def enc_first(val, **kwargs):
return encode(val, **kwargs)[0][1]
round(score * 20))
# Determine SENTENCE_SIMILARITY
score = 0
for word in stripped_sentence:
score += variable_vocabulary[variable][word]
score = 1 - ((score / 4) - min_variable_occurence) / (
max_variable_occurence - min_variable_occurence)
scores[subject][variable][Scores.SENTENCE_SIMILARITY] = int(
round(score * 20))
# Determine RHYTHMIC_SCORE
score = 0
phonetic_result = cologne_phonetics.encode(
sentence.strip(" ,;.:!?").lower())
sounds_to_word_groups = {}
for i in range(len(phonetic_result)):
for j in range(i + 1, len(phonetic_result)):
(word1, sound1) = phonetic_result[i]
(word2, sound2) = phonetic_result[j]
for x in range(1, 1 + min(len(sound1), len(sound2))):
if ((sound1[-x:] == sound2[-x:])
and not (word1 == word2)):
adj_word1 = word1
adj_word2 = word2
def test_returns_altered(self):
self.assertEqual(encode("bäTes")[0][0], "baetes")
def test_case_insensitive(self):
self.assertEqual(encode("foo"), encode("FoO"))
def test_concatenation(self):
self.assertTrue(encode("a-a") == encode("a a"))
self.assertEqual(encode("a-a", concat=True), [("a-a", '0')])
self.assertEqual(encode("a a", concat=True), [('a', '0'), ('a', '0')])
def multiple_before(self, char=None, before=None, exp=None):
for b in before:
self.assertEqual(encode(char + b), exp)
def authores_features(segcite, sowiport):
prep_data_ss = aux_author_features_gen(segcite, sowiport)
# First author features
# ============================
if len(list(prep_data_ss["first_aut_ln"])) > 0 and len(prep_data_ss['sowi_ln_norm_ls']):
First_author_OO = 1
if normalizeinput_author1(prep_data_ss["ln_ls_order"][0])[0] == prep_data_ss["sowi_ln_norm_ls"][0]:
First_author_Exact = 1
else:
First_author_Exact = 0
if len(prep_data_ss["ln_ls_order"]) > 0 and len(prep_data_ss["sowi_ls_ls"]) > 0:
if cologne_phonetics.encode(prep_data_ss["ln_ls_order"][0])[0][1] == \
cologne_phonetics.encode(prep_data_ss["sowi_ls_ls"][0])[0][1]:
First_author_Phono = 1
else:
First_author_Phono = 0
else:
First_author_Phono = 0
Pro_seg_fa = float(list(prep_data_ss['first_aut_ln'].values())[0])
else:
First_author_OO = 0
First_author_Exact = 0
First_author_Phono = 0
lsfatemp = list(prep_data_ss['first_aut_ln'].values())
if len(lsfatemp) > 0:
Pro_seg_fa = float(lsfatemp[0])
else:
Pro_seg_fa = 0
# levenstien Exact/Phono
# ==============================
lower_ls_sowi_ls = [x.lower().strip() for x in prep_data_ss['sowi_ls_ls']]
lower_ls_segcite_ls = [x.lower().strip() for x in prep_data_ss['ln_ls_order']]
phono_ls_sowi_ls = [cologne_phonetics.encode(x.lower().strip())[0][1] for x in prep_data_ss['sowi_ls_ls']]
phono_ls_seg_ls = [cologne_phonetics.encode(x.lower().strip())[0][1] for x in prep_data_ss['ln_ls_order']]
if len(lower_ls_sowi_ls) > 0 and len(lower_ls_segcite_ls) > 0:
if len(prep_data_ss['sowi_ls_ls']) > 0 and len(prep_data_ss['sowi_gn_ls']) > 0 and len(
prep_data_ss['ln_ls_order']) and len(prep_data_ss['gn_aux_dict']):
try:
ln_x = []
for inx, dls in enumerate(prep_data_ss['ln_ls_order']):
ln_x.append(
normalizeinput_author1(dls)[0] + prep_data_ss['gn_aux_dict'].get(inx, [""])[0][0].lower())
ln_y = []
for ixy, it_ls in enumerate(prep_data_ss['sowi_ls_ls']):
ln_y.append(normalizeinput_author1(it_ls)[0] + prep_data_ss['sowi_gn_ls'][ixy][0][0])
leven_pr_exact = levenshtein(ln_x, ln_y)
except:
leven_pr_exact = levenshtein(lower_ls_sowi_ls, lower_ls_segcite_ls)
else:
leven_pr_exact = levenshtein(lower_ls_sowi_ls, lower_ls_segcite_ls)
leven_pr_phono = levenshtein(phono_ls_sowi_ls, phono_ls_seg_ls)
Jacard_pr_exact = jaccard_similarity(lower_ls_sowi_ls, lower_ls_segcite_ls)
jacard_pr_phono = jaccard_similarity(phono_ls_sowi_ls, phono_ls_seg_ls)
author_oo = 1
else:
leven_pr_exact = 0
leven_pr_phono = 0
Jacard_pr_exact = 0
jacard_pr_phono = 0
author_oo = 0
intersect = list(set(lower_ls_sowi_ls).intersection(set(lower_ls_segcite_ls)))
# ==============================
keys_ls_dict = list(prep_data_ss["ln_dict_score"].keys())
ls_pr_ln_temp = []
for item_key in keys_ls_dict:
if item_key.lower() in intersect:
ls_pr_ln_temp.append(float(prep_data_ss["ln_dict_score"][item_key]))
if len(ls_pr_ln_temp) > 0:
jacard_pr_exact_scr = sum(ls_pr_ln_temp) / len(set(lower_ls_sowi_ls + lower_ls_sowi_ls))
else:
jacard_pr_exact_scr = 0
return [First_author_OO, author_oo, First_author_Exact, First_author_Phono, Pro_seg_fa, leven_pr_exact,
leven_pr_phono, Jacard_pr_exact, jacard_pr_phono, jacard_pr_exact_scr]