import streamlit as st
from streamlit_ace import st_ace
from streamlit_jupyter import st_jupyter
from transformers import pipeline, AutoTokenizer, AutoModelForSeq2SeqLM
import os
import subprocess
import black
from pylint import lint
from io import StringIO
import sys
import torch
from huggingface_hub import hf_hub_url, cached_download, HfApi
import re
from typing import List, Dict
# Access Hugging Face API key from secrets
hf_token = st.secrets["hf_token"]
if not hf_token:
st.error("Hugging Face API key not found. Please make sure it is set in the secrets.")
HUGGING_FACE_REPO_URL = "https://huggingface.co/spaces/acecalisto3/0shotTest"
PROJECT_ROOT = "projects"
AGENT_DIRECTORY = "agents"
AVAILABLE_CODE_GENERATIVE_MODELS = ["bigcode/starcoder", "Salesforce/codegen-350M-mono", "microsoft/CodeGPT-small"]
# Global state to manage communication between Tool Box and Workspace Chat App
if 'chat_history' not in st.session_state:
st.session_state.chat_history = []
if 'terminal_history' not in st.session_state:
st.session_state.terminal_history = []
if 'workspace_projects' not in st.session_state:
st.session_state.workspace_projects = {}
if 'available_agents' not in st.session_state:
st.session_state.available_agents = []
# AI Guide Toggle
ai_guide_level = st.sidebar.radio("AI Guide Level", ["Full Assistance", "Partial Assistance", "No Assistance"])
class AIAgent:
def __init__(self, name: str, description: str, skills: List[str]):
self.name = name
self.description = description
self.skills = skills
self._hf_api = HfApi() # Initialize HfApi here
def create_agent_prompt(self) -> str:
skills_str = '\n'.join([f"* {skill}" for skill in self.skills])
agent_prompt = f"""
As an elite expert developer, my name is {self.name}. I possess a comprehensive understanding of the following areas:
{skills_str}
I am confident that I can leverage my expertise to assist you in developing and deploying cutting-edge web applications. Please feel free to ask any questions or present any challenges you may encounter.
"""
return agent_prompt
def autonomous_build(self, chat_history: List[tuple[str, str]], workspace_projects: Dict[str, Dict],
project_name: str, selected_model: str, hf_token: str) -> tuple[str, str]:
summary = "Chat History:\n" + "\n".join([f"User: {u}\nAgent: {a}" for u, a in chat_history])
summary += "\n\nWorkspace Projects:\n" + "\n".join([f"{p}: {details}" for p, details in workspace_projects.items()])
next_step = "Based on the current state, the next logical step is to implement the main application logic."
return summary, next_step
def deploy_built_space_to_hf(self, project_name: str) -> str:
# Assuming you have a function that generates the space content
space_content = generate_space_content(project_name)
repository = self._hf_api.create_repo(
repo_id=project_name,
private=True,
token=hf_token,
exist_ok=True,
space_sdk="streamlit"
)
self._hf_api.upload_file(
path_or_fileobj=space_content,
path_in_repo="app.py",
repo_id=project_name,
repo_type="space",
token=hf_token
)
return repository.name
def has_valid_hf_token(self) -> bool:
return self._hf_api.whoami(token=hf_token) is not None
def process_input(input_text: str) -> str:
chatbot = pipeline("text-generation", model="microsoft/DialoGPT-medium", tokenizer="microsoft/DialoGPT-medium", clean_up_tokenization_spaces=True)
response = chatbot(input_text, max_length=50, num_return_sequences=1)[0]['generated_text']
return response
def run_code(code: str) -> str:
try:
result = subprocess.run(code, shell=True, capture_output=True, text=True)
return result.stdout
except Exception as e:
return str(e)
def workspace_interface(project_name: str) -> str:
project_path = os.path.join(PROJECT_ROOT, project_name)
if not os.path.exists(project_path):
os.makedirs(project_path)
st.session_state.workspace_projects[project_name] = {'files': []}
return f"Project '{project_name}' created successfully."
else:
return f"Project '{project_name}' already exists."
def add_code_to_workspace(project_name: str, code: str, file_name: str) -> str:
project_path = os.path.join(PROJECT_ROOT, project_name)
if not os.path.exists(project_path):
return f"Project '{project_name}' does not exist."
