-
Notifications
You must be signed in to change notification settings - Fork 1
/
company_score.py
196 lines (173 loc) · 8.27 KB
/
company_score.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
import json
import pandas as pd
import requests
from webhook import Webhook
from fullcontact import FullContact
from email_guess import EmailGuess
from queue import RQueue
from parse import Parse
import companies
from fuzzywuzzy import fuzz
from fuzzywuzzy import process
#from companies import Companies
from queue import RQueue
from rq import Queue
from worker import conn
q = Queue(connection=conn)
class CompanyScore:
def _bulk_info():
''' add industry, phones, and suff'''
def _company_info(self, company_name, api_key=""):
#TODO - company_name = self._remove_non_ascii(company_name) add to save
qry = {'where':json.dumps({'company_name': company_name}), 'limit':1000}
qry['order'] = '-createdAt'
crawls = Parse().get('CompanyInfoCrawl', qry).json()['results']
if not crawls:
# start crawls
return company_name
crawls = self._source_score(pd.DataFrame(crawls))
crawls = self._logo_score(crawls)
#crawls = crawls[crawls.api_key == api_key]
crawls['name_score'] = [fuzz.token_sort_ratio(row['name'], row.company_name)
for index, row in crawls.iterrows()]
crawls = crawls[crawls.name_score > 70].append(crawls[crawls.name.isnull()])
logo = crawls.sort("logo_score",ascending=False)
#logo=logo[(logo.logo != "") & (logo.logo.notnull())][["source","logo"]]
logo=logo[(logo.logo != "") & (logo.logo.notnull())].logo.tolist()
logo = logo[0] if logo else ""
#crawls = crawls[["press", 'source_score', 'source', 'createdAt', 'domain']]
final = {}
#print crawls.press.dropna()
for col in crawls.columns:
if col in ['source_score', 'source', 'createdAt']: continue
df = crawls[[col, 'source_score', 'source', 'createdAt']]
if df[col].dropna().empty: continue
if type(list(df[col].dropna())[0]) == list:
df[col] = df[col].dropna().apply(tuple)
try: df = df[df[col] != ""]
except: "lol"
try:
df = df[df[col].notnull()]
df = [source[1].sort('createdAt').drop_duplicates(col, True)
for source in df.groupby(col)]
df = [_df for _df in df if _df is not None]
df = [pd.DataFrame(columns=['source_score',col])] if len(df) is 0 else df
df = pd.concat(df).sort('source_score')[col]
if list(df): final[col] = list(df)[-1]
except: "lol"
if 'industry' in final.keys():
try:
final['industry'] = final['industry'][0]
except:
final["industry"] = ""
try:
final['industry_keywords'] = list(set(crawls.industry.dropna().sum()))
except:
final['industry_keywords'] = []
if 'address' in final.keys():
final['address'] = FullContact()._normalize_location(final['address'])
try:
final['handles'] = crawls[['source','handle']].dropna()
final['handles'] = final['handles'].drop_duplicates().to_dict('r')
except:
"lol"
tmp = crawls[['source','logo']].dropna()
#print tmp
#print "THE LOGO", logo
final["logo"] = logo
final['logos'] = tmp.drop_duplicates().to_dict('r')
try:
tmp = crawls[['source','phone']].dropna()
final['phones'] = tmp.drop_duplicates().to_dict('r')
except:
""" """
# TODO - if company_name exists update
# TODO - find if domain exists under different company_name then update
final = self._prettify_fields(final)
if "name_score" in final.keys(): del final["name_score"]
#print json.dumps(final)
self._add_to_clearspark_db('Company', 'company_name', company_name, final)
# TODO - find main domain from domain -> ie canon.ca should be canon.com
# clean data - ie titleify fields, and lowercase domain
# TODO - start a domain search with the deduced domain and the company_name
#print "RQUEUE CHECK"
if "domain" in final.keys():
domain = final["domain"]
'''
if len(RQueue()._results("{0}_{1}".format(company_name, api_key))) == 1:
q.enqueue(Companies()._domain_research, domain, api_key, company_name)
q.enqueue(Companies()._secondary_research, company_name, domain, api_key)
'''
if RQueue()._has_completed("{0}_{1}".format(company_name, api_key)):
#q.enqueue(Companies()._domain_research, domain, api_key, company_name)
#q.enqueue(Companies()._secondary_research, company_name, domain, api_key)
print "WEBHOOK <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<"
if "company_name" in final.keys():
Webhook()._update_company_info(final)
'''
job = q.enqueue(EmailGuess().search_sources, final["domain"],api_key,"")
job.meta["{0}_{1}".format(company_name, api_key)] = True
job.save()
for domain in crawls.domain.dropna().drop_duplicates():
job = q.enqueue(EmailGuess().search_sources, domain, api_key, "")
RQueue()._meta(job, "{0}_{1}".format(company_name, api_key))
'''
return final
def _prettify_fields(self, final):
if "domain" in final.keys():
final['domain'] = final['domain'].lower()
