Create app.py

app.py

@@ -0,0 +1,520 @@
+import streamlit as st
+from azure.cosmos import CosmosClient, exceptions
+import os
+import pandas as pd
+import traceback
+import shutil
+from github import Github
+from git import Repo
+from datetime import datetime
+import base64
+import json
+import uuid  # 🎲 For generating unique IDs
+from urllib.parse import quote  # 🔗 For encoding URLs
+from gradio_client import Client  # 🌐 For connecting to Gradio apps
+
+# 🎉 Welcome to our fun-filled Cosmos DB and GitHub Integration app!
+st.set_page_config(layout="wide")
+
+# 🌌 Cosmos DB configuration
+ENDPOINT = "https://acae-afd.documents.azure.com:443/"
+DATABASE_NAME = os.environ.get("COSMOS_DATABASE_NAME")
+CONTAINER_NAME = os.environ.get("COSMOS_CONTAINER_NAME")
+Key = os.environ.get("Key")  # 🔑 Don't forget your key!
+
+# 🏠 Your local app URL (Change this to your app's URL)
+LOCAL_APP_URL = "http://localhost:8501"
+
+# 🐙 GitHub configuration
+def download_github_repo(url, local_path):
+    # 🚚 Let's download that GitHub repo!
+    if os.path.exists(local_path):
+        shutil.rmtree(local_path)
+    Repo.clone_from(url, local_path)
+
+def create_zip_file(source_dir, output_filename):
+    # 📦 Zipping up files like a pro!
+    shutil.make_archive(output_filename, 'zip', source_dir)
+
+def create_repo(g, repo_name):
+    # 🛠️ Creating a new GitHub repo. Magic!
+    user = g.get_user()
+    return user.create_repo(repo_name)
+
+def push_to_github(local_path, repo, github_token):
+    # 🚀 Pushing code to GitHub. Hold on tight!
+    repo_url = f"https://{github_token}@github.com/{repo.full_name}.git"
+    local_repo = Repo(local_path)
+    
+    if 'origin' in [remote.name for remote in local_repo.remotes]:
+        origin = local_repo.remote('origin')
+        origin.set_url(repo_url)
+    else:
+        origin = local_repo.create_remote('origin', repo_url)
+    
+    if not local_repo.heads:
+        local_repo.git.checkout('-b', 'main')
+        current_branch = 'main'
+    else:
+        current_branch = local_repo.active_branch.name
+    
+    local_repo.git.add(A=True)
+    
+    if local_repo.is_dirty():
+        local_repo.git.commit('-m', 'Initial commit')
+    
+    origin.push(refspec=f'{current_branch}:{current_branch}')
+
+def get_base64_download_link(file_path, file_name):
+    # 🧙‍♂️ Generating a magical download link!
+    with open(file_path, "rb") as file:
+        contents = file.read()
+    base64_encoded = base64.b64encode(contents).decode()
+    return f'<a href="data:application/zip;base64,{base64_encoded}" download="{file_name}">⬇️ Download {file_name}</a>'
+
+
+# 🧭 New functions for dynamic sidebar navigation
+def get_databases(client):
+    # 📚 Fetching list of databases. So many options!
+    return [db['id'] for db in client.list_databases()]
+
+def get_containers(database):
+    # 📂 Getting containers. Containers within containers!
+    return [container['id'] for container in database.list_containers()]
+
+def get_documents(container, limit=None):
+    # 📝 Retrieving documents. Shhh, don't tell anyone!
