{title}

import nltk from nltk.tokenize import word_tokenize from nltk.corpus import stopwords from nltk.stem import WordNetLemmatizer # Téléchargez les ressources nécessaires pour nltk nltk.download('punkt') nltk.download('stopwords') nltk.download('wordnet') class HTMLGenerator: def __init__(self): self.lemmatizer = WordNetLemmatizer() def generate_html(self, title, content): html_code = f""" {title} {content} """ return html_code class Chatbot: def __init__(self): self.html_generator = HTMLGenerator() self.stop_words = set(stopwords.words('english')) def process_input(self, user_input): tokens = word_tokenize(user_input.lower()) filtered_tokens = [word for word in tokens if word.isalnum() and word not in self.stop_words] lemmatized_tokens = [self.html_generator.lemmatizer.lemmatize(word) for word in filtered_tokens] return lemmatized_tokens def respond(self, tokens): if 'create' in tokens and ('website' in tokens or 'web' in tokens): return "Sure! What should be the title of your website?" elif 'title' in tokens: return "Great! Now, what content would you like to include in your website?" elif 'content' in tokens: title_index = tokens.index('title') + 1 content_index = tokens.index('content') + 1 title = ' '.join(tokens[title_index:content_index-1]) content = ' '.join(tokens[content_index:]) html_code = self.html_generator.generate_html(title, content) return f"Here's the HTML code for your website:\n\n{html_code}" elif 'help' in tokens: return "You can ask me to create a website by providing a title and content. For example, 'Create a website with title My Website and content Welcome to my website!'" elif 'exit' in tokens or 'quit' in tokens: return "Goodbye!" else: return "I'm sorry, I didn't understand that. You can ask for help by typing 'help'." def main(): print("Welcome to the HTML website generator bot!") chatbot = Chatbot() while True: user_input = input("You: ") tokens = chatbot.process_input(user_input) response = chatbot.respond(tokens) print("Bot:", response) if 'exit' in tokens or 'quit' in tokens: break if __name__ == "__main__": main() # Installez Rasa avec la commande suivante : pip install rasa # Créez les fichiers suivants : nlu.yml, domain.yml et config.yml dans le même répertoire. # nlu.yml version: "2.0" language: "fr" intents: [greeting, weather] phrases: greeting: - Bonjour - Salut - Hello - Hi weather: - Quel temps fait-il ? - Est-ce qu'il pleut ? - Comment est la météo aujourd'hui ? # domain.yml version: "2.0" intents: [weather] responses: utter_ask_location: - Où est-ce que tu te trouves ? utter_goodbye: - Au revoir ! utter_unable_to_help: - Desole, je ne peux pas répondre à cette question. utter_ask_weather: - La météo dépend de ton emplacement. Puis-je savoir où es-tu situé ? actions: - utter_ask_location - utter_goodbye - action_weather slots: location: type: string required: true temperature: type: number weather_state: type: string types: WeatherState: values: ["ensoleillé", "nuageux", "pluvieux"] rules: - rule: Responds with goodbye when "utter_goodbye" is selected steps: - intent: utter_goodbye - rule: Responds with weather info based on slot values steps: - intent: weather - action: utter_ask_weather - action: utter_ask_location - form: null - slot: "temperature" - slot: "weather_state" - action: utter_goodbye policy: version: "2.0" policies: - name: "TemperaturePolicy" featurizer: class_name: MemoizationFeaturizer config: {} policy_class: PoliciesKitchenSinkPolicy configuration: {} - name: "ParagraphPredictorPolicy" featurizer: class_name: ParagraphEmbeddingFeaturizer config: {"embedding_model": "all-MiniLM-L6-v2"} policy_class: CountBasedFallbackPolicy configuration: {} # config.yml version: "2.0" language: "fr" pipeline: - name: WhitespaceTokenizer config: {} - name: RegexFeaturizer config: {} - name: LexicalSyntacticFeaturizer config: {} - name: CountVectorizerFeaturizer config: tokenizes_sentences: false min_ngram_size: 1 max_ngram_size: 4 include_positionals: true use_dict: true word_list_path: "" template_paths: [] - name: DIETClassifier config: epochs: 100 batch_size: 8 hidden_size: 256 num_layers: 2 random_seed: 1789 optimizer: adam loss_function: CrossEntropyLoss - name: EntitySynonymMapper config: {} - name: ResponseSelector config: epochs: 100 batch_size: 8 hidden_size: 256 num_layers: 2 random_seed: 1789 optimizer: adam loss_function: CrossEntropyLoss - name: FallbackClassifier config: {} policies: default_policy: class_name: RulePoliciesKitchenSinkPolicy config: {} fallback_policy: class_name: ParagraphPredictorPolicy config: {} featurizer: class_name: ParagraphEmbeddingFeaturizer