Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -1,6 +1,7 @@
|
|
| 1 |
import re
|
| 2 |
import json
|
| 3 |
import os
|
|
|
|
| 4 |
from datetime import datetime
|
| 5 |
from io import BytesIO
|
| 6 |
|
|
@@ -19,6 +20,7 @@ def configure_gemini(api_key):
|
|
| 19 |
|
| 20 |
# Read PDF content from a file-like object (from Streamlit uploader)
|
| 21 |
def read_pdf(file_obj):
|
|
|
|
| 22 |
text_content = []
|
| 23 |
pdf_reader = pypdf.PdfReader(file_obj)
|
| 24 |
for page in pdf_reader.pages:
|
|
@@ -51,6 +53,7 @@ Return ONLY valid JSON with this structure:
|
|
| 51 |
]
|
| 52 |
}"""
|
| 53 |
response = model.generate_content([prompt, text])
|
|
|
|
| 54 |
return response.text
|
| 55 |
|
| 56 |
# Generate financial report from aggregated JSON transactions and chosen sections
|
|
@@ -62,6 +65,7 @@ Generate a detailed financial report that includes the following sections: {', '
|
|
| 62 |
Ensure that each section is clearly formatted with headings and includes insights and summaries.
|
| 63 |
Return the complete report as plain text."""
|
| 64 |
response = model.generate_content([prompt])
|
|
|
|
| 65 |
return response.text
|
| 66 |
|
| 67 |
# Create a PDF file from the report text
|
|
@@ -80,11 +84,16 @@ def create_pdf_report(report_text):
|
|
| 80 |
return pdf_buffer
|
| 81 |
|
| 82 |
def main():
|
| 83 |
-
st.title("Quantitlytix AI
|
| 84 |
st.markdown("*Bank Statement Parser & Financial Report Generator*")
|
|
|
|
| 85 |
# Allow multiple PDF uploads
|
| 86 |
uploaded_files = st.file_uploader("Upload PDF bank statements", type="pdf", accept_multiple_files=True)
|
| 87 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 88 |
if uploaded_files:
|
| 89 |
try:
|
| 90 |
# Initialize the Gemini model
|
|
@@ -94,10 +103,19 @@ def main():
|
|
| 94 |
for uploaded_file in uploaded_files:
|
| 95 |
# Read PDF text directly from the uploaded file
|
| 96 |
pdf_text = read_pdf(uploaded_file)
|
|
|
|
|
|
|
|
|
|
|
|
|
| 97 |
with st.spinner(f"Processing {uploaded_file.name}..."):
|
| 98 |
json_response = process_with_gemini(model, pdf_text)
|
| 99 |
# Extract valid JSON from the response
|
| 100 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 101 |
json_str = json_str.replace('```json', '').replace('```', '')
|
| 102 |
data = json.loads(json_str)
|
| 103 |
transactions = data.get('transactions', [])
|
|
|
|
| 1 |
import re
|
| 2 |
import json
|
| 3 |
import os
|
| 4 |
+
import time
|
| 5 |
from datetime import datetime
|
| 6 |
from io import BytesIO
|
| 7 |
|
|
|
|
| 20 |
|
| 21 |
# Read PDF content from a file-like object (from Streamlit uploader)
|
| 22 |
def read_pdf(file_obj):
|
| 23 |
+
file_obj.seek(0) # Ensure the file pointer is at the start
|
| 24 |
text_content = []
|
| 25 |
pdf_reader = pypdf.PdfReader(file_obj)
|
| 26 |
for page in pdf_reader.pages:
|
|
|
|
| 53 |
]
|
| 54 |
}"""
|
| 55 |
response = model.generate_content([prompt, text])
|
| 56 |
+
time.sleep(7) # Sleep for 7 seconds to work around rate limit
|
| 57 |
return response.text
|
| 58 |
|
| 59 |
# Generate financial report from aggregated JSON transactions and chosen sections
|
|
|
|
| 65 |
Ensure that each section is clearly formatted with headings and includes insights and summaries.
|
| 66 |
Return the complete report as plain text."""
|
| 67 |
response = model.generate_content([prompt])
|
| 68 |
+
time.sleep(7) # Sleep for 7 seconds to work around rate limit
|
| 69 |
return response.text
|
| 70 |
|
| 71 |
# Create a PDF file from the report text
|
|
|
|
| 84 |
return pdf_buffer
|
| 85 |
|
| 86 |
def main():
|
| 87 |
+
st.title("Quantitlytix AI")
|
| 88 |
st.markdown("*Bank Statement Parser & Financial Report Generator*")
|
| 89 |
+
|
| 90 |
# Allow multiple PDF uploads
|
| 91 |
uploaded_files = st.file_uploader("Upload PDF bank statements", type="pdf", accept_multiple_files=True)
|
| 92 |
|
| 93 |
+
# Debug: Show number of files uploaded
|
| 94 |
+
if uploaded_files:
|
| 95 |
+
st.write(f"{len(uploaded_files)} file(s) uploaded.")
|
| 96 |
+
|
| 97 |
if uploaded_files:
|
| 98 |
try:
|
| 99 |
# Initialize the Gemini model
|
|
|
|
| 103 |
for uploaded_file in uploaded_files:
|
| 104 |
# Read PDF text directly from the uploaded file
|
| 105 |
pdf_text = read_pdf(uploaded_file)
|
| 106 |
+
if not pdf_text:
|
| 107 |
+
st.warning(f"No text found in {uploaded_file.name}.")
|
| 108 |
+
continue
|
| 109 |
+
|
| 110 |
with st.spinner(f"Processing {uploaded_file.name}..."):
|
| 111 |
json_response = process_with_gemini(model, pdf_text)
|
| 112 |
# Extract valid JSON from the response
|
| 113 |
+
start_idx = json_response.find('{')
|
| 114 |
+
end_idx = json_response.rfind('}') + 1
|
| 115 |
+
if start_idx == -1 or end_idx == -1:
|
| 116 |
+
st.warning(f"Invalid JSON response for {uploaded_file.name}.")
|
| 117 |
+
continue
|
| 118 |
+
json_str = json_response[start_idx:end_idx]
|
| 119 |
json_str = json_str.replace('```json', '').replace('```', '')
|
| 120 |
data = json.loads(json_str)
|
| 121 |
transactions = data.get('transactions', [])
|