file_path = os.path.join(project_path, file_name)
with open(file_path, "w") as file:
file.write(code)
st.session_state.workspace_projects[project_name]['files'].append(file_name)
return f"Code added to '{file_name}' in project '{project_name}'."
def display_chat_history(chat_history: List[tuple[str, str]]) -> str:
return "\n".join([f"User: {u}\nAgent: {a}" for u, a in chat_history])
def display_workspace_projects(workspace_projects: Dict[str, Dict]) -> str:
return "\n".join([f"{p}: {details}" for p, details in workspace_projects.items()])
def generate_space_content(project_name: str) -> str:
# Logic to generate the Streamlit app content based on project_name
# ... (This is where you'll need to implement the actual code generation)
return "import streamlit as st\nst.title('My Streamlit App')\nst.write('Hello, world!')"
# Function to display the AI Guide chat
def display_ai_guide_chat(chat_history: List[tuple[str, str]]):
st.markdown("
", unsafe_allow_html=True)
for user_message, agent_message in chat_history:
st.markdown(f"
{user_message}
", unsafe_allow_html=True)
st.markdown(f"
{agent_message}
", unsafe_allow_html=True)
st.markdown("
", unsafe_allow_html=True)
# Load the CodeGPT tokenizer explicitly
code_generator_tokenizer = AutoTokenizer.from_pretrained("microsoft/CodeGPT-small-py", clean_up_tokenization_spaces=True)
# Load the CodeGPT model for code completion
code_generator = pipeline("text-generation", model="microsoft/CodeGPT-small-py", tokenizer=code_generator_tokenizer)
def analyze_code(code: str) -> List[str]:
hints = []
# Example pointer: Suggest using list comprehensions
if re.search(r'for .* in .*:\n\s+.*\.append\(', code):
hints.append("Consider using a list comprehension instead of a loop for appending to a list.")
# Example pointer: Recommend using f-strings for string formatting
if re.search(r'\".*\%s\"|\'.*\%s\'', code) or re.search(r'\".*\%d\"|\'.*\%d\'', code):
hints.append("Consider using f-strings for cleaner and more efficient string formatting.")
# Example pointer: Avoid using global variables
if re.search(r'\bglobal\b', code):
hints.append("Avoid using global variables. Consider passing parameters or using classes.")
# Example pointer: Recommend using `with` statement for file operations
if re.search(r'open\(.+\)', code) and not re.search(r'with open\(.+\)', code):
hints.append("Consider using the `with` statement when opening files to ensure proper resource management.")
return hints
def get_code_completion(prompt: str) -> str:
# Generate code completion based on the current code input
# Use max_new_tokens instead of max_length
completions = code_generator(prompt, max_new_tokens=50, num_return_sequences=1)
return completions[0]['generated_text']
def lint_code(code: str) -> List[str]:
# Capture pylint output
pylint_output = StringIO()
sys.stdout = pylint_output
# Run pylint on the provided code
pylint.lint.Run(['--from-stdin'], do_exit=False, argv=[], stdin=StringIO(code))
# Reset stdout and fetch lint results
sys.stdout = sys.__stdout__
lint_results = pylint_output.getvalue().splitlines()
return lint_results
# Set page configuration
st.set_page_config(layout="wide", page_title="RoboCoders")
# Sidebar for chat interface
st.sidebar.title("Chat Interface")
user_input = st.sidebar.text_area("Type your idea, task, or request here:")
# Placeholder function to simulate code generation
def generate_code(user_input):
return f"# Generated code for: {user_input}\nprint('Hello, World!')"
# Main layout
col1, col2 = st.columns([1, 3])
with col1:
st.title("Code Editor")
if user_input:
code = generate_code(user_input)
else:
code = ""
code = st_ace(value=code, language='python', theme='monokai', height=400)
with col2:
st.title("Jupyter IPython Console")
st_jupyter()
st.title("Read-Only Terminal")
st.text_area("Terminal Output", height=200)
# Placeholder for autonomous agent logic
if user_input:
st.sidebar.write("Processing your request...")