+    query = "SELECT * FROM c ORDER BY c._ts DESC"
+    items = list(container.query_items(query=query, enable_cross_partition_query=True, max_item_count=limit))
+    return items
+
+
+# 🌟 Cosmos DB functions
+def insert_record(container, record):
+    try:
+        container.create_item(body=record)
+        return True, "Record inserted successfully! 🎉"
+    except exceptions.CosmosHttpResponseError as e:
+        return False, f"HTTP error occurred: {str(e)} 🚨"
+    except Exception as e:
+        return False, f"An unexpected error occurred: {str(e)} 😱"
+
+def update_record(container, updated_record):
+    try:
+        container.upsert_item(body=updated_record)
+        return True, f"Record with id {updated_record['id']} successfully updated. 🛠️"
+    except exceptions.CosmosHttpResponseError as e:
+        return False, f"HTTP error occurred: {str(e)} 🚨"
+    except Exception as e:
+        return False, f"An unexpected error occurred: {traceback.format_exc()} 😱"
+
+def delete_record(container, name, id):
+    try:
+        container.delete_item(item=id, partition_key=id)
+        return True, f"Successfully deleted record with name: {name} and id: {id} 🗑️"
+    except exceptions.CosmosResourceNotFoundError:
+        return False, f"Record with id {id} not found. It may have been already deleted. 🕵️‍♂️"
+    except exceptions.CosmosHttpResponseError as e:
+        return False, f"HTTP error occurred: {str(e)} 🚨"
+    except Exception as e:
+        return False, f"An unexpected error occurred: {traceback.format_exc()} 😱"
+
+# 🎲 Function to generate a unique UUID
+def generate_unique_id():
+    # 🧙‍♂️ Generating a unique UUID!
+    return str(uuid.uuid4())
+
+# 📦 Function to archive current container
+def archive_current_container(database_name, container_name, client):
+    try:
+        base_dir = "./cosmos_archive_current_container"
+        if os.path.exists(base_dir):
+            shutil.rmtree(base_dir)
+        os.makedirs(base_dir)
+        
+        db_client = client.get_database_client(database_name)
+        container_client = db_client.get_container_client(container_name)
+        items = list(container_client.read_all_items())
+        
+        container_dir = os.path.join(base_dir, container_name)
+        os.makedirs(container_dir)
+        
+        for item in items:
+            item_id = item.get('id', f"unknown_{datetime.now().strftime('%Y%m%d%H%M%S')}")
+            with open(os.path.join(container_dir, f"{item_id}.json"), 'w') as f:
+                json.dump(item, f, indent=2)
+        
+        archive_name = f"{container_name}_archive_{datetime.now().strftime('%Y%m%d%H%M%S')}"
+        shutil.make_archive(archive_name, 'zip', base_dir)
+        
+        return get_base64_download_link(f"{archive_name}.zip", f"{archive_name}.zip")
+    except Exception as e:
+        return f"An error occurred while archiving data: {str(e)} 😢"
+
+# 🔍 Search Glossary function
+def search_glossary(query):
+    # 🕵️‍♂️ Searching the glossary for: query
+    all_results = ""
+    st.markdown(f"- {query}")
+    
+    # 🔍 Run 1 - ArXiv RAG researcher expert ~-<>-~ Paper Summary & Ask LLM
+    client = Client("awacke1/Arxiv-Paper-Search-And-QA-RAG-Pattern")
+    response2 = client.predict(
+        query,	# str  in 'parameter_13' Textbox component
+        "google/gemma-7b-it",	# LLM Model Dropdown component
+        True,	# Stream output Checkbox component
+        api_name="/ask_llm"
+    )
+    st.write('🔍 Run of Multi-Agent System Paper Summary Spec is Complete')
+    st.markdown(response2)
+    
+    # ArXiv searcher ~-<>-~ Paper References - Update with RAG
+    client = Client("awacke1/Arxiv-Paper-Search-And-QA-RAG-Pattern")
+    response1 = client.predict(
+        query,	
+        10,	
+        "Semantic Search - up to 10 Mar 2024",	# Search Source Dropdown component
+        "mistralai/Mixtral-8x7B-Instruct-v0.1",	# LLM Model Dropdown component
+        api_name="/update_with_rag_md"
+    )
+    st.write('🔍 Run of Multi-Agent System Paper References is Complete')
+    responseall = response2 + response1[0] + response1[1]
+    st.markdown(responseall)
+    return responseall
+
+# 🎈 Let's modify the main app to be more fun!
