Spaces:
Sleeping
Sleeping
Update controllers/fetch_microsoft_mails.py
Browse files
controllers/fetch_microsoft_mails.py
CHANGED
@@ -1,4 +1,3 @@
|
|
1 |
-
@app.get("/emails")
|
2 |
def get_emails(request: Request):
|
3 |
access_token = request.session.get('access_token')
|
4 |
if not access_token:
|
@@ -41,4 +40,270 @@ def get_emails(request: Request):
|
|
41 |
|
42 |
emails.append(email_info)
|
43 |
|
44 |
-
return {"Successfully extracted the messages!!"}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
def get_emails(request: Request):
|
2 |
access_token = request.session.get('access_token')
|
3 |
if not access_token:
|
|
|
40 |
|
41 |
emails.append(email_info)
|
42 |
|
43 |
+
return {"Successfully extracted the messages!!"}
|
44 |
+
|
45 |
+
import base64
|
46 |
+
import json
|
47 |
+
import logging
|
48 |
+
import re
|
49 |
+
from concurrent.futures import ThreadPoolExecutor
|
50 |
+
from typing import Optional, List, Dict
|
51 |
+
import requests
|
52 |
+
from bs4 import BeautifulSoup
|
53 |
+
from models.models import Message, Attachment
|
54 |
+
from fastapi import WebSocket
|
55 |
+
from services import utils as ut
|
56 |
+
import asyncio
|
57 |
+
|
58 |
+
def get_company_type(company_name:str)->str:
|
59 |
+
company_types_dict ={'ao yun': 'wines and spirit', 'ardbeg': 'wines and spirit', 'belvedere': 'wines and spirit', 'bodega numanthia': 'wines and spirit', 'chandon': 'wines and spirit', 'château cheval blanc': 'wines and spirit', "château d'yquem": 'wines and spirit', 'château galoupet': 'wines and spirit', 'cheval des andes': 'wines and spirit', 'clos19': 'wines and spirit', 'cloudy bay': 'wines and spirit', 'colgin cellars': 'wines and spirit', 'dom pérignon': 'wines and spirit', 'domaine des lambrays': 'wines and spirit', 'eminente': 'wines and spirit', 'glenmorangie': 'wines and spirit', 'hennessy': 'wines and spirit', 'joseph phelps': 'wines and spirit', 'krug': 'wines and spirit', 'mercier': 'wines and spirit', 'moët & chandon': 'wines and spirit', 'newton vineyard': 'wines and spirit', 'ruinart': 'wines and spirit', 'terrazas de los andes': 'wines and spirit', 'veuve clicquot': 'wines and spirit', 'volcan de mi tierra': 'wines and spirit', 'woodinville': 'wines and spirit' , 'berluti': 'Fashion & Leather Goods', 'celine': 'Fashion & Leather Goods', 'christian dior': 'Fashion & Leather Goods', 'emilio pucci': 'Fashion & Leather Goods', 'fendi': 'Fashion & Leather Goods', 'givenchy': 'Fashion & Leather Goods', 'kenzo': 'Fashion & Leather Goods', 'loewe': 'Fashion & Leather Goods', 'loro piana': 'Fashion & Leather Goods', 'louis vuitton': 'Fashion & Leather Goods', 'marc jacobs': 'Fashion & Leather Goods', 'moynat': 'Fashion & Leather Goods', 