app.py

from smolagents import CodeAgent,DuckDuckGoSearchTool,HfApiModel,load_tool,tool
import datetime
import requests
import pytz
import yaml
from tools.final_answer import FinalAnswerTool

from Gradio_UI import GradioUI

import sys
import urllib.parse
import time
import re
import argparse

# Below is an example of a tool that does nothing. Amaze us with your creativity !!
@tool
def my_custom_tool(arg1:str, arg2:int)-> str: #it's import to specify the return type
    #Keep this format for the description / args / args description but feel free to modify the tool
    """A tool that does nothing yet [BLAST???]
    Args:
        arg1: the first argument
        arg2: the second argument
    """
    return "What magic will you build ?"

@tool
def get_current_time_in_timezone(timezone: str) -> str:
    """A tool that fetches the current local time in a specified timezone.
    Args:
        timezone: A string representing a valid timezone (e.g., 'America/New_York').
    """
    try:
        # Create timezone object
        tz = pytz.timezone(timezone)
        # Get current time in that timezone
        local_time = datetime.datetime.now(tz).strftime("%Y-%m-%d %H:%M:%S")
        return f"The current local time in {timezone} is: {local_time}"
    except Exception as e:
        return f"Error fetching time for timezone '{timezone}': {str(e)}"

@tool
def run_blast(arguments_string: str) -> str:
    """
    Submits a BLAST job to NCBI and retrieves the results.
    Parses arguments from a string input. Accepts a sequence directly as input, not a file.

    Args:
        arguments_string: A string containing the BLAST program, database, and query sequence.
            The string should be formatted as: "program database query_sequence".
            For example: "blastp nr ATGCGTAGCTAGCTAG...".
            The query sequence should be provided directly within the string.

    Returns:
        BLAST results in text format on success, or an error message string on failure.
    """
    # Use maxsplit=2 so that the query sequence remains intact even if it contains spaces.
    tokens = arguments_string.split(maxsplit=2)
    if len(tokens) < 3:
        return "Error: Invalid arguments. Usage: program database query_sequence"
    program, database, query_sequence = tokens
    # Remove all whitespace from the query sequence (if a contiguous sequence is desired)
    query_sequence = "".join(query_sequence.split())

    # Adjust program parameters for specific cases.
    if program.lower() == "megablast":
        program = "blastn&MEGABLAST=on"
    elif program.lower() == "rpsblast":
        program = "blastp&SERVICE=rpsblast"

    api_url = 'https://blast.ncbi.nlm.nih.gov/blast/Blast.cgi'
    payload = {
    'CMD': 'Put',
    'PROGRAM': program,
    'DATABASE': database,
    'QUERY': query_sequence,
    'FORMAT_TYPE': 'Text',
    'HITLIST_SIZE': 5,      # Return only the top 5 hits
    'DESCRIPTIONS': 5,       # Print only 5 descriptions
    'ALIGNMENTS': 5
    }


    try:
        response = requests.post(api_url, data=payload)
        response.raise_for_status()
    except requests.exceptions.RequestException as e:
        return f"Error submitting request: {e}"

    response_content = response.text

    # Use improved regex patterns to capture RID and RTOE
    rid_match = re.search(r"RID\s*=\s*(\S+)", response_content)
    rtoe_match = re.search(r"RTOE\s*=\s*(\d+)", response_content)

    if rid_match and rtoe_match:
        rid = rid_match.group(1).strip()
        try:
            rtoe = int(rtoe_match.group(1).strip())
        except ValueError:
            return "Error: Invalid RTOE value received from BLAST response."
    else:
        return "Error: Could not parse RID or RTOE from BLAST response.\nResponse content:\n" + response_content

    # Initial wait based on the estimated time
    time.sleep(rtoe)

    # Poll for results with a maximum number of attempts to avoid an infinite loop.
    max_polls = 60  # e.g., 60 polls * 5 seconds = 300 seconds maximum wait time
    poll_attempts = 0
    while poll_attempts < max_polls:
        time.sleep(5)
        poll_attempts += 1
        poll_url = f"{api_url}?CMD=Get&FORMAT_OBJECT=SearchInfo&RID={rid}"
        try:
            poll_response = requests.get(poll_url)
            poll_response.raise_for_status()
        except requests.exceptions.RequestException as e:
            return f"Error polling for results: {e}"

        status_content = poll_response.text

        if "Status=WAITING" in status_content:
            continue
        elif "Status=FAILED" in status_content:
            return f"Search {rid} failed; please report to blast-help@ncbi.nlm.nih.gov."
        elif "Status=UNKNOWN" in status_content:
            return f"Search {rid} expired."
        elif "Status=READY" in status_content:
            if "ThereAreHits=yes" in status_content:
                break  # Ready to fetch the results
            else:
                return "No hits found."
        else:
            return "Unknown error during polling.\nStatus response content:\n" + status_content
    else:
        return "Error polling for results: timed out."

    # Retrieve and return the BLAST results.
    result_url = f"{api_url}?CMD=Get&FORMAT_TYPE=Text&RID={rid}"
    try:
        result_response = requests.get(result_url)
        result_response.raise_for_status()
    except requests.exceptions.RequestException as e:
        return f"Error retrieving results: {e}"

    return result_response.text


final_answer = FinalAnswerTool()

# If the agent does not answer, the model is overloaded, please use another model or the following Hugging Face Endpoint that also contains qwen2.5 coder:
# model_id='https://pflgm2locj2t89co.us-east-1.aws.endpoints.huggingface.cloud' 

model = HfApiModel(
max_tokens=2096,
temperature=0.5,
model_id="Qwen/Qwen2.5-Coder-32B-Instruct", #'Qwen/Qwen2.5-Coder-32B-Instruct',# it is possible that this model may be overloaded,
custom_role_conversions=None,
)



# Import tool from Hub
image_generation_tool = load_tool("agents-course/text-to-image", trust_remote_code=True)

with open("prompts.yaml", 'r') as stream:
    prompt_templates = yaml.safe_load(stream)
    
agent = CodeAgent(
    model=model,
    tools=[DuckDuckGoSearchTool(),run_blast,image_generation_tool,final_answer], ## add your tools here (don't remove final answer)
    max_steps=6,
    verbosity_level=1,
    grammar=None,
    planning_interval=None,
    name=None,
    description=None,
    prompt_templates=prompt_templates
)


GradioUI(agent).launch()