+def main():
+    st.title("🐙Git🌌Cosmos💫 - Azure Cosmos DB and Github Agent")
+
+    # 🚦 Initialize session state
+    if 'logged_in' not in st.session_state:
+        st.session_state.logged_in = False
+    if 'selected_records' not in st.session_state:
+        st.session_state.selected_records = []
+    if 'client' not in st.session_state:
+        st.session_state.client = None
+    if 'selected_database' not in st.session_state:
+        st.session_state.selected_database = None
+    if 'selected_container' not in st.session_state:
+        st.session_state.selected_container = None
+    if 'selected_document_id' not in st.session_state:
+        st.session_state.selected_document_id = None
+    if 'current_index' not in st.session_state:
+        st.session_state.current_index = 0
+    if 'cloned_doc' not in st.session_state:
+        st.session_state.cloned_doc = None
+
+    # 🔍 Check for query parameters
+    query_params = st.experimental_get_query_params()
+    if 'q' in query_params:
+        query = query_params['q'][0]  # Get the first value of 'q'
+        # 🕵️‍♂️ We have a query! Let's search the glossary!
+        search_glossary(query)
+        st.stop()  # Stop further execution
+
+    # 🔐 Automatic Login
+    if Key:
+        st.session_state.primary_key = Key
+        st.session_state.logged_in = True
+    else:
+        st.error("Cosmos DB Key is not set in environment variables. 🔑❌")
+        return  # Can't proceed without a key
+
+    if st.session_state.logged_in:
+        # 🌌 Initialize Cosmos DB client
+        try:
+            if st.session_state.client is None:
+                st.session_state.client = CosmosClient(ENDPOINT, credential=st.session_state.primary_key)
+            
+            # 🗄️ Sidebar for database, container, and document selection
+            st.sidebar.title("🐙Git🌌Cosmos💫🗄️Navigator")
+            
+            databases = get_databases(st.session_state.client)
+            selected_db = st.sidebar.selectbox("🗃️ Select Database", databases)
+            
+            if selected_db != st.session_state.selected_database:
+                st.session_state.selected_database = selected_db
+                st.session_state.selected_container = None
+                st.session_state.selected_document_id = None
+                st.session_state.current_index = 0
+                st.rerun()
+            
+            if st.session_state.selected_database:
+                database = st.session_state.client.get_database_client(st.session_state.selected_database)
+                containers = get_containers(database)
+                selected_container = st.sidebar.selectbox("📁 Select Container", containers)
+                
+                if selected_container != st.session_state.selected_container:
+                    st.session_state.selected_container = selected_container
+                    st.session_state.selected_document_id = None
+                    st.session_state.current_index = 0
+                    st.rerun()
+                
+                if st.session_state.selected_container:
+                    container = database.get_container_client(st.session_state.selected_container)
+                    
+                    # 📦 Add Export button
+                    if st.button("📦 Export Container Data"):
+                        download_link = archive_current_container(st.session_state.selected_database, st.session_state.selected_container, st.session_state.client)
+                        if download_link.startswith('<a'):
+                            st.markdown(download_link, unsafe_allow_html=True)
+                        else:
+                            st.error(download_link)
+                    
+                    # Fetch documents
+                    documents = get_documents(container)
+                    total_docs = len(documents)
+                    
+                    if total_docs > 5:
+                        documents_to_display = documents[:5]
+                        st.info("Showing top 5 most recent documents.")
+                    else:
+                        documents_to_display = documents
+                        st.info(f"Showing all {len(documents_to_display)} documents.")