'patou': 'Fashion & Leather Goods', 'rimowa': 'Fashion & Leather Goods','acqua di parma': 'Perfumes & Cosmetics', 'benefit cosmetics': 'Perfumes & Cosmetics', 'cha ling': 'Perfumes & Cosmetics', 'fenty beauty by rihanna': 'Perfumes & Cosmetics', 'fresh': 'Perfumes & Cosmetics', 'givenchy parfums': 'Perfumes & Cosmetics', 'guerlain': 'Perfumes & Cosmetics', 'kenzo parfums': 'Perfumes & Cosmetics', 'kvd beauty': 'Perfumes & Cosmetics', 'loewe perfumes': 'Perfumes & Cosmetics', 'maison francis kurkdjian': 'Perfumes & Cosmetics', 'make up for ever': 'Perfumes & Cosmetics', 'officine universelle buly': 'Perfumes & Cosmetics', 'olehenriksen': 'Perfumes & Cosmetics', 'parfums christian dior': 'Perfumes & Cosmetics', 'stella by stella mccartney': 'Perfumes & Cosmetics','bulgari': 'Watches & Jewelry', 'chaumet': 'Watches & Jewelry', 'fred': 'Watches & Jewelry', 'hublot': 'Watches & Jewelry', 'repossi': 'Watches & Jewelry', 'tag heuer': 'Watches & Jewelry', 'tiffany & co.': 'Watches & Jewelry', 'zenith': 'Watches & Jewelry','24s': 'Selective retailing', 'dfs': 'Selective retailing', 'la grande epicerie de paris': 'Selective retailing', 'le bon marché rive gauche': 'Selective retailing', 'sephora': 'Selective retailing','belmond': 'Other activities', 'cheval blanc': 'Other activities', 'connaissance des arts': 'Other activities', 'cova': 'Other activities', 'investir': 'Other activities', "jardin d'acclimatation": 'Other activities', 'le parisien': 'Other activities', 'les echos': 'Other activities', 'radio classique': 'Other activities', 'royal van lent': 'Other activities'}
|
60 |
+
print(company_types_dict["louis vuitton"])
|
61 |
+
return company_types_dict.get(company_name.lower(), 'Others')
|
62 |
+
|
63 |
+
async def get_messages(code: str,websocket:WebSocket,brand_name: Optional[str] = None):
|
64 |
+
access_token = code
|
65 |
+
g_query = f'(subject:"your order" OR subject:receipts OR subject:receipt OR subject:aankoopbon OR subject:reçu OR subject:invoice OR subject:invoices OR category:purchases) has:attachment'
|
66 |
+
if brand_name is not None:
|
67 |
+
g_query = f'(subject:"your order" OR subject:receipts OR subject:receipt OR subject: aankoopbon OR subject:reçu OR subject:invoice OR subject:invoices OR category:purchases OR from:{brand_name}) AND subject:{brand_name} has:attachment'
|
68 |
+
page_token = None
|
69 |
+
messages = []
|
70 |
+
# max_results = 10
|
71 |
+
# gmail_url = f"https://www.googleapis.com/gmail/v1/users/me/messages?q={jobs_query}&maxResults={max_results}"
|
72 |
+
# gmail_response = requests.get(gmail_url, headers={"Authorization": f"Bearer {access_token}"})
|
73 |
+
# gmail_data = gmail_response.json()
|
74 |
+
# messages.append(gmail_data['messages'])
|
