Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -1,17 +1,18 @@
|
|
| 1 |
import os
|
| 2 |
import streamlit as st
|
| 3 |
import tempfile
|
| 4 |
-
from pymongo import MongoClient
|
| 5 |
from datetime import datetime
|
| 6 |
from pathlib import Path
|
| 7 |
-
from
|
| 8 |
from urllib.parse import quote_plus
|
|
|
|
| 9 |
|
| 10 |
# === MongoDB connection via Hugging Face secrets ===
|
| 11 |
user = quote_plus(os.getenv("MONGO_USER"))
|
| 12 |
password = quote_plus(os.getenv("MONGO_PASS"))
|
| 13 |
cluster = os.getenv("MONGO_CLUSTER")
|
| 14 |
db_name = os.environ.get("MONGO_DB", "grant_docs")
|
|
|
|
| 15 |
mongo_uri = f"mongodb+srv://{user}:{password}@{cluster}/{db_name}?retryWrites=true&w=majority&tls=true&tlsAllowInvalidCertificates=true"
|
| 16 |
client = MongoClient(mongo_uri, tls=True, tlsAllowInvalidCertificates=True, serverSelectionTimeoutMS=20000)
|
| 17 |
db = client[db_name]
|
|
@@ -22,12 +23,11 @@ st.title("📄 Document Chunker & Uploader")
|
|
| 22 |
|
| 23 |
with st.sidebar:
|
| 24 |
st.header("Settings")
|
| 25 |
-
|
| 26 |
-
# Fetch collection names for dropdown
|
| 27 |
try:
|
| 28 |
existing_collections = db.list_collection_names()
|
| 29 |
existing_collections.append("Create New Collection")
|
| 30 |
-
|
|
|
|
| 31 |
except Exception as e:
|
| 32 |
st.error(f"Failed to list collections: {e}")
|
| 33 |
selected_collection = "doc_chunks_cat"
|
|
@@ -42,33 +42,46 @@ with st.sidebar:
|
|
| 42 |
|
| 43 |
uploaded_file = st.file_uploader("Upload a DOCX, TXT, or PDF file", type=["docx", "txt", "pdf"])
|
| 44 |
|
| 45 |
-
|
|
|
|
| 46 |
temp_path = Path(tempfile.gettempdir()) / uploaded_file.name
|
| 47 |
with open(temp_path, "wb") as f:
|
| 48 |
f.write(uploaded_file.getbuffer())
|
| 49 |
|
| 50 |
-
st.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 51 |
|
| 52 |
-
|
| 53 |
-
collection = db[
|
| 54 |
|
| 55 |
-
if collection.find_one({"metadata.title":
|
| 56 |
st.warning("⚠️ This file already exists in the collection. Skipping...")
|
| 57 |
else:
|
| 58 |
st.write("⏳ Processing with DocumentChunker...")
|
| 59 |
chunker = DocumentChunker()
|
| 60 |
-
chunks = chunker.process_document(
|
| 61 |
|
| 62 |
if chunks:
|
| 63 |
for chunk in chunks:
|
| 64 |
chunk['metadata'].update({
|
| 65 |
-
"title":
|
| 66 |
-
"uploaded_at":
|
| 67 |
"is_grant_app": is_grant_app,
|
| 68 |
})
|
| 69 |
collection.insert_one(chunk)
|
| 70 |
|
| 71 |
-
st.success(f"✅ {len(chunks)} chunks inserted into `{
|
| 72 |
|
| 73 |
for i, c in enumerate(chunks[:3]):
|
| 74 |
st.subheader(f"Chunk {i+1}: {c['metadata'].get('header') or 'No Header'}")
|
|
@@ -78,14 +91,16 @@ if uploaded_file:
|
|
| 78 |
|
| 79 |
if len(chunks) > 3:
|
| 80 |
st.info(f"... and {len(chunks)-3} more chunks processed.")
|
| 81 |
-
|
| 82 |
else:
|
| 83 |
st.warning("⚠️ No chunks were generated.")