+                    
+                    if documents_to_display:
+                        # 🎨 Add Viewer/Editor selection
+                        view_options = ['Show as Markdown', 'Show as Code Editor', 'Show as Edit and Save', 'Clone Document', 'New Record']
+                        selected_view = st.selectbox("Select Viewer/Editor", view_options, index=2)
+                        
+                        if selected_view == 'Show as Markdown':
+                            # 🖌️ Show each record as Markdown with navigation
+                            total_docs = len(documents)
+                            doc = documents[st.session_state.current_index]
+                            st.markdown(f"#### Document ID: {doc.get('id', '')}")
+
+                            # 🕵️‍♂️ Let's extract values from the JSON that have at least one space
+                            values_with_space = []
+                            def extract_values(obj):
+                                if isinstance(obj, dict):
+                                    for k, v in obj.items():
+                                        extract_values(v)
+                                elif isinstance(obj, list):
+                                    for item in obj:
+                                        extract_values(item)
+                                elif isinstance(obj, str):
+                                    if ' ' in obj:
+                                        values_with_space.append(obj)
+
+                            extract_values(doc)
+
+                            # 🔗 Let's create a list of links for these values
+                            search_urls = {
+                                "🚀🌌ArXiv": lambda k: f"/?q={quote(k)}",
+                                "🃏Analyst": lambda k: f"/?q={quote(k)}-{quote('PromptPrefix')}",
+                                "📚PyCoder": lambda k: f"/?q={quote(k)}-{quote('PromptPrefix2')}",
+                                "🔬JSCoder": lambda k: f"/?q={quote(k)}-{quote('PromptPrefix3')}",
+                                "🏠": lambda k: f"{LOCAL_APP_URL}/?q={quote(k)}",
+                                "📖": lambda k: f"https://en.wikipedia.org/wiki/{quote(k)}",
+                                "🔍": lambda k: f"https://www.google.com/search?q={quote(k)}",
+                                "▶️": lambda k: f"https://www.youtube.com/results?search_query={quote(k)}",
+                                "🔎": lambda k: f"https://www.bing.com/search?q={quote(k)}",
+                                "🎥": lambda k: f"https://www.youtube.com/results?search_query={quote(k)}",
+                                "🐦": lambda k: f"https://twitter.com/search?q={quote(k)}",
+                            }
+
+                            st.markdown("#### 🔗 Links for Extracted Texts")
+                            for term in values_with_space:
+                                links_md = ' '.join([f"[{emoji}]({url(term)})" for emoji, url in search_urls.items()])
+                                st.markdown(f"**{term}** <small>{links_md}</small>", unsafe_allow_html=True)
+
+                            # Show the document content as markdown
+                            content = json.dumps(doc, indent=2)
+                            st.markdown(f"```json\n{content}\n```")
+
+                            # Navigation buttons
+                            col_prev, col_next = st.columns([1, 1])
+                            with col_prev:
+                                if st.button("⬅️ Previous", key='prev_markdown'):
+                                    if st.session_state.current_index > 0:
+                                        st.session_state.current_index -= 1
+                                        st.rerun()
+                            with col_next:
+                                if st.button("➡️ Next", key='next_markdown'):
+                                    if st.session_state.current_index < total_docs - 1:
+                                        st.session_state.current_index += 1
+                                        st.rerun()
+
+                        elif selected_view == 'Show as Code Editor':
+                            # 💻 Show each record in a code editor with navigation
+                            total_docs = len(documents)
+                            doc = documents[st.session_state.current_index]
+                            st.markdown(f"#### Document ID: {doc.get('id', '')}")
+                            doc_str = st.text_area("Edit Document", value=json.dumps(doc, indent=2), height=300, key=f'code_editor_{st.session_state.current_index}')
+                            col_prev, col_next = st.columns([1, 1])
+                            with col_prev:
+                                if st.button("⬅️ Previous", key='prev_code'):
+                                    if st.session_state.current_index > 0:
+                                        st.session_state.current_index -= 1
+                                        st.rerun()
+                            with col_next:
+                                if st.button("➡️ Next", key='next_code'):
+                                    if st.session_state.current_index < total_docs - 1:
+                                        st.session_state.current_index += 1
+                                        st.rerun()
+                            if st.button("💾 Save Changes", key=f'save_button_{st.session_state.current_index}'):
+                                try:
+                                    updated_doc = json.loads(doc_str)
+                                    success, message = update_record(container, updated_doc)
+                                    if success:
+                                        st.success(f"Document {updated_doc['id']} saved successfully.")
