Create multilingual_word_limit_issue_v3.py

lab/multilingual_word_limit_issue_v3.py

@@ -0,0 +1,331 @@
+import streamlit as st
+from graph import EssayWriter, RouteQuery, GraphState
+from language_options import language_options
+from crew import *
+import os
+import re
+import traceback
+import base64
+
+# Install Graphviz if not found
+if os.system("which dot") != 0:
+    os.system("apt-get update && apt-get install -y graphviz")
+
+st.markdown(
+    """
+    <h1 style="text-align: center; white-space: nowrap; font-size: 2.5em;">
+        Multi-Agent Essay Writing Assistant
+    </h1>
+    """,
+    unsafe_allow_html=True
+)
+
+# Ensure session state variables are initialized properly
+if "messages" not in st.session_state:
+    st.session_state["messages"] = [{"role": "assistant", "content": "Hello! How can I assist you today?"}]
+
+if "app" not in st.session_state:
+    st.session_state["app"] = None
+
+if "chat_active" not in st.session_state:
+    st.session_state["chat_active"] = True
+
+# Sidebar with essay settings and user-defined length
+# Sidebar with essay settings and user-defined length
+with st.sidebar:
+    st.subheader("📝 Note:")
+    st.info(
+        "\n\n 1. This app uses the 'gpt-4o-mini-2024-07-18' model."
+        "\n\n 2. Writing essays may take some time, approximately 1-2 minutes."
+    )
+
+    # API Key Retrieval
+    openai_key = st.secrets.get("OPENAI_API_KEY", "")
+
+    st.divider()
+
+    # User-defined essay length selection
+    st.subheader("⚙️🛠️ Configure Essay Settings:")
+    essay_length = st.number_input(
+        "Select Essay Length (words):",
+        min_value=150,
+        max_value=500,
+        value=250,
+        step=50
+    )
+
+    #st.divider()
+
+    #Language Selection
+    #st.subheader("🌍 Select Language:")    
+    selected_language = st.selectbox("Choose Language:", sorted(language_options.keys()), index=list(language_options.keys()).index("English"))
+
+    st.divider()
+
+    # Reference section
+    st.subheader("📖 References:")
+    st.markdown(
+        "[1. Multi-Agent System with CrewAI and LangChain](https://discuss.streamlit.io/t/new-project-i-have-build-a-multi-agent-system-with-crewai-and-langchain/84002)",
+        unsafe_allow_html=True
+    )
+
+
+# Initialize agents function
+def initialize_agents():
+    if not openai_key:
+        st.error("⚠️ OpenAI API key is missing! Please provide a valid key through Hugging Face Secrets.")
+        st.session_state["chat_active"] = True
+        return None
+
+    os.environ["OPENAI_API_KEY"] = openai_key
+    try:
+        # Prevent re-initialization
+        if "app" in st.session_state and st.session_state["app"] is not None:
+            return st.session_state["app"]
+        
+        # Initialize the full EssayWriter instance
+        essay_writer = EssayWriter()  # Store the full instance
+        st.session_state["app"] = essay_writer  # Now contains `graph`
+        st.session_state["chat_active"] = False  # Enable chat after successful initialization
+
+        return essay_writer
+    except Exception as e:
+        st.error(f"❌ Error initializing agents: {e}")
+        st.session_state["chat_active"] = True
+        return None
+
+
+# Automatically initialize agents on app load
+if st.session_state["app"] is None:
+    st.session_state["app"] = initialize_agents()
+
+if st.session_state["app"] is None:
+    st.error("⚠️ Failed to initialize agents. Please check your API key and restart the app.")
+
+app = st.session_state["app"]
+
+# Function to invoke the agent and generate a response
+def enforce_word_limit(text, limit):
+    """Enforces strict word limit by truncating extra words."""
+    words = re.findall(r'\b\w+\b', text)
+    return ' '.join(words[:limit]) if len(words) > limit else text
+
+def detect_unexpected_english(text, selected_language):
+    """Detects unintended English words when another language is selected."""
+    if selected_language != "English":
+        english_words = re.findall(r'\b(?:is|the|and|or|in|on|at|to|with|for|of|by|it|that|this|was|he|she|they|we|you|I)\b', text)
+        return len(english_words) > 5  # Allow a small tolerance
+
+def generate_response(topic, length, selected_language):
+    if not app or not hasattr(app, "graph"):
+        st.error("Agents are not initialized. Please check the system or restart the app.")
