def postprocess(word_as_list, fst):
if fst.postprocess_req:
if not fst.left_subseq:
word_as_list = word_as_list[::-1]
if fst.name == "Igbo ATR harmony":
post_fst = FST(vh_dataset['22B'])
output = post_fst.step(word_as_list)
word_as_list = output[::-1]
elif fst.name == "Diola-Fogny (Jola-Fonyi) ATR harmony":
word_as_list = fst.step(word_as_list)
word_as_list = word_as_list[::-1]
elif fst.name == \
"Pasiego vowel harmony (metaphony, raising, and centralization)":
word_as_list = word_as_list[::-1]
fst_b = FST(vh_dataset['26B'])
word_as_list = fst_b.step(word_as_list)
fst_c = FST(vh_dataset['26C'])
word_as_list = fst_c.step(word_as_list)[::-1]
elif fst.name == "Kalenjin ATR harmony":
# reversing this for lazy (but probably not efficient) parsing
word_as_list = word_as_list[::-1]
if '+' in word_as_list:
final_prefixes = word_as_list[word_as_list.index('+'
):][::-1]
stem = word_as_list[:word_as_list.index('+')][::-1]
else:
final_prefixes = []
stem = word_as_list
if '-' in stem:
stem = stem[:stem.index('-')]
post_language = vh_dataset['24B']
post_fst = FST(post_language)
suffix_as_list = []
[suffix_as_list.extend(suff) for suff in fst.suffix]
processed_stem_and_suffix = post_fst.step(stem +
suffix_as_list)
word_as_list = final_prefixes + processed_stem_and_suffix
else:
if fst.name == "Maasai (Eastern Nilotic) ATR harmony":
word_as_list = word_as_list[::-1]
fst_b = vh_dataset['28B']
word_as_list = fst_b.step(word_as_list)[::-1]
if fst.name == "Nawuri (North Guang) ATR harmony":
fst_b = vh_dataset['31B']
reversed = word_as_list[::-1]
word_as_list = fst_b.step(reversed)[::-1]
return "".join(word_as_list)
def preprocess(word_as_list, prefix_as_list, stem_as_list, suffix_as_list,
fst):
fst.prefix = prefix_as_list
fst.suffix = suffix_as_list
# empty input
if word_as_list == []:
return False, []
# if preprocessing isn't required on the input string
if not fst.preprocess_req:
return True, word_as_list
# Preprocessing for left-subsequential languages
if fst.left_subseq:
if fst.name == "Kisa applicative suffix Vlɑ"\
or fst.name == "Kisa reversative suffix Vlɑ":
return True, word_as_list[:-3]
elif fst.name in {
"Uyghur backness harmony", "Uyghur plural suffix -lVr",
'Uyghur dative suffix -' + U_F_V + 'V'
}:
# run the preprocess step
preliminary_language = vh_dataset['5P']
preliminary_fst = FST(preliminary_language)
if fst.name == "Uyghur backness harmony":
return True, preliminary_fst.step(word_as_list)
elif fst.name == "Uyghur plural suffix -lVr":
word_as_list = word_as_list[:-3]
return True, preliminary_fst.step(word_as_list)
elif fst.name == 'Uyghur dative suffix -' + U_F_V + 'V':
word_as_list = word_as_list[:-2]
return True, preliminary_fst.step(word_as_list)
elif fst.name == "Halh (Mongolic) rounding harmony":
# run the preprocess step
preliminary_language = vh_dataset['8P']
preliminary_fst = FST(preliminary_language)
return True, preliminary_fst.step(word_as_list)
elif fst.name == "Kalmyk (Oirat) harmony":
'''AI---------------------------------------------------------------
Rules for language 17
---------------------------------------------------------------AI'''
# If there is a suffix, hyphenate
# Check if i is the only type of vowel in the stem
# if not, run stem + suffix through 17
# if yes, don't process the stem. Input the suffix into 17P.
# Output is the stored stem + output of 17P
alphabet_as_set = set(fst.alphabet)
vowels_in_stem = set()
for ch in stem_as_list:
if ch in fst.alphabet:
vowels_in_stem.add(ch)
if vowels_in_stem.intersection(alphabet_as_set) != {'i'}:
return True, word_as_list
else:
if suffix_as_list != []:
preliminary_language = vh_dataset['17P']
preliminary_fst = FST(preliminary_language)
fst.suffix = preliminary_fst.step(suffix_as_list[0])
return False, stem_as_list + fst.suffix
elif fst.name in {
"Maasai (Eastern Nilotic) ATR harmony",
"Nawuri (North Guang) ATR harmony"
}:
stem_as_list.insert(0, '!')
stem_as_list.append('&')
return True, prefix_as_list + stem_as_list + suffix_as_list
elif fst.name == "Kashaya (Pomoan) translaryngeal harmony":
fst_p = vh_dataset['29P']
return True, fst_p.step(word_as_list)
elif fst.name == "Standard Hungarian palatal harmony of alternating suffixes":
stem_as_list.insert(0, '!')
return True, prefix_as_list + stem_as_list + suffix_as_list
elif fst.name == "Tunica harmony":
word_as_list.append('#')
return True, word_as_list
else:
return True, word_as_list
# Preprocessing for right-subsequential languages
else:
# languages 9,10,11,12,13
if fst.name in {
'Jingulu nominal root with non-neuter gender suffix',
'Jingulu adjectivial root with non-neuter gender suffix',
'Jingulu verbal root with subject agreement-marking suffix',
'Jingulu verbal root with motion-imperative suffix',
'Jingulu verbal root with negative imperative suffix',
}:
return True, word_as_list[::-1]
# if not "u" in fst.suffix[0] and not "i" in fst.suffix[0]:
# return False, word_as_list
# else:
# suffix_start = word_as_list.index('-')
# return True, word_as_list[:suffix_start][::-1]
elif fst.name == 'Yoruba ATR harmony':
return True, word_as_list[::-1]
elif fst.name == "Kalenjin ATR harmony":
language = vh_dataset[24]
def add_delimiters(affix_list):
for affix in affix_list:
if affix in language['n-a_suff']:
affix.insert(0, '&')
elif affix in language['n-a_r&pre']:
affix.append('!')
return affix_list
# prefix pre-processing
prefix_as_list = add_delimiters(prefix_as_list)
suffix_as_list = add_delimiters(suffix_as_list)
stem_as_list = add_delimiters(stem_as_list)
prefix_flat = []
[prefix_flat.extend(prefix) for prefix in prefix_as_list]
prefix_as_list = prefix_flat
if len(suffix_as_list) >= 1:
suffix_to_add = suffix_as_list[0]
else:
suffix_to_add = []
return True, (prefix_as_list + stem_as_list + suffix_to_add)[::-1]
# h e h i + h i + h i + a w q - a e
elif fst.name==\
"Asturian Lena (Romance) height harmony with inflectional suffixes":
return True, word_as_list[::-1]
return True, word_as_list[::-1]