tools.py

from smolagents import DuckDuckGoSearchTool
from smolagents import Tool
from huggingface_hub import InferenceClient
import soundfile as sf
import torch
from transformers import AutoModelForSpeechSeq2Seq, AutoProcessor, pipeline
from datasets import load_dataset
from transformers import Qwen2VLForConditionalGeneration, AutoTokenizer, AutoProcessor
from qwen_vl_utils import process_vision_info
import requests
from bs4 import BeautifulSoup
from pydub import AudioSegment
import numpy as np
from scipy.io import wavfile
import pandas as pd
from PIL import Image
import html5lib

class read_python_file(Tool):
    name = "reader_python"
    description = "Read a python file. Will return the entire code in the file"
    inputs = {
        "file_name": {
            "type": "string",
            "description": "The python file path"
        },
    }
    output_type = "string"

    def forward(self, file_name):

        try:
            with open(file_name, "r", encoding="utf-8") as fichier:
                contenu = fichier.read()
                print("Contenu du fichier :\n")
                return contenu
        
        except FileNotFoundError:
            print(f"Erreur : le fichier '{fichier_cible}' n'existe pas.")
            return
        except Exception as e:
            print(f"Une erreur est survenue : {e}")
            return

class read_excel_file(Tool):
    name = "reader_excel"
    description = "Read a excel file. Will return the entire info in the file"
    inputs = {
        "file_name": {
            "type": "string",
            "description": "The excel file path"
        },
    }
    output_type = "string"

    def forward(self, file_name):

        try:
            # Lecture de toutes les feuilles
            xls = pd.ExcelFile(file_name)
            print("Sheets :", xls.sheet_names)
        
            # Lecture d'une feuille spécifique (ici, la première)
            df = pd.read_excel(xls, sheet_name=xls.sheet_names[0])
            print("\nContent of the sheet :\n")
            return df
        
        except FileNotFoundError:
            print(f"Erreur : le fichier '{fichier_excel}' n'existe pas.")
            return
        except Exception as e:
            print(f"Une erreur est survenue : {e}")
            return 

        
class is_commutative(Tool):
    name = "commutative"
    description = "Performs a study on a table set to see if it is commutative."
    inputs = {
        "set": {
            "type": "array",
            "items": {"type": "string"},
            "description": "The set defined, for example : {'a','b','c'}."
        },
        "table": {
            "type": "string",
            "description": "The table in markdown format with rows separated by '\n'. Give only the table after the '|---|---|---|---|---|---|' symbol, starting with the '\n', ending with '\n\n', as a string."
        }
    }
    output_type = "string"

    def forward(self, set, table):
        set_0=list(sorted(set))
        table_0=np.empty((len(set_0)+1,len(set_0)+1),dtype='<U5')
            
        for i in range(len(set_0)):
            table_0[0,i+1]=set_0[i]
        
        k=1
        l=0
        for item in table:
            if item=='\n' or item=='|':
                table_0=table_0
            else:
                table_0[k,l]=item
                l=l+1
                if(l==6):
                    l=0
                    k=k+1
        
        for i in range(1,len(set_0)+1):
            for j in range(1,len(set_0)+1):
                if (table_0[i,j]!=table_0[j,i]):
                    return f"Not commutative, because of the elements: {set_0[i-1]} and {set_0[j-1]}"
        return "It is commutative"
        
class Web_research(Tool):
    name="web_search"
    description = "Web search on a specific topic."
    inputs = {
        "topic": {
            "type": "string",
            "description": "The topic on which the user wants the latest news"
        }
    }
    output_type = "string"

    def forward(self, topic: str):
        search_tool = DuckDuckGoSearchTool()
        # Example usage
        results = search_tool(f"{topic}")
        return f"Here is what we can find on the web for {topic} : str({results})"

class Find_wikipedia_URL(Tool):
    name="wiki_url_tool"
    description = "Always use to check a wikipedia ENGLISH URL page before trying to acces the URL. For another langage, you just have to change the beginning of the url (here, it is en for english)"
    inputs = {
        "subject": {
            "type": "string",
            "description": "The name or topic on which you want the Wikipedia URL"
        }
    }
    output_type = "string"

