MistriDevLab / app.py
acecalisto3's picture
Update app.py
4fd2fa4 verified
raw
history blame
9.58 kB
import os
import subprocess
import random
import json
from datetime import datetime
from huggingface_hub import InferenceClient, cached_download, hf_hub_url
import gradio as gr
from safe_search import safe_search
from i_search import google, i_search as i_s
from agent import (
ACTION_PROMPT,
ADD_PROMPT,
COMPRESS_HISTORY_PROMPT,
LOG_PROMPT,
LOG_RESPONSE,
MODIFY_PROMPT,
PREFIX,
SEARCH_QUERY,
READ_PROMPT,
TASK_PROMPT,
UNDERSTAND_TEST_RESULTS_PROMPT,
)
from utils import parse_action, parse_file_content, read_python_module_structure
class App:
def __init__(self):
self.app_state = {"components": []}
self.terminal_history = ""
self.components_registry = {
"Button": {
"properties": {
"label": "Click Me",
"onclick": ""
},
"description": "A clickable button",
"code_snippet": "gr.Button(value='{{label}}', variant='primary')"
},
"Text Input": {
"properties": {
"value": "",
"placeholder": "Enter text"
},
"description": "A field for entering text",
"code_snippet": "gr.Textbox(label='{{placeholder}}')"
},
"Image": {
"properties": {
"src": "#",
"alt": "Image"
},
"description": "Displays an image",
"code_snippet": "gr.Image(label='{{alt}}')"
},
"Dropdown": {
"properties": {
"choices": ["Option 1", "Option 2"],
"value": ""
},
"description": "A dropdown menu for selecting options",
"code_snippet": "gr.Dropdown(choices={{choices}}, label='Dropdown')"
}
}
self.nlp_model_names = [
"google/flan-t5-small",
"Qwen/CodeQwen1.5-7B-Chat-GGUF",
"bartowski/Codestral-22B-v0.1-GGUF",
"bartowski/AutoCoder-GGUF"
]
self.nlp_models = []
self.initialize_nlp_models()
def initialize_nlp_models(self):
for nlp_model_name in self.nlp_model_names:
try:
cached_download(hf_hub_url(nlp_model_name, revision="main"))
self.nlp_models.append(InferenceClient(nlp_model_name))
except:
self.nlp_models.append(None)
def get_nlp_response(self, input_text, model_index):
if self.nlp_models[model_index]:
response = self.nlp_models[model_index].text_generation(input_text)
return response.generated_text
else:
return "NLP model not available."
class Component:
def __init__(self, type, properties=None, id=None):
self.id = id or random.randint(1000, 9999)
self.type = type
self.properties = properties or self.components_registry[type]["properties"].copy()
def to_dict(self):
return {
"id": self.id,
"type": self.type,
"properties": self.properties,
}
def render(self):
if self.type == "Dropdown":
self.properties["choices"] = str(self.properties["choices"]).replace("[", "").replace("]", "").replace("'", "")
return self.components_registry[self.type]["code_snippet"].format(**self.properties)
def update_app_canvas(self):
components_html = "".join([f"<div>Component ID: {component['id']}, Type: {component['type']}, Properties: {component['properties']}</div>" for component in self.app_state["components"]])
return components_html
def add_component(self, component_type):
if component_type in self.components_registry:
new_component = self.Component(component_type)
self.app_state["components"].append(new_component.to_dict())
return (
self.update_app_canvas(),
f"System: Added component: {component_type}\n",
)
else:
return None, f"Error: Invalid component type: {component_type}\n"
def run_terminal_command(self, command, history):
output = ""
try:
if command.startswith("add "):
component_type = command.split("add ")[1]
return self.add_component(component_type)
elif command.startswith("search "):
query = command.split("search ")[1]
return google(query)
elif command.startswith("i search "):
query = command.split("i search ")[1]
return i_s(query)
elif command.startswith("safe search "):
query = command.split("safesearch ")[1]
return safe_search(query)
elif command.startswith("read "):
file_path = command.split("read ")[1]
return parse_file_content(file_path)
elif command == "task":
return TASK_PROMPT
elif command == "modify":
return MODIFY_PROMPT
elif command == "log":
return LOG_PROMPT
elif command.startswith("understand test results "):
test_results = command.split("understand test results ")[1]
return self.understand_test_results(test_results)
elif command.startswith("compress history"):
return self.compress_history(history)
elif command == "help":
return self.get_help_message()
elif command == "exit":
exit()
else:
output = subprocess.check_output(command, shell=True).decode("utf-8")
except Exception as e:
output = str(e)
return output or "No output\n"
def compress_history(self, history):
compressed_history = ""
lines = history.strip().split("\n")
for line in lines:
if not line.strip().startswith("#"):
compressed_history += line + "\n"
return compressed_history
def understand_test_results(self, test_results):
# Logic to understand test results
return UNDERSTAND_TEST_RESULTS_PROMPT
def get_help_message(self):
return """
Available commands:
- add [component_type]: Add a component to the app canvas
- search [query]: Perform a Google search
- i search [query]: Perform an intelligent search
- safe search [query]: Perform a safe search
- read [file_path]: Read and parse the content of a Python module
- task: Prompt for a task to perform
- modify: Prompt to modify a component property
- log: Prompt to log a response
- understand test results [test_results]: Understand test results
- compress history: Compress the terminal history by removing comments
- help: Show this help message
- exit: Exit the program
"""
def process_input(self, input_text):
if input_text.strip().startswith("/"):
command = input_text.strip().lstrip("/")
output = self.run_terminal_command(command, self.terminal_history)
self.terminal_history += f"{input_text}\n{output}\n"
return output
else:
model_index = random.randint(0, len(self.nlp_models)-1)
response = self.get_nlp_response(input_text, model_index)
component_id, action, property_name, property_value = parse_action(response)
if component_id:
component = next((comp for comp in self.app_state["components"] if comp["id"] == component_id), None)
if component:
if action == "update":
component["properties"][property_name] = property_value
return (
self.update_app_canvas(),
f"System: Updated property '{property_name}' of component with ID {component_id}\n",
)
elif action == "remove":
self.app_state["components"].remove(component)
return (
self.update_app_canvas(),
f"System: Removed component with ID {component_id}\n",
)
else:
return None, f"Error: Invalid action: {action}\n"
else:
return None, f"Error: Component with ID {component_id} not found\n"
else:
return None, f"Error: Failed to parse action from NLP response\n"
def run(self):
print("Welcome to the Python App Builder!")
print("Type 'help' to see the available commands.")
print("-" * 50)
try:
while True:
try:
input_text = input("Enter input: ")
except EOFError:
print("Error: Input reading interrupted. Please provide valid input.")
continue
output, system_message = self.process_input(input_text)
if output:
print(output)
if system_message:
print(system_message)
except KeyboardInterrupt:
print("\nApplication stopped by user.")
if __name__ == "__main__":
app = App()
app.run()