DrishtiSharma commited on
Commit
13627f6
·
verified ·
1 Parent(s): 8908600

Delete bad_app.py

Browse files
Files changed (1) hide show
  1. bad_app.py +0 -237
bad_app.py DELETED
@@ -1,237 +0,0 @@
1
- import sys
2
- import os
3
- import re
4
- import shutil
5
- import time
6
- import streamlit as st
7
- import nltk
8
- import tempfile
9
- import subprocess
10
- import base64 # For embedding PDF content
11
-
12
- # Pin NLTK to version 3.9.1
13
- REQUIRED_NLTK_VERSION = "3.9.1"
14
- subprocess.run([sys.executable, "-m", "pip", "install", f"nltk=={REQUIRED_NLTK_VERSION}"])
15
-
16
- # Set up temporary directory for NLTK resources
17
- nltk_data_path = os.path.join(tempfile.gettempdir(), "nltk_data")
18
- os.makedirs(nltk_data_path, exist_ok=True)
19
- nltk.data.path.append(nltk_data_path)
20
-
21
- # Download 'punkt_tab' for compatibility
22
- try:
23
- print("Ensuring NLTK 'punkt_tab' resource is downloaded...")
24
- nltk.download("punkt_tab", download_dir=nltk_data_path)
25
- except Exception as e:
26
- print(f"Error downloading NLTK 'punkt_tab': {e}")
27
- raise e
28
-
29
- sys.path.append(os.path.abspath("."))
30
- from langchain.chains import ConversationalRetrievalChain
31
- from langchain.memory import ConversationBufferMemory
32
- from langchain.llms import OpenAI
33
- from langchain.document_loaders import UnstructuredPDFLoader
34
- from langchain.vectorstores import Chroma
35
- from langchain.embeddings import HuggingFaceEmbeddings
36
- from langchain.text_splitter import NLTKTextSplitter
37
- from patent_downloader import PatentDownloader
38
-
39
- PERSISTED_DIRECTORY = tempfile.mkdtemp()
40
-
41
- # Fetch API key securely from the environment
42
- OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
43
- if not OPENAI_API_KEY:
44
- st.error("Critical Error: OpenAI API key not found in the environment variables. Please configure it.")
45
- st.stop()
46
-
47
- def check_poppler_installed():
48
- if not shutil.which("pdfinfo"):
49
- raise EnvironmentError(
50
- "Poppler is not installed or not in PATH. Install 'poppler-utils' for PDF processing."
51
- )
52
-
53
- check_poppler_installed()
54
-
55
- def load_docs(document_path):
56
- try:
57
- loader = UnstructuredPDFLoader(
58
- document_path,
59
- mode="elements",
60
- strategy="fast",
61
- ocr_languages=None
62
- )
63
- documents = loader.load()
64
- text_splitter = NLTKTextSplitter(chunk_size=1000)
65
- split_docs = text_splitter.split_documents(documents)
66
-
67
- # Filter metadata to only include str, int, float, or bool
68
- for doc in split_docs:
69
- if hasattr(doc, "metadata") and isinstance(doc.metadata, dict):
70
- doc.metadata = {
71
- k: v for k, v in doc.metadata.items()
72
- if isinstance(v, (str, int, float, bool))
73
- }
74
- return split_docs
75
- except Exception as e:
76
- st.error(f"Failed to load and process PDF: {e}")
77
- st.stop()
78
-
79
- def already_indexed(vectordb, file_name):
80
- indexed_sources = set(
81
- x["source"] for x in vectordb.get(include=["metadatas"])["metadatas"]
82
- )
83
- return file_name in indexed_sources
84
-
85
- def load_chain(file_name=None):
86
- loaded_patent = st.session_state.get("LOADED_PATENT")
87
-
88
- vectordb = Chroma(
89
- persist_directory=PERSISTED_DIRECTORY,
90
- embedding_function=HuggingFaceEmbeddings(),
91
- )
92
- if loaded_patent == file_name or already_indexed(vectordb, file_name):
93
- st.write("✅ Already indexed.")
94
- else:
95
- vectordb.delete_collection()
96
- docs = load_docs(file_name)
97
- st.write("🔍 Number of Documents: ", len(docs))
98
-
99
- vectordb = Chroma.from_documents(
100
- docs, HuggingFaceEmbeddings(), persist_directory=PERSISTED_DIRECTORY
101
- )
102
- vectordb.persist()
103
- st.session_state["LOADED_PATENT"] = file_name
104
-
105
- memory = ConversationBufferMemory(
106
- memory_key="chat_history",
107
- return_messages=True,
108
- input_key="question",
109
- output_key="answer",
110
- )
111
- return ConversationalRetrievalChain.from_llm(
112
- OpenAI(temperature=0, openai_api_key=OPENAI_API_KEY),
113
- vectordb.as_retriever(search_kwargs={"k": 3}),
114
- return_source_documents=False,
115
- memory=memory,
116
- )
117
-
118
- def extract_patent_number(url):
119
- pattern = r"/patent/([A-Z]{2}\d+)"
120
- match = re.search(pattern, url)
121
- return match.group(1) if match else None
122
-
123
- def download_pdf(patent_number):
124
- try:
125
- patent_downloader = PatentDownloader(verbose=True)
126
- output_path = patent_downloader.download(patents=patent_number, output_path=tempfile.gettempdir())
127
- return output_path[0]
128
- except Exception as e:
129
- st.error(f"Failed to download patent PDF: {e}")
130
- st.stop()
131
-
132
- def embed_pdf(file_path):
133
- """Convert PDF file to base64 and embed it in an iframe."""