    def forward(self, subject: str):
        words=subject.split()
        url_wiki="https://en.wikipedia.org/wiki/"
        for i in range(len(words)):
            if(i==0):
                url_wiki+=str(words[i])
            if(i!=0):
                url_wiki+='_'+str(words[i])
        return f"Here is what we url to use : str({url_wiki}). If it does not work, change the first letters of {subject} to be upper or lower, but never change anything else"

class translate_everything(Tool):
    name="translator"
    description = "You do not understand a sentence? It does not look like any language you know? Try this tool, maybe the sentence is just reversed!"
    inputs = {
        "sentence": {
            "type": "string",
            "description": "The sentence to translate"
        }
    }
    output_type = "string"

    def forward(self, sentence: str):
        # Input string
        reversed_words = sentence.split() #' '.join(s.split()[::-1])
        right_sentence=[]
        for word in reversed_words:
            right_sentence.append(word[::-1])
        
        translated_sentence = " ".join(right_sentence[::-1])
        return f"The translated sentence is : {translated_sentence}"

class image_interpreter(Tool):
    name = "image_tool"
    description = "Allows you to convert an image to text (the function will describe the image)."
    inputs = {
        'image': {"type": "image", "description": "The image of interest, png format or jpeg (must be the local path of the image)"},
        'prompt': {"type": "string", "description": "The function returns a description of the image, but you can request specific details with this prompt to ensure they are included in the description."},
    }
    output_type = "string"

    def forward(self, image, prompt):
        device = "cuda" if torch.cuda.is_available() else "cpu"
        model = Qwen2VLForConditionalGeneration.from_pretrained(
        "Qwen/Qwen2-VL-7B-Instruct", torch_dtype="auto", device_map="auto"
        )
        processor = AutoProcessor.from_pretrained("Qwen/Qwen2-VL-7B-Instruct")
        
        messages = [
            {
                "role": "user",
                "content": [
                    {"type": "image", "image": Image.open(image).convert("RGB")},
                    {"type": "text", "text": "describe this image, with as much details as you can" + str(prompt)},
                ],
            }
        ]

        text = processor.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
        image_inputs, video_inputs = process_vision_info(messages)
        inputs = processor(
            text=[text],
            images=image_inputs,
            padding=True,
            return_tensors="pt",
        ).to(device)

        generated_ids = model.generate(**inputs, max_new_tokens=128)
        generated_ids_trimmed = [
            out_ids[len(in_ids):] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
        ]
        output_text = processor.batch_decode(
            generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
        )

        return output_text[0]


class audio_or_mp3__interpreter(Tool):
    name="audio_tool"
    description = "Allows you to convert audio into text. It uses Whisper, it is a state-of-the-art model for automatic speech recognition (ASR) and speech translation"
    inputs = {
        'audio': {"type": "audio", "description": "the audio of interest (must be the local path to this audio file). Must be in the format mp3."}
    }
    output_type = "string"
    
    def forward(self, audio):
        device = "cuda:0" if torch.cuda.is_available() else "cpu"
        torch_dtype = torch.float16 if torch.cuda.is_available() else torch.float32
        
        model_id = "openai/whisper-large-v3"
        model = AutoModelForSpeechSeq2Seq.from_pretrained(model_id, torch_dtype=torch_dtype)
        model.to(device)
        processor = AutoProcessor.from_pretrained(model_id)
        
        pipe = pipeline(
            "automatic-speech-recognition",
            model=model,
            tokenizer=processor.tokenizer,
            feature_extractor=processor.feature_extractor,
            torch_dtype=torch_dtype,
            device=device
        )
        
        audio = AudioSegment.from_mp3(audio)
        audio = audio.set_channels(1).set_frame_rate(16000)
        audio.export("output.wav", format="wav")
        
        result = pipe("output.wav", return_timestamps=True)
        return result["text"]

class Wikipedia_reader(Tool):
    name="wiki_tool"
    description = "To be used whenever you need to read a Wikipedia page. Will return all wikitables, to easily read it and find information"
    inputs = {
        "url": {
            "type": "string",
            "description": "The wikippedia url page"
        }
    }
    output_type = "string"

    def forward(self, url: str):
        #try:
        #    page = requests.get({url})
        #except Exception as e:
         #   print('Error downloading page: ',e)
        #soup = BeautifulSoup(page.text, 'html.parser')
        tables=pd.read_html(url)
        return tables