|
| 84 |
|
| 85 |
-
|
| 86 |
-
|
| 87 |
-
|
| 88 |
-
|
| 89 |
-
|
| 90 |
|
| 91 |
-
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
import os
|
| 2 |
import streamlit as st
|
| 3 |
import tempfile
|
|
|
|
| 4 |
from datetime import datetime
|
| 5 |
from pathlib import Path
|
| 6 |
+
from pymongo import MongoClient
|
| 7 |
from urllib.parse import quote_plus
|
| 8 |
+
from document_chunker import DocumentChunker
|
| 9 |
|
| 10 |
# === MongoDB connection via Hugging Face secrets ===
|
| 11 |
user = quote_plus(os.getenv("MONGO_USER"))
|
| 12 |
password = quote_plus(os.getenv("MONGO_PASS"))
|
| 13 |
cluster = os.getenv("MONGO_CLUSTER")
|
| 14 |
db_name = os.environ.get("MONGO_DB", "grant_docs")
|
| 15 |
+
|
| 16 |
mongo_uri = f"mongodb+srv://{user}:{password}@{cluster}/{db_name}?retryWrites=true&w=majority&tls=true&tlsAllowInvalidCertificates=true"
|
| 17 |
client = MongoClient(mongo_uri, tls=True, tlsAllowInvalidCertificates=True, serverSelectionTimeoutMS=20000)
|
| 18 |
db = client[db_name]
|
|
|
|
| 23 |
|
| 24 |
with st.sidebar:
|
| 25 |
st.header("Settings")
|
|
|
|
|
|
|
| 26 |
try:
|
| 27 |
existing_collections = db.list_collection_names()
|
| 28 |
existing_collections.append("Create New Collection")
|
| 29 |
+
default_index = existing_collections.index("doc_chunks_cat") if "doc_chunks_cat" in existing_collections else 0
|
| 30 |
+
selected_collection = st.selectbox("Choose MongoDB Collection", existing_collections, index=default_index)
|
| 31 |
except Exception as e:
|
| 32 |
st.error(f"Failed to list collections: {e}")
|
| 33 |
selected_collection = "doc_chunks_cat"
|
|
|
|
| 42 |
|
| 43 |
uploaded_file = st.file_uploader("Upload a DOCX, TXT, or PDF file", type=["docx", "txt", "pdf"])
|
| 44 |
|
| 45 |
+
# === Store session state after upload ===
|
| 46 |
+
if uploaded_file and "ready_to_process" not in st.session_state:
|
| 47 |
temp_path = Path(tempfile.gettempdir()) / uploaded_file.name
|
| 48 |
with open(temp_path, "wb") as f:
|
| 49 |
f.write(uploaded_file.getbuffer())
|
| 50 |
|
| 51 |
+
st.session_state["uploaded_file_name"] = uploaded_file.name
|
| 52 |
+
st.session_state["collection_name"] = selected_collection
|
| 53 |
+
st.session_state["is_grant_app"] = is_grant_app
|
| 54 |
+
st.session_state["temp_path"] = str(temp_path)
|
| 55 |
+
st.session_state["ready_to_process"] = True
|
| 56 |
+
st.rerun()
|
| 57 |
+
|
| 58 |
+
# === Process document ===
|
| 59 |
+
if st.session_state.get("ready_to_process"):
|
| 60 |
+
file_name = st.session_state["uploaded_file_name"]
|
| 61 |
+
collection_name = st.session_state["collection_name"]
|
| 62 |
+
is_grant_app = st.session_state["is_grant_app"]
|
| 63 |
+
temp_path = st.session_state["temp_path"]
|
| 64 |
|
| 65 |
+
st.success(f"Uploaded `{file_name}`")
|
| 66 |
+
collection = db[collection_name]
|
| 67 |
|
| 68 |
+
if collection.find_one({"metadata.title": file_name}):
|
| 69 |
st.warning("⚠️ This file already exists in the collection. Skipping...")
|
| 70 |
else:
|
| 71 |
st.write("⏳ Processing with DocumentChunker...")
|
| 72 |
chunker = DocumentChunker()
|
| 73 |
+
chunks = chunker.process_document(temp_path)
|
| 74 |
|
| 75 |
if chunks:
|
| 76 |
for chunk in chunks:
|
| 77 |
chunk['metadata'].update({
|
| 78 |
+
"title": file_name,
|
| 79 |
+
"uploaded_at": datetime.now().isoformat(),
|
| 80 |
"is_grant_app": is_grant_app,
|
| 81 |
})
|
| 82 |
collection.insert_one(chunk)
|
| 83 |
|
| 84 |
+
st.success(f"✅ {len(chunks)} chunks inserted into `{collection_name}`")
|
| 85 |
|
| 86 |
for i, c in enumerate(chunks[:3]):
|
| 87 |
st.subheader(f"Chunk {i+1}: {c['metadata'].get('header') or 'No Header'}")
|
|
|
|
| 91 |
|
| 92 |
if len(chunks) > 3:
|
| 93 |
st.info(f"... and {len(chunks)-3} more chunks processed.")
|
|
|
|
| 94 |
else:
|
| 95 |
st.warning("⚠️ No chunks were generated.")
|
| 96 |
|
| 97 |
+
# Clean up
|
| 98 |
+
try:
|
| 99 |
+
os.remove(temp_path)
|
| 100 |
+
except Exception as e:
|
| 101 |
+
st.warning(f"⚠️ Could not delete temp file: {e}")
|
| 102 |
|
| 103 |
+
# Reset session
|
| 104 |
+
for key in ["uploaded_file_name", "collection_name", "is_grant_app", "temp_path", "ready_to_process"]:
|
| 105 |
+
st.session_state.pop(key, None)
|
| 106 |
+
st.rerun()
|