# titlify everything else ?
return final
def _add_to_clearspark_db(self, class_name, column, value, data):
qry = json.dumps({column: value})
obj = Parse().get(class_name, {'where': qry}).json()['results']
#print obj
if obj:
print "NEW UPDATE OLD", class_name+'/'+obj[0]['objectId']
print Parse().update(class_name+'/'+obj[0]['objectId'], data).json()
else:
print "NEW CREATE NEW"
print Parse().create(class_name, data).json()
def _company_check(self, company_name, domain, data, class_name="Company"):
qry = json.dumps({'domain': domain})
domain_check = Parse().get(class_name, {'where': qry}).json()['results']
qry = json.dumps({'company_name': company_name})
name_check = Parse().get(class_name, {'where': qry}).json()['results']
if domain_check == [] and name_check == []:
print "NEW CREATE NEW"
print Parse().create(class_name, data).json()
else:
print "NEW UPDATE OLD", class_name+'/'+domain_check[0]['objectId']
print Parse().update(class_name+'/'+domain_check[0]['objectId'], data).json()
# TODO - add search query
def _email_webhook_should_be_called(self, crawls):
return True
def _webhook_should_be_called(self, crawls):
print crawls.source.drop_duplicates().shape[0]
return True
def _source_score(self, df):
df.ix[df.source == "linkedin", 'source_score'] = 10
df.ix[df.source == "zoominfo", 'source_score'] = 9
df.ix[df.source == "yelp", 'source_score'] = 2
df.ix[df.source == "yellowpages", 'source_score'] = 3
df.ix[df.source == "facebook", 'source_score'] = 1
df.ix[df.source == "twitter", 'source_score'] = 0
df.ix[df.source == "businessweek", 'source_score'] = 4
df.ix[df.source == "forbes", 'source_score'] = 5
df.ix[df.source == "hoovers", 'source_score'] = 6
df.ix[df.source == "crunchbase", 'source_score'] = 7
df.ix[df.source == "glassdoor", 'source_score'] = 8
return df
def _logo_score(self, df):
df.ix[df.source == "linkedin", 'logo_score'] = 8
df.ix[df.source == "zoominfo", 'logo_score'] = 5
df.ix[df.source == "yelp", 'logo_score'] = 1
df.ix[df.source == "yellowpages", 'logo_score'] = 2
df.ix[df.source == "facebook", 'logo_score'] = 9
df.ix[df.source == "twitter", 'logo_score'] = 10
df.ix[df.source == "businessweek", 'logo_score'] = 3
df.ix[df.source == "forbes", 'logo_score'] = 4
df.ix[df.source == "hoovers", 'logo_score'] = 0
df.ix[df.source == "crunchbase", 'logo_score'] = 6
df.ix[df.source == "glassdoor", 'logo_score'] = 7
return df