134
- with open(file_path, "rb") as f:
135
- base64_pdf = base64.b64encode(f.read()).decode("utf-8")
136
- pdf_display = f"""
137
- <iframe src="data:application/pdf;base64,{base64_pdf}" width="700" height="1000" style="border: none;"></iframe>
138
- """
139
- return pdf_display
140
-
141
- if __name__ == "__main__":
142
- st.set_page_config(
143
- page_title="Patent Chat: Google Patents Chat Demo",
144
- page_icon="📖",
145
- layout="wide",
146
- initial_sidebar_state="expanded",
147
- )
148
- st.header("���� Patent Chat: Google Patents Chat Demo")
149
-
150
- # Fetch query parameters safely
151
- query_params = st.query_params
152
- default_patent_link = query_params.get("patent_link", "https://patents.google.com/patent/US8676427B1/en")
153
-
154
- # Input for Google Patent Link
155
- patent_link = st.text_area("Enter Google Patent Link:", value=default_patent_link, height=100)
156
-
157
- # Button to start processing
158
- if st.button("Load and Process Patent"):
159
- if not patent_link:
160
- st.warning("Please enter a Google patent link to proceed.")
161
- st.stop()
162
-
163
- patent_number = extract_patent_number(patent_link)
164
- if not patent_number:
165
- st.error("Invalid patent link format. Please provide a valid Google patent link.")
166
- st.stop()
167
-
168
- st.write(f"Patent number: **{patent_number}**")
169
-
170
- # Define PDF path in temp directory
171
- pdf_path = os.path.join(tempfile.gettempdir(), f"{patent_number}.pdf")
172
- if os.path.isfile(pdf_path):
173
- st.write("✅ File already downloaded.")
174
- else:
175
- st.write("📥 Downloading patent file...")
176
- pdf_path = download_pdf(patent_number)
177
- st.write(f"✅ File downloaded: {pdf_path}")
178
-
179
- # Display a preview of the downloaded PDF
180
- st.write("📄 Preview of the downloaded patent PDF:")
181
- if os.path.isfile(pdf_path):
182
- with open(pdf_path, "rb") as pdf_file:
183
- st.download_button(
184
- label="Download PDF",
185
- data=pdf_file,
186
- file_name=f"{patent_number}.pdf",
187
- mime="application/pdf"
188
- )
189
- # Embed PDF content using base64
190
- st.write("📋 PDF Content:")
191
- pdf_view = embed_pdf(pdf_path)
192
- st.components.v1.html(pdf_view, height=1000)
193
-
194
- st.write("🔄 Loading document into the system...")
195
-
196
- # Persist the chain in session state to prevent reloading
197
- if "chain" not in st.session_state or st.session_state.get("loaded_file") != pdf_path:
198
- st.session_state.chain = load_chain(pdf_path)
199
- st.session_state.loaded_file = pdf_path
200
- st.session_state.messages = [{"role": "assistant", "content": "Hello! How can I assist you with this patent?"}]
201
-
202
- st.success("🚀 Document successfully loaded! You can now start asking questions.")
203
-
204
- # Initialize messages if not already done
205
- if "messages" not in st.session_state:
206
- st.session_state.messages = [{"role": "assistant", "content": "Hello! How can I assist you with this patent?"}]
207
-
208
- # Display previous chat messages
209
- for message in st.session_state.messages:
210
- with st.chat_message(message["role"]):
211
- st.markdown(message["content"])
212
-
213
- # Chat Input Section
214
- if "chain" in st.session_state:
215
- if user_input := st.chat_input("What is your question?"):
216
- # Append user message
217
- st.session_state.messages.append({"role": "user", "content": user_input})
218
- with st.chat_message("user"):
219
- st.markdown(user_input)
220
-
221
- # Generate assistant response
222
- with st.chat_message("assistant"):
223
- message_placeholder = st.empty()
224
- full_response = ""
225
-
226
- with st.spinner("Generating response..."):
227
- try:
228
- assistant_response = st.session_state.chain({"question": user_input})
229
- full_response = assistant_response["answer"]
230
- except Exception as e:
231
- full_response = f"An error occurred: {e}"
232
-
233
- # Display assistant response
234
- message_placeholder.markdown(full_response)
235
- st.session_state.messages.append({"role": "assistant", "content": full_response})
236
- else:
237
- st.info("Press the 'Load and Process Patent' button to start processing.")