75 |
+
def fetch_message_wrapper(message_data):
|
76 |
+
message_id = message_data.get("id")
|
77 |
+
if message_id:
|
78 |
+
return fetch_message_data(access_token, message_id)
|
79 |
+
|
80 |
+
return None
|
81 |
+
|
82 |
+
while True:
|
83 |
+
gmail_url = f"https://www.googleapis.com/gmail/v1/users/me/messages?q={g_query}"
|
84 |
+
if page_token:
|
85 |
+
gmail_url += f"&pageToken={page_token}"
|
86 |
+
|
87 |
+
gmail_response = requests.get(gmail_url, headers={"Authorization": f"Bearer {access_token}"})
|
88 |
+
gmail_data = gmail_response.json()
|
89 |
+
print(len(gmail_data))
|
90 |
+
print(gmail_data)
|
91 |
+
|
92 |
+
if "messages" in gmail_data:
|
93 |
+
with ThreadPoolExecutor(max_workers=15) as executor:
|
94 |
+
|
95 |
+
|
96 |
+
|
97 |
+
futures=[executor.submit(fetch_message_wrapper, message_data) for message_data in
|
98 |
+
gmail_data["messages"]]
|
99 |
+
for future in futures:
|
100 |
+
message = future.result()
|
101 |
+
if message:
|
102 |
+
messages.append(message)
|
103 |
+
for message_data in messages:
|
104 |
+
await process_message(message_data,websocket,10000)
|
105 |
+
|
106 |
+
if "nextPageToken" in gmail_data:
|
107 |
+
page_token = gmail_data["nextPageToken"]
|
108 |
+
else:
|
109 |
+
break
|
110 |
+
print("printing messages")
|
111 |
+
print(messages)
|
112 |
+
return messages
|
113 |
+
|
114 |
+
async def process_message(message:Message, websocket:WebSocket, chunk_size:int):
|
115 |
+
logging.info("process_message")
|
116 |
+
if message:
|
117 |
+
message_json = message.to_json()
|
118 |
+
logging.info(f"{message_json}")
|
119 |
+
await send_message_in_chunks(websocket, message_json, chunk_size)
|
120 |
+
await websocket.send_text("NEXT_MESSAGE")
|
121 |
+
|
122 |
+
|
123 |
+
def fetch_message_data(access_token: str, message_id: str) -> Message:
|
124 |
+
message_url = f"https://www.googleapis.com/gmail/v1/users/me/messages/{message_id}"
|
125 |
+
message_response = requests.get(message_url, headers={"Authorization": f"Bearer {access_token}"})
|
126 |
+
message_data = message_response.json()
|
127 |
+
# print(message_data)
|
128 |
+
subject = extract_subject_from_mail(message_data)
|
129 |
+
company_from_mail = extract_domain_name(message_data['payload']['headers'], subject)
|
130 |
+
|
131 |
+
body = extract_body_from_mail(message_data)
|
132 |
+
|
133 |
+
attachments,structed_attachment_data = extract_attachments_from_mail(access_token, message_data)
|
134 |
+
high_level_company_type = get_company_type(company_from_mail)
|
135 |
+
# structed_attachment_data = extract_json_from_attachments(access_token , message_data)
|
136 |
+
|
137 |
+
|
138 |
+
body_len = 0
|
139 |
+
if body is not None :
|
140 |
+
body_len = len(body)
