def add_necessary_breathing(w):
s = syllabify(w)
o, n, c = onset_nucleus_coda(s[0])
if o == "" and not breathing(n):
a = accent(n)
if a:
if len(n) == 2:
n = n[0] + add_diacritic(add_diacritic(base(n[1]), SMOOTH), a)
else:
n = add_diacritic(add_diacritic(base(n), SMOOTH), a)
else:
if len(n) == 2:
n = n[0] + add_diacritic(base(n[1]), SMOOTH)
else:
n = add_diacritic(base(n), SMOOTH)
return o + n + c + "".join(s[1:])
else:
return w
def syllable_accent(s):
for ch in nucleus(s):
a = accent(ch)
if a:
return a
def generate(self, lemma, parse, allow_form_override=True, context=None):
answers = []
stems = None
accent_override = None
is_enclitic = False
ending_override = None
if lemma in self.lexicon:
if allow_form_override:
answer = self.lexicon[lemma].get("forms", {}).get(parse)
if answer:
return answer
stems = self.regex_list(lemma, parse, context)
if "." in parse:
accents = self.lexicon[lemma].get("accents", {}).get(parse.split(".")[0])
if accents == "enclitic":
is_enclitic = True
else:
accent_override = accents
ending_override = self.lexicon[lemma].get("endings", {}).get(parse)
if stems is None:
return
else:
stems = stems.split("/")
if parse not in stemming_rules:
return
for stem in stems:
stem = debreath(stem)
pairs = stemming_rules[parse]
while isinstance(pairs, dict) and "ref" in pairs:
if pairs["ref"] in stemming_rules:
pairs = stemming_rules[pairs["ref"]]
else:
# @@@ raise error?
return
base_endings = []
default = []
for rule in pairs:
s1, s234, s5 = rule.split("|")
s2, s34 = s234.split(">")
s3, s4 = s34.split("<")
if stem.endswith(strip_accents(s1 + s2)):
if s2:
base = stem[:-len(s2)]
else:
base = stem
else:
continue
if ending_override:
ending_list = ending_override.split("/")
else:
ending_list = [s3 + s5]
if s1 + s2:
base_endings.append((base, ending_list))
else:
default.append((base, ending_list))
# only use default if there are no other options
if len(base_endings) == 0 and len(default) > 0:
base_endings = default
for base, ending_list in base_endings:
for ending in ending_list:
if accent(ending):
answers.append((base + ending).replace("|", ""))
elif is_enclitic:
answers.append(make_oxytone(base + ending).replace("|", ""))
else:
if parse[2] == "P":
if accent_override:
answers.append(persistent(base + ending, accent_override))
elif parse == "AAP.NSM" and ending == "ων":
answers.append(make_oxytone(base + ending).replace("|", ""))
elif parse == "AAP.NSM" and ending == "_3+ς":
answers.append(make_oxytone(base + ending).replace("|", ""))
elif parse == "PAP.NSM" and ending == "_3+ς":
answers.append(make_oxytone(base + ending).replace("|", ""))
elif parse[0:3] == "AAP" and parse != "AAP.NSM":
# calculate NSM
nsms = self.generate(lemma, "AAP.NSM", context=context)
nsms = nsms.split("/")
for nsm in nsms:
if nsm.endswith(("ών", "ούς")):
answers.append(persistent(base + ending, nsm))
else:
answers.append(persistent(base + ending, lemma))
elif parse[0:3] == "PAP" and parse != "PAP.NSM":
# calculate NSM
nsms = self.generate(lemma, "PAP.NSM").split("/")
for nsm in nsms:
nsm = strip_length(nsm)
answers.append(persistent(base + ending, nsm))
else:
answers.append(recessive(base + ending, default_short=True))
elif parse[0:3] in ["AAN", "XAN", "XMN", "XPN"]:
answers.append(on_penult(base + ending, default_short=True))
elif parse[0:3] == "PAN" and stem.endswith("!"):
answers.append(on_penult(base + ending, default_short=True))
else:
answers.append(recessive(base + ending, default_short=True))
return "/".join(remove_duplicates(rebreath(w) for w in answers))