Spaces:
Runtime error
Runtime error
File size: 5,631 Bytes
180a0cc 3073f86 180a0cc 3073f86 180a0cc 3073f86 180a0cc 3073f86 89d3f24 180a0cc 89d3f24 180a0cc 89d3f24 180a0cc 3073f86 180a0cc 3073f86 180a0cc 89d3f24 3073f86 67b4e55 3073f86 89d3f24 67b4e55 3073f86 89d3f24 180a0cc 3073f86 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 |
import streamlit as st
from PyPDF2 import PdfReader
from anthropic import Anthropic
from prompts import DIFFERENTIATE_PROMPT
def extract_differences(input_text):
input_text = input_text.strip()
qa_pairs = input_text.split('</difference>')
# Initialize an empty list to hold the parsed dictionary objects
parsed_data = []
# Iterate over each QA pair
for pair in qa_pairs:
# Check if the pair has both question and answer (ignoring the last one)
if '<text1_section>' in pair and '</text1_section>' in pair and '<text2_section>' in pair and '</text2_section>' in pair and '<explanation>' in pair and '</explanation>' in pair:
# Extract the question and answer text
text1 = pair.split('<text1_section>')[1].split('</text1_section>')[0]
text2 = pair.split('<text2_section>')[1].split('</text2_section>')[0]
explanation = pair.split('<explanation>')[1].split('</explanation>')[0]
# Create a dictionary for the current pair and append it to the list
parsed_data.append({'text1': text1.strip(), 'text2': text2.strip(), 'explanation': explanation.strip()})
return parsed_data
st.cache_data()
def make_llm_api_call(prompt):
client = Anthropic()
message = client.messages.create(
model="claude-3-haiku-20240307",
max_tokens=4096,
temperature=0,
messages=[{"role": "user", "content": [{"type": "text", "text": prompt}]}],
)
return message
def get_llm_response(extractedtext1, extractedtext2):
prompt = DIFFERENTIATE_PROMPT.format(text1=extractedtext1, text2=extractedtext2)
message = make_llm_api_call(prompt)
message_text = message.content[0].text
try:
before_differences = message_text.split("<differences>")[0]
after_differences = message_text.split("</differences>")[1]
differences_list = extract_differences(message_text.split("<differences>")[1].split("</differences>")[0].strip())
except Exception as e:
print("Error:", e)
return message_text, []
difference_content = "\n\n\n".join([f"**Text1:**\n\n{d['text1']}\n\n**Text2:**\n\n{d['text2']}\n\n**Explanation:**\n\n{d['explanation']}\n\n----------------------" for d in differences_list])
display_text = f"{before_differences}\n\n{difference_content}\n\n{after_differences}"
return display_text, differences_list
def extract_text_with_pypdf(pdf_path):
reader = PdfReader(pdf_path)
text = ""
for page in reader.pages:
text += page.extract_text() + "\n"
return text
def main():
st.set_page_config(layout="wide") # Enable wide layout
st.markdown('<div style="text-align: center;">' + '<h1>PDF Upload and Compare App</h1>' + '</div>', unsafe_allow_html=True)
# Create columns for side-by-side buttons
col1, col2 = st.columns([2, 2])
# Add upload button to left column
uploaded_file1 = col1.file_uploader("**Text 1**", type="pdf")
# Add upload button to right column
uploaded_file2 = col2.file_uploader("**Text 2**", type="pdf")
# Check if both files are uploaded
if uploaded_file1 and uploaded_file2:
# Get filenames from uploaded files
filename1 = uploaded_file1.name
filename2 = uploaded_file2.name
try:
extracted_text1 = extract_text_with_pypdf(uploaded_file1)
extracted_text2 = extract_text_with_pypdf(uploaded_file2)
with col1.expander(filename1):
st.write("\n\n".join(extracted_text1.splitlines()))
with col2.expander(filename2):
st.write("\n\n".join(extracted_text2.splitlines()))
st.success(f"Content of files **{filename1}** and **{filename2}** have been extracted successfully.")
except Exception as e:
st.error(f"Error saving files: {str(e)}")
# Add button at the bottom to run Find Differences function
if st.button("Find Differences"):
try:
display_text, parsed_data = get_llm_response(extracted_text1, extracted_text2)
display_text1 = extracted_text1
display_text2 = extracted_text2
for diff in parsed_data:
diff_text1 = diff['text1'].strip()
diff_text2 = diff['text2'].strip()
diff_text1_phrase = "\n".join([f"<span style='background-color: grey;'>{t}</span>" for t in diff_text1.splitlines()])
diff_text2_phrase = "\n".join([f"<span style='background-color: grey;'>{t}</span>" for t in diff_text2.splitlines()])
display_text1 = diff_text1_phrase.join(display_text1.split(diff_text1)) if diff_text1 in display_text1 else display_text1
display_text2 = diff_text2_phrase.join(display_text2.split(diff_text2)) if diff_text2 in display_text2 else display_text2
with col1.expander(f"{filename1} Difference Highlighted"):
st.write("\n\n".join(display_text1.splitlines()), unsafe_allow_html=True)
with col2.expander(f"{filename2} Difference Highlighted"):
st.write("\n\n".join(display_text2.splitlines()), unsafe_allow_html=True)
st.markdown(display_text)
except Exception as e:
st.error(f"Error finding differences: {str(e)}")
if __name__ == "__main__":
main() |