+        return {"response": "Error: Agents not initialized."}
+
+    # Dynamically adjust structure based on length
+    if length <= 250:
+        intro_limit, body_limit, conclusion_limit = length // 5, length // 2, length // 5
+        num_sections = 2  # Shorter essays should have fewer sections
+    elif length <= 350:
+        intro_limit, body_limit, conclusion_limit = length // 6, length // 1.8, length // 6
+        num_sections = 3
+    else:
+        intro_limit, body_limit, conclusion_limit = length // 7, length // 1.7, length // 7
+        num_sections = 4
+
+    # Optimized Structured Prompt
+    refined_prompt = f"""
+    Write a **well-structured, informative, and engaging** essay on "{topic}" **strictly in {selected_language}.**
+    
+    **Word Limit:** Exactly {length} words. **Do not exceed or fall short of this limit.**
+    **Language Rules:** Use natural linguistic style from {selected_language}. **Do not use English** unless explicitly requested.
+    
+    **Essay Structure:**
+    - **Title**: Max 10 words.
+    - **Introduction ({intro_limit} words max)**:
+      - Clearly define the topic and its significance.
+      - Provide a strong thesis statement.
+      - Preview the key points covered in the essay.
+    - **Main Body ({body_limit} words max, {num_sections} sections)**:
+      - Each section must have:
+        - A **clear subheading**.
+        - A concise topic sentence with supporting details.
+        - Relevant **examples, statistics, or historical references**.
+      - Maintain natural **flow** between sections.
+    - **Conclusion ({conclusion_limit} words max)**:
+      - Summarize key insights **without repetition**.
+      - Reinforce the thesis **based on discussion**.
+      - End with a strong **closing statement** (reflection or call to action).
+    
+    **Hard Rules:**
+    - **Use only {selected_language}**. No English unless explicitly requested.
+    - **Do not exceed {length} words.** Absolute limit.
+    - **Write concisely and avoid fluff**. No redundancy.
+    - **Merge similar ideas** to maintain smooth readability.
+    - **Ensure strict adherence to section word limits**.
+    """
+
+    # Invoke AI model with enforced word limit
+    response = app.graph.invoke(input={
+        "topic": topic,
+        "length": length,
+        "prompt": refined_prompt,
+        "language": selected_language,
+        "max_tokens": length + 10  # Small buffer for better trimming
+    })
+
+    # Strict word limit enforcement
+    essay_text = enforce_word_limit(response.get("essay", ""), length)
+
+    # Detect unintended English words in non-English essays
+    if detect_unexpected_english(essay_text, selected_language):
+        return {"response": f"⚠️ Warning: Some English words were detected in the {selected_language} essay. Try regenerating."}
+
+    return {"essay": essay_text}
+
+
+
+# Define Tabs
+tab1, tab2 = st.tabs(["📜 Essay Generation", "📊 Workflow Viz"])
+
+# 📜 Tab 1: Essay Generation
+with tab1:
+    # Display chat messages from the session
+    if "messages" not in st.session_state:
+        st.session_state["messages"] = [{"role": "assistant", "content": "Hello! How can I assist you today?"}]
+
+    for message in st.session_state["messages"]:
+        with st.chat_message(message["role"]):
+            st.markdown(message["content"], unsafe_allow_html=True)
+
+    # Input
+    topic = st.text_input("📝 Provide an essay topic:", value="Write an essay on the cultural diversity of India")
+
+    # Add spacing
+    st.write("")
+
+    # Generate button
+    if st.button("🚀 Generate Essay"):
+        if topic and topic.strip():  # Ensure it's not empty
+            # Store user message only if it's not already stored
+            if not any(msg["content"] == topic for msg in st.session_state["messages"]):
+                st.session_state["messages"].append({"role": "user", "content": topic})
+
+            with st.spinner("⏳ Generating your essay..."):
+                response = None
+                if app:
+                    response = app.write_essay({"topic": topic})
+                else:
+                    st.error("⚠️ Agents are not initialized. Please check the system or restart the app.")
+
+            # Store and display assistant response
+            if response and "essay" in response:
+                essay = response["essay"]
+
+                assistant_response = f"Here is your {essay_length}-word essay preview and the download link."