config: {"embedding_model": "all-MiniLM-L6-v2"} policy_class: CountBasedFallbackPolicy configuration: {} # Exécutez le script suivant pour entrainement : ```python import os from rasa_sdk import ActionServer from rasa_sdk.events import SlotSet, EventType from rasa_sdk.executor import CollectingDispatcher actions = ['action_weather'] if __name__ == "__main__": from rasa_sdk.events import SlotSet, IntentForm, ActiveLoopConfirmed actions += [IntentForm(), ActiveLoopConfirmed()] dispatcher = CollectingDispatcher.from_commands([], actions) loop_started_event = EventType([]) loop_finished_event = EventType([]) action_completed_event = EventType([]) slot_set_event = EventType([]) @dispatcher.rule('Responds with goodbye when "utter_goodbye" is selected') def handle_goodbyes(dispatcher: CollectingDispatcher) -> List[Dict[Text, Any]]: return [SlotSet("anything", None), ActionSet(["utter_goodbye"])]) @action("action_weather") async def weather_(self, dispatcher: CollectingDispatcher, tracker: Tracker, domain: Dict[Text, Any]) -> List[Dict[Text, Any]]: return [slot_set_event, SlotSet("location", self._parse_location(tracker)), ActionSet(["utter_ask_weather"])] @staticmethod def _parse_location(tracker: Tracker) -> Text: return tracker.get_slot("location") loop_started = False loop_finished = False while not loop_finished: try: action_server = ActionServer.create( os.environ.get("SERVICE_URL", "http://localhost:5055"), actions=actions, project="testproject", model_directory="models", credential_file="credentials.json", config_checkpoint_interval=1, debug=False, cors_origins=[], verify_ssl=True, ) dispatcher.reset() loop_started = True action_server.handle_new_messages() except Exception as e: print(str(e)) continue finally: if loop_started and loop_finished: loop_finished = False loop_started = False action_server.stop() action_server.wait_for_termination() if __name__ == "__main__": main() rasa shell --debugversion: "2.0" language: fr intents: - greetings - ask_weather - create_website - give_website_details phrases: greetings: - bonjour - salut - hello - hi ask_weather: - quel temps fait-il ? - est-ce qu'il pleut ? - comment est la météo aujourd'hui ? create_website: - créer un site web - faire un nouveau site give_website_details: - le titre doit être «{}» - contenu «{}»action_endpoint: url: "http://localhost:5055/webhook" rasa_nlu_path: ./models/nlu from typing import Any, Text, Dict, List from rasa_sdk import Action, Tracker from rasa_sdk.executor import CollectingDispatcher from rasa_sdk.forms import FormValidationAction class ValidateWebsiteNameForm(FormValidationAction): def name(self) -> Text: return "validate_website_name_form" async def validate_website_name( self, slot_value: Any, dispatcher: CollectingDispatcher, tracker: Tracker, domain: Dict[Text, Any], ) -> Dict[Text, Any]: if len(slot_value.strip().split()) > 1: dispatcher.utter_message(text="Je suis désolé, j'ai besoin d'un seul mot pour le nom du site.") raise ValueError("Un seul mot attendu") return {"website_name": slot_value} class GenerateHTMLForm(FormValidationAction): def name(self) -> Text: return "generate_html_form" async def validate_site_title( self, slot_value: Any, dispatcher: CollectingDispatcher, tracker: Tracker, domain: Dict[Text, Any], ) -> Dict[Text, Any]: if '"' not in slot_value: dispatcher.utter_message(text="Le titre doit être entouré de guillemets.") raise ValueError("Guillemets requis") return {"site_title": slot_value} async def validate_site_content( self, slot_value: Any, dispatcher: CollectingDispatcher, tracker: Tracker, domain: Dict[Text, Any], ) -> Dict[Text, Any]: if '"' not in slot_value: dispatcher.utter_message(text="Le contenu doit être entouré de guillemets.") raise ValueError("Guillemets requis") return {"site_content": slot_value} class ActionGenerateHTML(Action): def name(self) -> Text: return "action_generate_html" def run(self, dispatcher: CollectingDispatcher, tracker: Tracker, domain: Dict[Text, Any]) -> List[EventType]: html = f""" {tracker.latest_message['entities'][0]['value']} {tracker.latest_message['entities'][1]['value']} """ dispatcher.utter_message(text="Votre code HTML a été généré.", html_format=True, message=html) return [] ... policies: ... fallback: type: TEDPolicy nlu_threshold: 0.4 core_threshold: 0.4 slots_threshold: 0.4 ...

Assistants

Popular assistants made by the community

Mathssolver by leon

Sport predictor

AI.co

Artgenerate

HTMLgenius

Camerbot

PromptAssist

Webdesignchatbot

FootFlicks

AIgenerate

ScriptMaster ActionGenius

HTMLBOT

Webbot

CalcSciBot

Posebot

Ok

Chat pronostic