Upload 3 files

app.py

@@ -0,0 +1,111 @@
+import os
+import streamlit as st
+from io import BytesIO
+from tempfile import NamedTemporaryFile
+from langchain.chains import create_retrieval_chain
+from langchain.chains.combine_documents import create_stuff_documents_chain
+from langchain_core.prompts import ChatPromptTemplate
+from langchain_openai import ChatOpenAI
+from langchain_community.document_loaders import PyPDFLoader
+from langchain_core.vectorstores import InMemoryVectorStore
+from langchain_openai import OpenAIEmbeddings
+from langchain_text_splitters import RecursiveCharacterTextSplitter
+
+# Function to process PDF, run Q&A, and return results
+def process_pdf(api_key, uploaded_file, questions_path, prompt_path, display_placeholder):
+    # Set up OpenAI API key
+    os.environ["OPENAI_API_KEY"] = api_key
+
+    # Temporarily save the uploaded file to disk
+    with NamedTemporaryFile(delete=False, suffix=".pdf") as temp_pdf:
+        temp_pdf.write(uploaded_file.read())  # Write the uploaded file to the temp file
+        temp_pdf_path = temp_pdf.name
+
+    # Load the PDF document using PyPDFLoader
+    loader = PyPDFLoader(temp_pdf_path)
+    docs = loader.load()
+
+    # Split the document into smaller chunks for embedding
+    text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=200)
+    splits = text_splitter.split_documents(docs)
+
+    # Create vector store and retriever
+    vectorstore = InMemoryVectorStore.from_documents(
+        documents=splits, embedding=OpenAIEmbeddings()
+    )
+    retriever = vectorstore.as_retriever(search_kwargs={"k": 10})
+
+    # Read the system prompt from a Markdown (.md) file
+    if os.path.exists(prompt_path):
+        with open(prompt_path, "r") as file:
+            system_prompt = file.read()
+    else:
+        raise FileNotFoundError(f"The specified file was not found: {prompt_path}")
+
+    # Ensure the system prompt includes {context} for document input
+    prompt = ChatPromptTemplate.from_messages(
+        [
+            ("system", system_prompt),
+            ("human", "{input}"),
+        ]
+    )
+
+    # Create the retrieval and question-answering chains
+    llm = ChatOpenAI(model="gpt-4o")
+    question_answer_chain = create_stuff_documents_chain(llm, prompt, document_variable_name="context")
+    rag_chain = create_retrieval_chain(retriever, question_answer_chain)
+
+    # Load questions from a Markdown file
+    if os.path.exists(questions_path):
+        with open(questions_path, "r") as file:
+            questions = [line.strip() for line in file.readlines() if line.strip()]
+    else:
+        raise FileNotFoundError(f"The specified file was not found: {questions_path}")
+
+    # Generate question and answer pairs incrementally
+    qa_results = []
+    for question in questions:
+        result = rag_chain.invoke({"input": question})
+        answer = result["answer"]
+        qa_text = f"### Question: {question}\n**Answer:** {answer}\n"
+        qa_results.append(qa_text)
+        # Update the placeholder with each new Q&A pair
+        display_placeholder.markdown("\n".join(qa_results), unsafe_allow_html=True)
+
+    # Clean up the temporary file
+    os.remove(temp_pdf_path)
+
+    return qa_results
+
+# Streamlit app layout
+st.title("Climate Policy Summary Tool")
+
+# Input OpenAI API key
+api_key = st.text_input("Enter your OpenAI API key:", type="password")
+
+# File upload section for PDF
+uploaded_file = st.file_uploader("Upload a PDF document", type="pdf")
+
+# Define static paths for prompt and questions
+prompt_file_path = "summary_tool_system_prompt.md"
+questions_file_path = "summary_tool_questions.md"
+
+# When user clicks "Generate"
+if st.button("Generate") and api_key and uploaded_file:
+    # Create a placeholder to update with each Q&A
+    display_placeholder = st.empty()
+
+    with st.spinner("Processing..."):
+        try:
+            results = process_pdf(api_key, uploaded_file, questions_file_path, prompt_file_path, display_placeholder)
+            
+            # Allow the user to download the results as a Markdown file
+            markdown_text = "\n".join(results)
+            st.download_button(
+                label="Download Results as Markdown",
+                data=markdown_text,
+                file_name="qa_results.md",
+                mime="text/markdown"
+            )
+        except Exception as e:
+            st.error(f"An error occurred: {e}")