+                st.session_state["messages"].append({"role": "assistant", "content": assistant_response})
+
+                st.chat_message("assistant").markdown(assistant_response)
+
+                # Create Two-Column Layout
+                col1, col2 = st.columns(2)
+
+                with col1:
+                    st.markdown(f"### 📝 Essay Preview ({essay_length} words)")
+                    st.markdown(f"#### {essay['header']}")
+                    st.markdown(essay["entry"])
+
+                    for para in essay["paragraphs"]:
+                        st.markdown(f"**{para['sub_header']}**")
+                        st.markdown(para["paragraph"])
+
+                    st.markdown("**🖊️ Conclusion:**")
+                    st.markdown(essay["conclusion"])
+
+                with col2:
+                    st.markdown("### ✍️ Edit Your Essay:")
+
+                    # Combine all parts of the essay into one editable text field
+                    full_essay_text = f"## {essay['header']}\n\n{essay['entry']}\n\n"
+                    for para in essay["paragraphs"]:
+                        full_essay_text += f"### {para['sub_header']}\n{para['paragraph']}\n\n"
+                    full_essay_text += f"**Conclusion:**\n{essay['conclusion']}"
+
+                    # Editable text area for the user
+                    edited_essay = st.text_area("Edit Here:", value=full_essay_text, height=300)
+
+                    # Save and Download buttons
+                    save_col1, save_col2 = st.columns(2)
+
+                    with save_col1:
+                        if st.button("💾 Save as TXT"):
+                            with open("edited_essay.txt", "w", encoding="utf-8") as file:
+                                file.write(edited_essay)
+                            with open("edited_essay.txt", "rb") as file:
+                                st.download_button(label="⬇️ Download TXT", data=file, file_name="edited_essay.txt", mime="text/plain")
+
+                    with save_col2:
+                        if st.button("📄 Save as PDF"):
+                            from fpdf import FPDF
+
+                            pdf = FPDF()
+                            pdf.set_auto_page_break(auto=True, margin=15)
+                            pdf.add_page()
+                            pdf.set_font("Arial", size=12)
+
+                            for line in edited_essay.split("\n"):
+                                pdf.cell(200, 10, txt=line, ln=True, align='L')
+
+                            pdf.output("edited_essay.pdf")
+
+                            with open("edited_essay.pdf", "rb") as file:
+                                st.download_button(label="⬇️ Download PDF", data=file, file_name="edited_essay.pdf", mime="application/pdf")
+
+                # Provide download link for the original PDF 
+                pdf_name = response.get("pdf_name")
+                if pdf_name and os.path.exists(pdf_name):
+                    with open(pdf_name, "rb") as pdf_file:
+                        b64 = base64.b64encode(pdf_file.read()).decode()
+                        href = f"<a href='data:application/octet-stream;base64,{b64}' download='{pdf_name}'>📄 Click here to download the original PDF</a>"
+                        st.markdown(href, unsafe_allow_html=True)
+
+                # Save response in session state
+                st.session_state["messages"].append(
+                    {"role": "assistant", "content": f"Here is your {essay_length}-word essay preview and the download link."}
+                )
+            elif response:
+                st.markdown(response["response"])
+                st.session_state["messages"].append({"role": "assistant", "content": response["response"]})
+            else:
+                st.error("⚠️ No response received. Please try again.")
+
+
+
+# 📊 Tab 2: Workflow Visualization
+with tab2:
+    #st.subheader("📊 Multi-Agent Essay Writer Workflow Viz")
+
+    try:
+        graph_path = "/tmp/graph.png"  
+        if os.path.exists(graph_path):
+            st.image(graph_path, caption="Multi-Agent Essay Writer Workflow Visualization", use_container_width=True)
+        else:
+            st.warning("⚠️ Workflow graph not found. Please run `graph.py` to regenerate `graph.png`.")
+
+    except Exception as e:
+        st.error("❌ An error occurred while generating the workflow visualization.")
+        st.text_area("Error Details:", traceback.format_exc(), height=500)
+
+
+# Acknowledgement Section
+st.markdown(
+    """
+    <div style="text-align: center; font-size: 14px; color: #555; padding-top: 200px; margin-top: 200px;">
+        <strong>Acknowledgement:</strong> This app is based on Mesut Duman's work: 
+        <a href="https://github.com/mesutdmn/Autonomous-Multi-Agent-Systems-with-CrewAI-Essay-Writer/tree/main" 
+           target="_blank" style="color: #007BFF; text-decoration: none;">
+           CrewAI Essay Writer
+        </a>
+    </div>
+    """,
+    unsafe_allow_html=True,
+)