-
Notifications
You must be signed in to change notification settings - Fork 0
/
search_engine_best.py
183 lines (165 loc) · 7.33 KB
/
search_engine_best.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
import os
import pandas as pd
from parser_module_stamming import Parse_stem
from reader import ReadFile
from configuration import ConfigClass
from parser_module import Parse
from indexer import Indexer
from searcher_word2vec import Searcher
from timeit import default_timer as timer
from datetime import timedelta
from gensim.models import KeyedVectors
import utils
#word2vec
# DO NOT CHANGE THE CLASS NAME
class SearchEngine:
# DO NOT MODIFY THIS SIGNATURE
# You can change the internal implementation, but you must have a parser and an indexer.
def __init__(self, config=None):
if config == None:
config = ConfigClass()
self._config = config
if config.toStem:
self._parser = Parse_stem()
else:
self._parser = Parse()
self._indexer = Indexer(config)
self._model = None
# DO NOT MODIFY THIS SIGNATURE
# You can change the internal implementation as you see fit.
def build_index_from_parquet(self, fn):
"""
Reads parquet file and passes it to the parser, then indexer.
Input:
fn - path to parquet file
Output:
No output, just modifies the internal _indexer object.
"""
config = self._config
indexer = self._indexer
number_of_documents = 0
if(config.getoneFile()):
df = pd.read_parquet(fn, engine="pyarrow")
documents_list = df.values.tolist()
# Iterate over every document in the file
for idx, document in enumerate(documents_list):
# parse the document
parsed_document = self._parser.parse_doc(document)
number_of_documents += 1
# index the document data
self._indexer.add_new_doc(parsed_document)
self._indexer.calculationSummerize()
else:
r = ReadFile(corpus_path=config.get__corpusPath())
for root, dirs, files in os.walk(config.get__corpusPath(), topdown=True):
for name in files:
ext = name.split('.')[-1]
if ext == 'parquet':
documents_list = r.read_folder(root, file_name=name)
# Iterate over every document in the file
for idx, document in enumerate(documents_list):
# parse the document
parsed_document = self._parser.parse_doc(document)
number_of_documents += 1
# index the document data
indexer.add_new_doc(parsed_document)
# indexer.update_posting_files()
# indexer.reset_cach()
self._indexer.save_index('inverted_idx')
print('Finished parsing and indexing.')
# def get_full_text(self, d_id):
# return self._indexer.documents_data[d_id][4]
# DO NOT MODIFY THIS SIGNATURE
# You can change the internal implementation as you see fit.
def load_index(self, fn):
"""
Loads a pre-computed index (or indices) so we can answer queries.
Input:
fn - file name of pickled index.
"""
self._indexer.load_index(fn)
# DO NOT MODIFY THIS SIGNATURE
# You can change the internal implementation as you see fit.
def load_precomputed_model(self, model_dir=None):#TODO implement
"""
Loads a pre-computed model (or models) so we can answer queries.
This is where you would load models like word2vec, LSI, LDA, etc. and
assign to self._model, which is passed on to the searcher at query time.
"""
self._model = KeyedVectors.load_word2vec_format('GoogleNews-vectors-negative300.bin', binary=True)
# self._model = KeyedVectors.load_word2vec_format(self._config.google_news_vectors_negative300_path, binary=True)
# DO NOT MODIFY THIS SIGNATURE
# You can change the internal implementation as you see fit.
def search(self, query):
"""
Executes a query over an existing index and returns the number of
relevant docs and an ordered list of search results.
Input:
query - string.
Output:
A tuple containing the number of relevant search results, and
a list of tweet_ids where the first element is the most relavant
and the last is the least relevant result.
"""
if self._indexer.inverted_idx == None:
print("can't run query without inverted index been loaded")
return
searcher = Searcher(self._parser, self._indexer, model=self._model)
return searcher.search(query)
def main(self, queries = None, num_docs_to_retrieve= None):
config = self._config
# config.set_corpusPath(corpus_path)
# config.set_savedFileMainFolder(output_path)
# config.set_toStem(stemming)
self.load_precomputed_model()
vectorModel = self._model
start = timer()
print("----started parsing and indexer----")
self.build_index_from_parquet('data/benchmark_data_train.snappy.parquet')
end = timer()
print("Process ends..")
# print(timedelta(seconds=end - start))
if num_docs_to_retrieve == None:
num_docs_to_retrieve = 2000
inverted_index = self.load_index('inverted_idx')
if queries == None: #for end users use
user_query = input("Please enter a query")
user_num = int(input("How many tweets you want to get (maximum)?"))
res = self.search(user_query)
print("here are the links to the tweets related to use query")
for x in range(user_num):
try:
tweeter_start_link = 'https://twitter.com/IsraelHayomHeb/status/'
tweet_id = (res[1][x])
print(tweeter_start_link+tweet_id)
except:
pass #less than number of doc to retrive found
else:#for engenier use
import csv
with open('queries_output.csv', 'w', newline='') as file:
writer = csv.writer(file)
writer.writerow(["query", "tweet"])
if not isinstance(queries, list):
try:
f = open(queries, "r+", encoding='utf-8')
# queries = f.read()
queries = f.readlines()
f.close()
except Exception:
raise
print("fail in reading queries file")
# see numbers for the document file
i = 0
for querie in queries:
print("querie number" + str(i))
print(querie)
i += 1
res = self.search(querie)
for s in range(num_docs_to_retrieve):
try:
writer.writerow([i,res[1][s]])
except:
pass # less than number of doc to retrive found
# for doc in res[1]:
# #print('Tweet id: {}, Score: {}'.format(doc_tuple[1], doc_tuple[0]))
# writer.writerow([i, doc])