summary_tool_questions.md

@@ -0,0 +1,21 @@
+Provide a summary of information available from the plan including the city name, population, geography, major industries, per capita income, or similar summary information.
+
+To what extent does the plan address resilience, mitigation, and adaptation?
+
+Does the plan describe a scenario of climate change the planning jurisdiction will face and for which it is planning climate policies for resilience, mitigation, and adaptation? If so, please summarize any such scenario.
+
+What climate impacts does the plan identify as presenting climate risks for the planning jurisdiction? For each identified climate impact, summarize the plan’s description of: 1. The climate impact. 2. The climate risks it poses to the planning jurisdiction. 3. Areas or populations with the most climate vulnerability to the identified climate impact.
+
+What resilience options does the plan identify? For each identified resilience option, summarize the information provided in the plan regarding cost, timing, location, and other important attributes.
+
+What mitigation options does the plan identify? For each identified mitigation option, summarize the information provided in the plan regarding cost, timing, location, and other important attributes.
+
+What adaptation options does the plan identify? For each identified adaptation option, summarize: 1. The climate threat it is addressing. 2. Information provided in the plan regarding cost, timing, location, and other important attributes.
+
+What climate policies does the plan identify to advance climate justice? For each climate policy, summarize: 1. The climate threat it is addressing. 2. Areas or populations with the most climate vulnerability to the identified climate impact.
+
+Is there any climate policy identified in the plan that could be a potential maladaptation? For each identified climate policy, summarize: 1. The climate threat it is addressing. 2. The resilience, mitigation, or adaptation option that the climate policy proposes.
+
+Summarize the plan's description of how the planning jurisdiction will monitor and evaluate progress on the implementation of resilience options, mitigation options, and adaptation options proposed in the plan.
+
+Generate a table that summarizes the following features from the document: Location/City Name, Population, Threats Identified, Measures Identified

summary_tool_system_prompt.md

@@ -0,0 +1,30 @@
+You are a researcher specializing in extracting climate-related information from climate adaptation and resilience plans. You are provided with a document (referred to as the "plan") from a state or city in the United States (referred to as the "planning jurisdiction") that has adopted policies addressing climate change, which may pertain to resilience, mitigation, and adaptation.
+
+Your task is to analyze the plan and answer a consistent set of questions based strictly on the content provided. Your responses must:
+
+- **Base all answers strictly on the document.**
+- **Include direct quotations formatted with citations** in the format *(Chapter name, Section header, Page number etc. if applicable)*.
+- **Format all responses using Markdown syntax.**
+
+### **Definitions**
+
+Refer to the following definitions when answering the questions:
+
+- **Climate Change:** A long-term shift in weather patterns and temperatures, primarily caused by human activities emitting greenhouse gases (GHGs).
+- **Greenhouse Gases (GHGs):** Atmospheric gases like CO₂, CH₄, N₂O that absorb and emit radiation, leading to the greenhouse effect.
+- **Anthropogenic Emissions:** Emissions of GHGs resulting from human activities such as burning fossil fuels and deforestation.
+- **Climate Impacts:** Consequences of climate-related hazards on natural and human systems, affecting lives, ecosystems, economies, and infrastructure.
+- **Climate Risk:** Potential negative consequences from climate impacts, resulting from the interaction of hazard, exposure, and vulnerability.
+- **Climate Vulnerability:** The degree to which a system is susceptible to harm from climate change and its ability to adapt.
+- **Climate Policies:** Strategies and measures adopted to implement resilience, mitigation, and adaptation options.
+- **Resilience:** The ability of systems to cope with climate hazards by maintaining essential functions and adapting to changes.
+- **Resilience Options:** Strategies to build resilience through policy changes, infrastructure improvements, planning, etc.
+- **Mitigation:** Efforts to reduce or prevent emission of GHGs.
+- **Mitigation Options:** Technologies or practices that contribute to mitigation, like renewable energy or waste minimization.
+- **Adaptation:** Adjusting systems to actual or expected climate changes to minimize harm or exploit beneficial opportunities.
+- **Adaptation Options:** Strategies addressing climate change adaptation, including structural, institutional, ecological, and behavioral measures.
+- **Climate Justice:** Ensuring equitable sharing of the burdens and benefits of climate change impacts.
+- **Maladaptation:** Actions that may increase vulnerability to climate change or diminish resilience.
+- **Scenario:** A plausible description of how the future may develop based on a coherent set of assumptions.
+
+{context}