|
141 |
+
|
142 |
+
# print("subject: ")
|
143 |
+
# print(subject)
|
144 |
+
# print("company name: ")
|
145 |
+
# print(company_from_mail)
|
146 |
+
# print("Printing the body of the mail: ")
|
147 |
+
# print(body)
|
148 |
+
# print("Printing attachment Data: ")
|
149 |
+
# print(attachments)
|
150 |
+
# print("Completed this mail.")
|
151 |
+
return Message(message_id=message_id, body_len=body_len,body=body, attachments=attachments, company=company_from_mail,high_level_company_type=high_level_company_type,structured_data = structed_attachment_data)
|
152 |
+
|
153 |
+
|
154 |
+
|
155 |
+
def extract_subject_from_mail(message_data: dict) -> str:
|
156 |
+
if 'payload' in message_data and 'headers' in message_data['payload']:
|
157 |
+
headers = message_data['payload']['headers']
|
158 |
+
for header in headers:
|
159 |
+
if header['name'] == 'Subject':
|
160 |
+
return header['value']
|
161 |
+
return ""
|
162 |
+
else:
|
163 |
+
return ""
|
164 |
+
|
165 |
+
|
166 |
+
def extract_domain_name(payload: dict, subject: str) -> str:
|
167 |
+
domain_name = 'others'
|
168 |
+
for fromdata in payload:
|
169 |
+
if fromdata['name'] == 'From':
|
170 |
+
domain_name = extract_domain_from_email(fromdata['value'])
|
171 |
+
break
|
172 |
+
if 'chanel' in subject.lower():
|
173 |
+
return 'chanel'
|
174 |
+
if 'louis vuitton' in subject.lower():
|
175 |
+
return 'Louis Vuitton'
|
176 |
+
return domain_name
|
177 |
+
|
178 |
+
|
179 |
+
def extract_domain_from_email(email_string: str) -> Optional[str]:
|
180 |
+
email_address = re.search(r'[\w\.-]+@[\w\.-]+', email_string).group()
|
181 |
+
domain = email_address.split('@')[-1].split('.')[0]
|
182 |
+
if email_address and domain:
|
183 |
+
return domain
|
184 |
+
else:
|
185 |
+
return None
|
186 |
+
|
187 |
+
|
188 |
+
# def extract_body_from_mail(message_data: dict) -> str:
|
189 |
+
# body = None
|
190 |
+
# if "payload" in message_data and "parts" in message_data["payload"]:
|
191 |
+
# for part in message_data["payload"]["parts"]:
|
192 |
+
# if 'mimeType' in part and (part['mimeType'] == 'text/plain' or part['mimeType'] == 'text/html'):
|
193 |
+
# body_data = part['body'].get('data', '')
|
194 |
+
# body_base64 = base64.urlsafe_b64decode(body_data)
|
195 |
+
# body = extract_text(body_base64)
|
196 |
+
# return body
|
197 |
+
|
198 |
+
|
199 |
+
def extract_body_from_mail(message_data: dict) -> str:
|
200 |
+
body = None
|
201 |
+
if "payload" in message_data:
|
202 |
+
payload = message_data["payload"]
|
203 |
+
if "parts" in payload:
|
204 |
+
for part in payload["parts"]:
|
205 |
+
if 'mimeType' in part and (part['mimeType'] == 'text/plain' or part['mimeType'] == 'text/html'):
|
206 |
+
body_data = part['body'].get('data', '')
|
207 |
+
if body_data:
|
208 |
+
body_base64 = base64.urlsafe_b64decode(body_data)
|
209 |
+
body = extract_text(body_base64)
|
210 |
+
|
211 |
+
elif 'body' in payload:
|
212 |
+
body_data = payload['body'].get('data', '')
|
213 |
+
if body_data:
|
214 |
+
body_base64 = base64.urlsafe_b64decode(body_data)
|
215 |
+
body = extract_text(body_base64)
|
216 |
+
elif 'parts' in payload['body']:
|
217 |
+
for part in payload['body']['parts']:
|
218 |
+
if 'mimeType' in part and (part['mimeType'] == 'text/plain' or part['mimeType'] == 'text/html'):
|
219 |
+
body_data = part['body'].get('data', '')
|
220 |
+
if body_data:
|
221 |
+
body_base64 = base64.urlsafe_b64decode(body_data)
|
222 |
+
body = extract_text(body_base64)
|
223 |
+
|
224 |
+
if not body:
|
225 |
+
body = message_data.get('snippet', '')
|
226 |
+
return body
|
227 |
+
|
228 |
+
|
229 |
+
def fetch_attachment_data(access_token: str, message_id: str, attachment_id: str) -> Dict:
|
230 |
+