# Example: Generate a simple "Hello, World!" Streamlit app
generated_code = code_generator(f"Create a Streamlit app that displays 'Hello, World!'", max_new_tokens=50, num_return_sequences=1)[0]['generated_text']
st.sidebar.write("Generated code:")
st.sidebar.code(generated_code, language="python")
# Update the code editor
code = generated_code
# ... (Additional logic for code analysis, project management, etc.)
# ... (Update the Jupyter console and terminal output as needed)
# ... (Interact with the AI guide chatbot)
if __name__ == "__main__":
st.sidebar.title("Navigation")
app_mode = st.sidebar.selectbox("Choose the app mode", ["Home", "Terminal", "Explorer", "Code Editor", "Build & Deploy"])
if app_mode == "Home":
st.title("Welcome to AI-Guided Development")
st.write("This application helps you build and deploy applications with the assistance of an AI Guide.")
st.write("Toggle the AI Guide from the sidebar to choose the level of assistance you need.")
elif app_mode == "Terminal":
st.header("Terminal")
terminal_input = st.text_input("Enter a command:")
if st.button("Run"):
output = run_code(terminal_input)
st.session_state.terminal_history.append((terminal_input, output))
st.code(output, language="bash")
if ai_guide_level != "No Assistance":
st.write("Run commands here to add packages to your project. For example: pip install .")
if terminal_input and "install" in terminal_input:
package_name = terminal_input.split("install")[-1].strip()
st.write(f"Package {package_name} will be added to your project.")
elif app_mode == "Explorer":
st.header("Explorer")
uploaded_file = st.file_uploader("Upload a file", type=["py"])
if uploaded_file:
file_details = {"FileName": uploaded_file.name, "FileType": uploaded_file.type}
st.write(file_details)
save_path = os.path.join(PROJECT_ROOT, uploaded_file.name)
with open(save_path, "wb") as f:
f.write(uploaded_file.getbuffer())
st.success(f"File {uploaded_file.name} saved successfully!")
st.write("Drag and drop files into the 'app' folder.")
for project, details in st.session_state.workspace_projects.items():
st.write(f"Project: {project}")
for file in details['files']:
st.write(f" - {file}")
if st.button(f"Move {file} to app folder"):
# Logic to move file to 'app' folder
pass
if ai_guide_level != "No Assistance":
st.write("You can upload files and move them into the 'app' folder for building your application.")
elif app_mode == "Code Editor":
st.header("Code Editor")
code_editor = st.text_area("Write your code:", height=300)
if st.button("Save Code"):
# Logic to save code
pass
if ai_guide_level != "No Assistance":
st.write("The function foo() requires the bar package. Add it to requirements.txt.")
# Analyze code and provide real-time hints
hints = analyze_code(code_editor)
if hints:
st.write("**Helpful Hints:**")
for hint in hints:
st.write(f"- {hint}")
if st.button("Get Code Suggestion"):
# Provide a predictive code completion
completion = get_code_completion(code_editor)
st.write("**Suggested Code Completion:**")
st.code(completion, language="python")
if st.button("Check Code"):
# Analyze the code for errors and warnings
lint_results = lint_code(code_editor)
if lint_results:
st.write("**Errors and Warnings:**")
for result in lint_results:
st.write(result)
else:
st.write("No issues found! Your code is clean.")
elif app_mode == "Build & Deploy":
st.header("Build & Deploy")
project_name_input = st.text_input("Enter Project Name for Automation:")
if st.button("Automate"):
selected_agent = st.selectbox("Select an AI agent", st.session_state.available_agents)
selected_model = st.selectbox("Select a code-generative model", AVAILABLE_CODE_GENERATIVE_MODELS)
agent = AIAgent(selected_agent, "", []) # Load the agent without skills for now
summary, next_step = agent.autonomous_build(st.session_state.chat_history, st.session_state.workspace_projects, project_name_input, selected_model, hf_token)
st.write("Autonomous Build Summary:")
st.write(summary)
st.write("Next Step:")
st.write(next_step)
if agent._hf_api and agent.has_valid_hf_token():
repository_name = agent.deploy_built_space_to_hf(project_name_input)
st.markdown("## Congratulations! Successfully deployed Space 🚀 ##")
st.markdown(f"[Check out your new Space here](hf.co/{repository_name})")
# AI Guide Chat
if ai_guide_level != "No Assistance":
display_ai_guide_chat(st.session_state.chat_history)
# Add a text input for user to interact with the AI Guide
user_input = st.text_input("Ask the AI Guide a question:", key="user_input")
if st.button("Send"):
if user_input:
# Process the user's input and get a response from the AI Guide
agent_response = process_input(user_input)
st.session_state.chat_history.append((user_input, agent_response))
# Clear the user input field
st.session_state.user_input = ""
# CSS for styling
st.markdown("""
""", unsafe_allow_html=True)