Spaces:
Sleeping
Sleeping
Upload function.py
Browse files- function.py +67 -0
function.py
ADDED
@@ -0,0 +1,67 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from langchain.llms import OpenAI
|
2 |
+
from langchain.chains import AnalyzeDocumentChain
|
3 |
+
from langchain.document_loaders import PyPDFLoader
|
4 |
+
from langchain.document_loaders import UnstructuredEmailLoader
|
5 |
+
from langchain.document_loaders import Docx2txtLoader
|
6 |
+
from langchain.chains.question_answering import load_qa_chain
|
7 |
+
import os
|
8 |
+
from pandas_ai import csv_file,excel_file
|
9 |
+
|
10 |
+
|
11 |
+
def pdf_file(file_upload,message):
|
12 |
+
page_text = ''
|
13 |
+
loader = PyPDFLoader(file_upload)
|
14 |
+
pages = loader.load_and_split()
|
15 |
+
for page in pages:
|
16 |
+
page_text = page_text + page.page_content
|
17 |
+
os.environ['OPENAI_API_KEY'] = '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'
|
18 |
+
llm = OpenAI(temperature=0)
|
19 |
+
qa_chain = load_qa_chain(llm, chain_type="map_reduce")
|
20 |
+
qa_pdf_chain = AnalyzeDocumentChain(combine_docs_chain=qa_chain)
|
21 |
+
output = qa_pdf_chain.run(input_document=page_text,question=message)
|
22 |
+
return output
|
23 |
+
|
24 |
+
def email_file(file_upload,message):
|
25 |
+
email_text = ''
|
26 |
+
loader = UnstructuredEmailLoader(file_upload,mode="elements",process_attachments=True)
|
27 |
+
emails = loader.load()
|
28 |
+
for email in emails:
|
29 |
+
email_text = email_text + email.page_content
|
30 |
+
os.environ['OPENAI_API_KEY'] = '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'
|
31 |
+
llm = OpenAI(temperature=0)
|
32 |
+
qa_chain = load_qa_chain(llm, chain_type="map_reduce")
|
33 |
+
qa_email_chain = AnalyzeDocumentChain(combine_docs_chain=qa_chain)
|
34 |
+
output = qa_email_chain.run(input_document=email_text,question=message)
|
35 |
+
return output
|
36 |
+
|
37 |
+
def docx_file(file_upload,message):
|
38 |
+
doc_text = ''
|
39 |
+
loader = Docx2txtLoader(file_upload)
|
40 |
+
documents = loader.load()
|
41 |
+
for doc in documents:
|
42 |
+
doc_text = doc_text + doc.page_content
|
43 |
+
os.environ['OPENAI_API_KEY'] = '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'
|
44 |
+
llm = OpenAI(temperature=0)
|
45 |
+
qa_chain = load_qa_chain(llm, chain_type="map_reduce")
|
46 |
+
qa_document_chain = AnalyzeDocumentChain(combine_docs_chain=qa_chain)
|
47 |
+
output = qa_document_chain.run(input_document=doc_text,question=message)
|
48 |
+
return output
|
49 |
+
|
50 |
+
def initiate_process(file_upload,extention,message):
|
51 |
+
# try:
|
52 |
+
if extention=='pdf':
|
53 |
+
output = pdf_file(file_upload,message)
|
54 |
+
elif extention=='eml':
|
55 |
+
output = email_file(file_upload,message)
|
56 |
+
elif extention=='docx':
|
57 |
+
output = docx_file(file_upload,message)
|
58 |
+
elif extention=='csv':
|
59 |
+
output = csv_file(file_upload,message)
|
60 |
+
elif extention=='xlsx':
|
61 |
+
output = excel_file(file_upload,message)
|
62 |
+
else:
|
63 |
+
output = "Please upload correct file format"
|
64 |
+
return output
|
65 |
+
# except Exception as e:
|
66 |
+
# output = str(e)
|
67 |
+
# return output
|