def phrase_search(search_phrase, mode):
# len_of_query = preprocess_squery(search_phrase,mode)[1]
term = preprocess_squery(search_phrase, mode)
len_of_query = term.__len__()
t1 = set(term_psearch(term[0], mode))
t2 = set(term_psearch(term[-1], mode))
term_ids = list(t1 & t2)
i = 0
# {'1901-00001': {'pos': [59], 'tf': 1}, 'df': 1}
IDtftl = [] # doc_id contains the first and last word
term_dic_f = read_index(term[0], mode)
term_dic_l = read_index(term[-1], mode)
if term_dic_f == None or term_dic_l == None:
return 'None'
else:
for docid in term_ids:
if term_dic_f.__contains__(docid) and term_dic_l.__contains__(
docid):
posif = term_dic_f.get(docid)['pos']
posil = term_dic_l.get(docid)['pos']
m = 0
n = 0
while m < posif.__len__():
while n < posil.__len__():
if (int(posil[n]) - int(posif[m]) == len_of_query - 1):
IDtftl.append(docid)
n += 1
m += 1
docid_phrase = []
if len_of_query > 2:
for id_query in IDtftl:
for i in range(len_of_query - 1):
dictf = read_index(term[i], mode)
dictl = read_index(term[i + 1], mode)
if dictf.__contains__(id_query) and dictl.__contains__(
id_query):
posif = dictf.get(id_query)['pos']
posil = dictl.get(id_query)['pos']
m = 0
n = 0
while m < posif.__len__():
while n < posil.__len__():
if (int(posil[n]) - int(posif[m]) == 1):
docid_phrase.append(id_query)
n += 1
m += 1
else:
docid_phrase = IDtftl
phrase_docid = set(docid_phrase)
if docid_phrase.__len__() == 0:
return 'None'
else:
return phrase_docid
def term_psearch(term, mode):
docid_list = []
term_dic = read_index(term, mode)
# print(term_dic) {'_id': 'constant', 'df': 2277, '1901-00002': {'tf': 2, 'pos': [28, 40]}, '1901-00018': {'pos': [76], 'tf': 1}, '1901-00028': {'pos': [10], 'tf': 1}, '1901-00119': {'pos': [42, 76], 'tf': 2}, '1901-00210': {'pos': [93], 'tf': 1},
if term_dic == None: #term is not in index
return '0'
else:
docid = term_dic.keys()
for item in docid:
if item != 'df' and item != '_id':
docid_list.append(item)
return docid_list
def term_psearch(term,mode):
docid_list = []
term_dic = read_index(term,mode)
if term_dic == None or len(term_dic) == 1: #term is not in index
return '0'
else :
docid = term_dic.keys()
for item in docid:
if item!= '_id':
docid_list.append(item)
return docid_list
def term_search(term,mode):
# posi_list = []
# if readindex(file_path).__contains__(term):
# posi_list = list(text[term]['docdict'].keys())
# return posi_list
docid_list = []
term_dic = read_index(term, mode)
if term_dic == None: #term is not in index
return '0'
else :
docid = term_dic.keys()
for item in docid:
if item != 'df' and item!= '_id':
docid_list.append(item)
return docid_list