+                                        st.session_state.selected_document_id = updated_doc['id']
+                                        st.rerun()
+                                    else:
+                                        st.error(message)
+                                except json.JSONDecodeError as e:
+                                    st.error(f"Invalid JSON: {str(e)} 🚫")
+                        elif selected_view == 'Show as Edit and Save':
+                            # ✏️ Show as Edit and Save in columns
+                            st.markdown("#### Edit the document fields below:")
+                            
+                            # Create columns for each document
+                            num_cols = len(documents_to_display)
+                            cols = st.columns(num_cols)
+                            
+                            for idx, (col, doc) in enumerate(zip(cols, documents_to_display)):
+                                with col:
+                                    st.markdown(f"##### Document ID: {doc.get('id', '')}")
+                                    editable_id = st.text_input("ID", value=doc.get('id', ''), key=f'edit_id_{idx}')
+                                    # Remove 'id' from the document for editing other fields
+                                    editable_doc = doc.copy()
+                                    editable_doc.pop('id', None)
+                                    doc_str = st.text_area("Document Content (in JSON format)", value=json.dumps(editable_doc, indent=2), height=300, key=f'doc_str_{idx}')
+                                    if st.button("💾 Save Changes", key=f'save_button_{idx}'):
+                                        try:
+                                            updated_doc = json.loads(doc_str)
+                                            updated_doc['id'] = editable_id  # Include the possibly edited ID
+                                            success, message = update_record(container, updated_doc)
+                                            if success:
+                                                st.success(f"Document {updated_doc['id']} saved successfully.")
+                                                st.session_state.selected_document_id = updated_doc['id']
+                                                st.rerun()
+                                            else:
+                                                st.error(message)
+                                        except json.JSONDecodeError as e:
+                                            st.error(f"Invalid JSON: {str(e)} 🚫")
+                        elif selected_view == 'Clone Document':
+                            # 🧬 Clone Document per record
+                            st.markdown("#### Clone a document:")
+                            for idx, doc in enumerate(documents_to_display):
+                                st.markdown(f"##### Document ID: {doc.get('id', '')}")
+                                if st.button("📄 Clone Document", key=f'clone_button_{idx}'):
+                                    cloned_doc = doc.copy()
+                                    # Generate a unique ID
+                                    cloned_doc['id'] = generate_unique_id()
+                                    st.session_state.cloned_doc = cloned_doc
+                                    st.session_state.cloned_doc_str = json.dumps(cloned_doc, indent=2)
+                                    st.session_state.clone_mode = True
+                                    st.rerun()
+                            if st.session_state.get('clone_mode', False):
+                                st.markdown("#### Edit Cloned Document:")
+                                cloned_doc_str = st.text_area("Cloned Document Content (in JSON format)", value=st.session_state.cloned_doc_str, height=300)
+                                if st.button("💾 Save Cloned Document"):
+                                    try:
+                                        new_doc = json.loads(cloned_doc_str)
+                                        success, message = insert_record(container, new_doc)
+                                        if success:
+                                            st.success(f"Cloned document saved with id: {new_doc['id']} 🎉")
+                                            st.session_state.selected_document_id = new_doc['id']
+                                            st.session_state.clone_mode = False
+                                            st.session_state.cloned_doc = None
+                                            st.session_state.cloned_doc_str = ''
+                                            st.rerun()
+                                        else:
+                                            st.error(message)
+                                    except json.JSONDecodeError as e:
+                                        st.error(f"Invalid JSON: {str(e)} 🚫")
+                        elif selected_view == 'New Record':
+                            # 🆕 New Record
+                            st.markdown("#### Create a new document:")
+                            new_id = st.text_input("ID", value=generate_unique_id(), key='new_id')
