripped(BaseHyphenator.hyph, self, subword)
else:
sub_hword = self.stripper.apply_stripped(ExplicitHyphenator.hyph, self, subword)
if sub_hword is None:
sub_hword = self.stripper.apply_stripped(self.hyph, self, subword)
if sub_hword is None:
sub_hword = HyphenatedWord(subword, hyphenations=[])
sub_hwords.append(sub_hword)
#end handle subword
if len(sub_hwords) > 1:
return HyphenatedWord.join(sub_hwords)
else:
return sub_hwords[0] # Kann auch None sein.
def hyph(self, aWord):
assert isinstance(aWord, unicode_type)
hword = HyphenatedWord(aWord, hyphenations=self.zerlegeWort(aWord))
# None (unknown) kann hier nicht vorkommen, da der
# Algorithmus musterbasiert funktioniert und die Wörter
# sowieso nicht "kennt" oder "nicht kennt".
return hword
def i_hyphenate_derived(self,aWord):
"""
You can use this method in classes derived from ExplicitHyphenator.
It will first split the word using BaseHyphenator,
then for each "subword" it will call ExplicitHyphenator,
and only call the derived classes hyph method for the still
unknown subwords.
TODO: The implementation does not match the docstring
test: "hohenlimburg.de", "hohenlimburg.de)"
"""
#print "ExplicitHyphenator.i_hyphenate_derived", aWord
assert isinstance(aWord, unicode)
# Helper function
sub_hwords = []
hword = BaseHyphenator.i_hyphenate(self,aWord)
#print "BaseHyphenator.i_hyphenate returned %r" % hword
if hword is None:
hword = HyphenatedWord(aWord,hyphenations=[])
base_hyph_points = hword.hyphenations
last_indx = 0
nr = 0
for hpnum, hp in enumerate(base_hyph_points):
if isinstance(hp, int):
hp = HyphenationPoint(hp, quality=5, sl=SHY)
subword = hword[last_indx+nr:hp.indx]
# handle subword
if SHY in subword:
sub_hword = self.stripper.apply_stripped(BaseHyphenator.hyph, self, subword)
else:
sub_hword = self.stripper.apply_stripped(ExplicitHyphenator.hyph, self, subword)
if sub_hword is None:
sub_hword = self.stripper.apply_stripped(self.hyph, self, subword)
if sub_hword is None:
sub_hword = HyphenatedWord(subword, hyphenations=[])
sub_hwords.append(sub_hword)
# end handle subword
last_indx = hp.indx
nr = hp.nr
# Now the last subword
subword = hword[last_indx:]
# handle subword
if SHY in subword:
sub_hword = self.stripper.apply_stripped(BaseHyphenator.hyph, self, subword)
else:
sub_hword = self.stripper.apply_stripped(ExplicitHyphenator.hyph, self, subword)
if sub_hword is None:
sub_hword = self.stripper.apply_stripped(self.hyph, self, subword)
if sub_hword is None:
sub_hword = HyphenatedWord(subword, hyphenations=[])
sub_hwords.append(sub_hword)
#end handle subword
if len(sub_hwords) > 1:
return HyphenatedWord.join(sub_hwords)
else:
return sub_hwords[0] # Kann auch None sein.
def hyph(self, word):
#print "ExplicitHyphenator hyph", word
lenword = len(word)
for (lae, L) in self.sonderfaelle:
if lae == lenword:
trennung = L.get(word.lower(), None)
if trennung is not None:
hword = HyphenatedWord(word, decodeTrennung(trennung))
return hword
break
# Wort nicht gefunden
return None
def hyph(self, word):
log.debug("DCW hyphenate %r", word)
assert isinstance(word, unicode)
loesungen = self.zerlegeWort(word)
if len(loesungen) > 1:
# Trennung ist nicht eindeutig, z.B. bei WachsTube oder WachStube.
#hword.info = ("AMBIGUOUS", loesungen)
# nimm nur solche Trennstellen, die in allen Lösungen vorkommen,
# und für die Qualität nimm die schlechteste.
loesung = []
loesung0, andere = loesungen[0], loesungen[1:]
for i, hp in enumerate(loesung0):
q = hp.quality
for a in andere:
if q:
for hp1 in a:
if hp1.indx==hp.indx \
and hp1.nl==hp.nl and hp1.sl==hp.sl \
and hp1.nr==hp.nr and hp1.sr==hp.sr:
q = min(q, hp1.quality)
break
else:
# Trennstelle nicht in der anderen Lösung enthalten
q = 0
if q:
loesung.append(
HyphenationPoint(hp.indx, q, hp.nl, hp.sl, hp.nr,
hp.sr))
if loesung:
# Es gibt mindestens eine Trennstelle, die bei allen Varianten
# enthalten ist, z.b. Wachstu-be.
pass
# hword.info = ("HYPHEN_OK", loesung)
else:
# Es gibt keine Trennstelle.
pass
elif len(loesungen) == 1:
# Trennung ist eindeutig
loesung = loesungen[0]
#hword.info = ("HYPHEN_OK", loesung)
if not loesung:
pass # hword.info = ("NOT_HYPHENATABLE", aWord)
else:
# Das Wort ist uns unbekannt.
return None
return HyphenatedWord(word, loesung)