-
Notifications
You must be signed in to change notification settings - Fork 0
/
engine.py
262 lines (209 loc) · 8.57 KB
/
engine.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
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
import zipfile
import json
import re
import pickle
from sklearn.feature_extraction.text import TfidfVectorizer
from urllib.parse import urlparse
from nltk.stem import PorterStemmer
from bs4 import BeautifulSoup
from dataclasses import dataclass,field
from collections import Counter, defaultdict
from typing import DefaultDict, List, Set
from flask import Flask, render_template, request
@dataclass
class indexStats():
"""
Class that keeps track of all the stats of the index
"""
numDocs: int = 0
length: int=0
uniquePages: int = 0
totalSize: int = 0
hits: int = 0
tf: int = 0
ifd: int = 0
tf_idf: int = 0
ps: PorterStemmer = PorterStemmer()
indexDict: DefaultDict[str, List] = field(default_factory = lambda: defaultdict(list))
tf_idf_values: DefaultDict[str, List] = field(default_factory = lambda: defaultdict(list))
uniqueTokens: Set[str] = field(default_factory = set)
searchTokens: List[str] = field(default_factory = list)
top_urls: List[str] = field(default_factory = list)
import_words: List[str] = field(default_factory = list)
similarity_matrix: list[list[float]] = None
app = Flask(__name__, static_url_path='/static')
stats = indexStats()
@app.route("/")
def home():
return render_template("index.html")
@app.route("/search", methods=['GET'])
def search():
query = request.args.get('query')
# global stats
if bool(
re.search(r'\band\b', query.lower(), flags=re.IGNORECASE)):
tokens = query.split("and")
stats.searchTokens = [stats.ps.stem(token) for token in tokens]
else:
tokens = query.split(" ")
stats.searchTokens = [stats.ps.stem(token) for token in tokens]
searchIndex()
return render_template("index.html", urls = stats.top_urls, search=query, hits=stats.hits)
def valid(url, content):
"""checks if url is a valid html file"""
parsed = urlparse(url)
if not parsed.path.endswith(".html"):
return False
try:
soup = BeautifulSoup(content, "html.parser")
except:
return False
if "html" not in soup.contents[0]:
return False
return True
def index(path:str):
"""
Takes in a path to a zip file, and reads its contents
:param path:
path to the zip file
:return: None
"""
# Open the zip file
with zipfile.ZipFile(path, 'r') as zip_ref:
# List all files in the zip archive
file_list = zip_ref.namelist()
# Iterate through each file in the zip archive
for file_name in file_list:
stats.numDocs+=1
# Check if the file is a JSON file (you can customize this condition)
if file_name.endswith('.json'):
# Read the JSON content directly from the zip archive
with zip_ref.open(file_name) as json_file:
# Load JSON content
json_data = json.load(json_file)
if not valid(json_data['url'], json_data['content']):
pass
stats.numDocs += 1
soup = BeautifulSoup(json_data['content'], "html.parser")
alphanumeric_words = re.findall(r'\b\w+\b', soup.get_text())
corpus = [stats.ps.stem(word.lower()) for word in alphanumeric_words] #all words in a file
num_of_words_in_file = len(corpus)
counter = Counter(corpus)
#gets each word and frequency in each file
for word, count in counter.items():
stats.indexDict[word].append((json_data['url'], count))
def merge(num:int):
"""
merges partial indexes
:param num:
:return:
"""
for i in range(num):
with open(f"partial indexes/index_{i+1}.pkl", "rb") as file:
indexData = pickle.load(file)
stats.indexDict.update(indexData)
def mergeTifidf(num:int):
"""
Merges the indexed files into the dictionary that holds values of tfidf
"""
for i in range(num):
with open(f"index_w_tfidf_{i+1}.pkl", "rb") as file:
indexData = pickle.load(file)
stats.tf_idf_values.update(indexData)
def create_partial_index():
num_partitions = 3
partSize = len(stats.indexDict) // num_partitions
for i in range(num_partitions):
start = i * partSize
end = (i+1) * partSize
splitData = {k: stats.indexDict[k] for k in list(stats.indexDict)[start:end]}
with open(f'index_w_tfidf_{i + 1}.pkl', 'wb') as file:
pickle.dump(splitData, file)
def calculateTFIDF(path: str):
"""
Takes in a path to a zip file, and reads its contents
:param path:
path to the zip file
:return: None
"""
# Initialize TfidfVectorizer
vectorizer = TfidfVectorizer()
# Open the zip file
with zipfile.ZipFile(path, 'r') as zip_ref:
# List all files in the zip archive
file_list = zip_ref.namelist()
# Iterate through each file in the zip archive
i = 0
for file_name in file_list:
stats.numDocs += 1
# Check if the file is a JSON file
if file_name.endswith('.json'):
# Read the JSON content directly from the zip archive
with zip_ref.open(file_name) as json_file:
# Load JSON content
json_data = json.load(json_file)
if not valid(json_data['url'], json_data['content']):
pass
stats.numDocs += 1
soup = BeautifulSoup(json_data['content'], "html.parser")
if len(stats.import_words) == 0:
stats.import_words = [heading.text.lower() for heading in soup.find_all(['h1', 'h2', 'h3'])] + [word for bold_tag in soup.find_all('b') for word in re.findall(r'\b\w+\b', bold_tag.get_text())]
alphanumeric_words = re.findall(r'\b\w+\b', soup.get_text())
corpus = [stats.ps.stem(word.lower()) for word in
alphanumeric_words] # all words in a file
# Convert the list of words into a string for TfidfVectorizer
document_text = ' '.join(corpus)
# Vectorize the document using TF-IDF
try:
tfidf_matrix = vectorizer.fit_transform([document_text])
except ValueError:
pass
feature_names = vectorizer.get_feature_names_out()
# Get each word and its TF-IDF score in the document
for word, tfidf_score in zip(feature_names, tfidf_matrix.toarray()[0]):
stats.indexDict[word].append((json_data['url'], tfidf_score))
if word.lower in stats.import_words: #check if word is important word
stats.indexDict.get(word.lower)[1] += 2
if i % 100 == 0:
print(i)
i += 1
def create_partial_index():
num_partitions = 3
partSize = len(stats.indexDict) // num_partitions
for i in range(num_partitions):
start = i * partSize
end = (i+1) * partSize
splitData = {k: stats.indexDict[k] for k in list(stats.indexDict)[start:end]}
with open(f'index_w_tfidf_{i + 1}.pkl', 'wb') as file:
pickle.dump(splitData, file)
def searchIndex():
matching_urls = None
for token in stats.searchTokens:
result = stats.tf_idf_values.get(token)
# Update matching_urls based on the current token
if result:
if matching_urls is None:
matching_urls = set(result)
else:
matching_urls.intersection_update(result)
# store top 5 urls tht contain all tokens
if matching_urls:
sorted_data = sorted(matching_urls, key=lambda x: x[1], reverse=True)
stats.top_urls = sorted_data[:5]
stats.hits = len(sorted_data)
else:
stats.top_urls = ["No matching URLs found."]
if __name__ == "__main__":
command = input('Type "index" to create index or type "run" to run application: ')
if command == "index":
path = input("Enter the path to zip file: ").strip('"')
calculateTFIDF(path)
create_partial_index()
if command == "run":
stats = indexStats()
if not stats.tf_idf_values:
print("Indexing...")
mergeTifidf(3)
# Wait for both threads to finish
print("Index Complete!")
app.run(port = 8000)