+                            new_doc_str = st.text_area("Document Content (in JSON format)", value='{}', height=300)
+                            if st.button("➕ Create New Document"):
+                                try:
+                                    new_doc = json.loads(new_doc_str)
+                                    new_doc['id'] = new_id  # Use the provided ID
+                                    success, message = insert_record(container, new_doc)
+                                    if success:
+                                        st.success(f"New document created with id: {new_doc['id']} 🎉")
+                                        st.session_state.selected_document_id = new_doc['id']
+                                        # Switch to 'Show as Edit and Save' mode
+                                        st.rerun()
+                                    else:
+                                        st.error(message)
+                                except json.JSONDecodeError as e:
+                                    st.error(f"Invalid JSON: {str(e)} 🚫")
+                    else:
+                        st.sidebar.info("No documents found in this container. 📭")
+            
+            # 🎉 Main content area
+            st.subheader(f"📊 Container: {st.session_state.selected_container}")
+            if st.session_state.selected_container:
+                if documents_to_display:
+                    df = pd.DataFrame(documents_to_display)
+                    st.dataframe(df)
+                else:
+                    st.info("No documents to display. 🧐")
+                    
+            # 🐙 GitHub section
+            st.subheader("🐙 GitHub Operations")
+            github_token = os.environ.get("GITHUB")  # Read GitHub token from environment variable
+            source_repo = st.text_input("Source GitHub Repository URL", value="https://github.com/AaronCWacker/AIExamples-8-24-Streamlit")
+            new_repo_name = st.text_input("New Repository Name (for cloning)", value=f"AIExample-Clone-{datetime.now().strftime('%Y%m%d_%H%M%S')}")
+
+            col1, col2 = st.columns(2)
+            with col1:
+                if st.button("📥 Clone Repository"):
+                    if github_token and source_repo:
+                        try:
+                            local_path = f"./temp_repo_{datetime.now().strftime('%Y%m%d%H%M%S')}"
+                            download_github_repo(source_repo, local_path)
+                            zip_filename = f"{new_repo_name}.zip"
+                            create_zip_file(local_path, zip_filename[:-4])
+                            st.markdown(get_base64_download_link(zip_filename, zip_filename), unsafe_allow_html=True)
+                            st.success("Repository cloned successfully! 🎉")
+                        except Exception as e:
+                            st.error(f"An error occurred: {str(e)} 😢")
+                        finally:
+                            if os.path.exists(local_path):
+                                shutil.rmtree(local_path)
+                            if os.path.exists(zip_filename):
+                                os.remove(zip_filename)
+                    else:
+                        st.error("Please ensure GitHub token is set in environment variables and source repository URL is provided. 🔑❓")
+        
+            with col2:
+                if st.button("📤 Push to New Repository"):
+                    if github_token and source_repo:
+                        try:
+                            g = Github(github_token)
+                            new_repo = create_repo(g, new_repo_name)
+                            local_path = f"./temp_repo_{datetime.now().strftime('%Y%m%d%H%M%S')}"
+                            download_github_repo(source_repo, local_path)
+                            push_to_github(local_path, new_repo, github_token)
+                            st.success(f"Repository pushed successfully to {new_repo.html_url} 🚀")
+                        except Exception as e:
+                            st.error(f"An error occurred: {str(e)} 😢")
+                        finally:
+                            if os.path.exists(local_path):
+                                shutil.rmtree(local_path)
+                        else:
+                            st.error("Please ensure GitHub token is set in environment variables and source repository URL is provided. 🔑❓")
+    
+        except exceptions.CosmosHttpResponseError as e:
+            st.error(f"Failed to connect to Cosmos DB. HTTP error: {str(e)} 🚨")
+        except Exception as e:
+            st.error(f"An unexpected error occurred: {str(e)} 😱")
+
+    # 🚪 Logout button
+    if st.session_state.logged_in and st.sidebar.button("🚪 Logout"):
+        st.session_state.logged_in = False
+        st.session_state.selected_records.clear()
+        st.session_state.client = None
+        st.session_state.selected_database = None
+        st.session_state.selected_container = None
+        st.session_state.selected_document_id = None
+        st.session_state.current_index = 0
+        st.rerun()
+
+if __name__ == "__main__":
+    main()