attachment_url = f"https://www.googleapis.com/gmail/v1/users/me/messages/{message_id}/attachments/{attachment_id}"
|
231 |
+
attachment_response = requests.get(attachment_url, headers={"Authorization": f"Bearer {access_token}"})
|
232 |
+
return attachment_response.json()
|
233 |
+
|
234 |
+
|
235 |
+
def extract_attachments_from_mail(access_token: str, message_data: dict) -> List[Attachment]:
|
236 |
+
attachments = []
|
237 |
+
structured_data = []
|
238 |
+
if "payload" in message_data and "parts" in message_data["payload"]:
|
239 |
+
for part in message_data["payload"]["parts"]:
|
240 |
+
if "body" in part and "attachmentId" in part["body"]:
|
241 |
+
attachment_id = part["body"]["attachmentId"]
|
242 |
+
attachment_data = fetch_attachment_data(access_token, message_data["id"], attachment_id)
|
243 |
+
filename = part.get("filename", "untitled.txt")
|
244 |
+
if filename.endswith(".zip") or filename.endswith(".txt") or filename.endswith(".png") or filename.endswith(".jpg") or filename.endswith(".jpeg") or filename.endswith(".gif"):
|
245 |
+
continue
|
246 |
+
data = attachment_data.get("data", "")
|
247 |
+
raw_text=ut.extract_text_from_attachment(filename , data)
|
248 |
+
struct_data = ut.strcuture_document_data(raw_text)
|
249 |
+
if struct_data:
|
250 |
+
structured_data.append(struct_data)
|
251 |
+
|
252 |
+
attachments.append(Attachment(attachment_len = len(attachment_data.get("data", "")),filename=filename, data=attachment_data.get("data", "")))
|
253 |
+
return attachments,structured_data
|
254 |
+
|
255 |
+
|
256 |
+
def extract_text(html_content: str) -> str:
|
257 |
+
if not html_content:
|
258 |
+
raise ValueError("HTML content is empty or None")
|
259 |
+
|
260 |
+
soup = BeautifulSoup(html_content, 'html.parser')
|
261 |
+
text = soup.get_text(separator=' ')
|
262 |
+
text = re.sub(r'\s+', ' ', text).strip()
|
263 |
+
return text
|
264 |
+
|
265 |
+
|
266 |
+
async def websocket_main(code: str, websocket: WebSocket,brand_name: Optional[str] = None):
|
267 |
+
access_token = code
|
268 |
+
# messages = get_messages(access_token,websocket,brand_name)
|
269 |
+
await get_messages(access_token,websocket,brand_name)
|
270 |
+
# print("websocket_main")
|
271 |
+
# print(messages)
|
272 |
+
# # logging.info(f"brand_name:{brand_name}")
|
273 |
+
# await websocket.send_json({"total_messages": len(messages)})
|
274 |
+
# print("Total Length of messages")
|
275 |
+
# print(len(messages))
|
276 |
+
# chunk_size = 100000
|
277 |
+
# i=0
|
278 |
+
# for message in messages:
|
279 |
+
# message_json = message.to_json()
|
280 |
+
|
281 |
+
# logging.info(f"{i} th message")
|
282 |
+
# i=i+1
|
283 |
+
# await send_message_in_chunks(websocket, message_json, chunk_size)
|
284 |
+
# await websocket.send_text("NEXT_MESSAGE")
|
285 |
+
|
286 |
+
|
287 |
+
await websocket.close()
|
288 |
+
|
289 |
+
|
290 |
+
async def send_message_in_chunks(websocket: WebSocket, message_json: dict, chunk_size: int):
|
291 |
+
|
292 |
+
# if message_json['attachments'] is not None :
|
293 |
+
# for attch in message_json['attachments']:
|
294 |
+
# attachment_len = attch['attachment_len']
|
295 |
+
|
296 |
+
|
297 |
+
# print(body_len)
|
298 |
+
# print(attachment_len)
|
299 |
+
# if attachment_len == 0:
|
300 |
+
# attachment_len = None
|
301 |
+
# await websocket.send_json({"body_len":body_len ,"attachment_len":attachment_len})
|
302 |
+
|
303 |
+
message_str = json.dumps(message_json)
|
304 |
+
# print("Printing message_str")
|
305 |
+
# print(message_str)
|
306 |
+
# logging.info(message_str)
|
307 |
+
# await websocket.send_json({"file_len":len(file)})
|
308 |
+
for i in range(0, len(message_str), chunk_size):
|
309 |
+
await websocket.send_text(message_str[i:i + chunk_size])
|