Update src/streamlit_app.py

src/streamlit_app.py

@@ -1,40 +1,213 @@
-import altair as alt
-import numpy as np
-import pandas as pd
-import streamlit as st
+"""
+CodeAlpha - Task 1: Language Translation Tool
+Author: (Loic HOUNYOVI)
+Description: Application web permettant de traduire du texte d'une langue
+             source vers une langue cible en utilisant un vrai modèle de
+             deep learning (Meta NLLB-200) chargé localement via la
+             librairie Hugging Face `transformers`, avec en bonus une
+             synthèse vocale (text-to-speech) et un bouton copier.
 
+Modèle utilisé : facebook/nllb-200-distilled-600M
+  - Modèle Seq2Seq (encoder-decoder) multilingue, pré-entraîné par Meta AI
+  - Supporte ~200 langues avec un seul et même modèle
+  - Chargé via transformers.pipeline("translation", ...)
 """
-# Welcome to Streamlit!
 
-Edit `/streamlit_app.py` to customize this app to your heart's desire :heart:.
-If you have any questions, checkout our [documentation](https://docs.streamlit.io) and [community
-forums](https://discuss.streamlit.io).
+import io
 
-In the meantime, below is an example of what you can do with just a few lines of code:
-"""
+import streamlit as st
+from transformers import AutoModelForSeq2SeqLM, AutoTokenizer
+from gtts import gTTS
+
+try:
+    from langdetect import detect as detect_lang
+    LANGDETECT_AVAILABLE = True
+except ImportError:
+    LANGDETECT_AVAILABLE = False
+
+# ---------------------------------------------------------
+# Configuration de la page
+st.set_page_config(
+    page_title="CodeAlpha Language Translator (AI Model)",
+    page_icon="🌐",
+    layout="centered",
+)
+
+# ---------------------------------------------------------
+# Langues supportées : code affichage -> (code NLLB / FLORES-200, code gTTS)
+LANGUAGES = {
+    "en": "Anglais",
+    "fr": "Français",
+    "es": "Espagnol",
+    "de": "Allemand",
+    "it": "Italien",
+    "pt": "Portugais",
+    "ar": "Arabe",
+    "zh-CN": "Chinois (simplifié)",
+    "ja": "Japonais",
+    "ko": "Coréen",
+    "ru": "Russe",
+    "hi": "Hindi",
+    "nl": "Néerlandais",
+    "tr": "Turc",
+}
+
+NLLB_CODES = {
+    "en": "eng_Latn",
+    "fr": "fra_Latn",
+    "es": "spa_Latn",
+    "de": "deu_Latn",
+    "it": "ita_Latn",
+    "pt": "por_Latn",
+    "ar": "arb_Arab",
+    "zh-CN": "zho_Hans",
+    "ja": "jpn_Jpan",
+    "ko": "kor_Hang",
+    "ru": "rus_Cyrl",
+    "hi": "hin_Deva",
+    "nl": "nld_Latn",
+    "tr": "tur_Latn",
+}
+
+# gTTS accepte directement des codes ISO classiques (tous nos codes conviennent)
+TTS_SUPPORTED = set(LANGUAGES.keys())
+
+
+# ---------------------------------------------------------
+# Chargement du modèle (mis en cache : ne se recharge pas à chaque interaction)
+MODEL_NAME = "facebook/nllb-200-distilled-600M"
+
+
+@st.cache_resource(show_spinner="Chargement du modèle NLLB-200 (peut prendre un moment la 1ère fois)...")
+def load_model():
+    tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
+    model = AutoModelForSeq2SeqLM.from_pretrained(MODEL_NAME)
+    return tokenizer, model
+
+
+def translate_text(text: str, src_code: str, tgt_code: str) -> str:
+    """Traduit `text` de src_code vers tgt_code (codes FLORES-200) via NLLB-200."""
+    tokenizer, model = load_model()
+    tokenizer.src_lang = src_code
+    inputs = tokenizer(text, return_tensors="pt", truncation=True)
+    forced_bos_token_id = tokenizer.convert_tokens_to_ids(tgt_code)
+    generated_tokens = model.generate(
+        **inputs,
+        forced_bos_token_id=forced_bos_token_id,
+        max_length=400,
+    )
+    return tokenizer.batch_decode(generated_tokens, skip_special_tokens=True)[0]
+
+# ---------------------------------------------------------
+# En-tête
+st.title("🌐 Language Translation Tool")
+st.caption("CodeAlpha Artificial Intelligence Internship — Task 1")
+
+st.write(
+    "Entrez votre texte, choisissez la langue source et la langue cible, "
+    "puis cliquez sur **Traduire**."
+)
+
+# ---------------------------------------------------------
+# Zone de saisie
+input_text = st.text_area("Texte à traduire :", height=150, placeholder="Écrivez ou collez votre texte ici...")
+
+col1, col2 = st.columns(2)
+with col1:
+    source_options = ["auto"] + list(LANGUAGES.keys()) if LANGDETECT_AVAILABLE else list(LANGUAGES.keys())
+    source_lang = st.selectbox(
+        "Langue source",
+        options=source_options,
+        format_func=lambda code: "Détection automatique" if code == "auto" else LANGUAGES[code],
+        index=0,
+    )
+with col2:
+    # Langue cible par défaut : anglais si la source n'est pas anglais, sinon français
+    target_options = [c for c in LANGUAGES.keys() if c != "auto"]
+    target_lang = st.selectbox(
+        "Langue cible",
+        options=target_options,
+        format_func=lambda code: LANGUAGES[code],
+        index=target_options.index("fr") if "fr" in target_options else 0,
+    )
+
+translate_clicked = st.button("🔄 Traduire", type="primary", use_container_width=True)
+
+# ---------------------------------------------------------
+# Logique de traduction
+if "translated_text" not in st.session_state:
+    st.session_state.translated_text = ""
+
+if translate_clicked:
+    if not input_text.strip():
+        st.warning("Merci d'entrer du texte avant de traduire.")
+    else:
+        try:
+            # --- Détection automatique de la langue source si demandé ---
+            actual_source = source_lang
+            if source_lang == "auto":
+                if LANGDETECT_AVAILABLE:
+                    detected = detect_lang(input_text)
+                    # langdetect renvoie parfois "zh-cn" au lieu de "zh-CN"
+                    detected = "zh-CN" if detected.lower().startswith("zh") else detected
+                    actual_source = detected if detected in LANGUAGES else "en"
+                else:
+                    actual_source = "en"
+
+            src_code = NLLB_CODES.get(actual_source)
+            tgt_code = NLLB_CODES.get(target_lang)
+
+            if src_code is None or tgt_code is None:
+                st.error("Langue non supportée par le modèle.")
+            else:
+                with st.spinner("Traduction en cours (inférence du modèle)..."):
+                    result = translate_text(input_text, src_code, tgt_code)
+                    st.session_state.translated_text = result
+                    if source_lang == "auto":
+                        st.caption(f"🔍 Langue détectée : {LANGUAGES.get(actual_source, actual_source)}")
+        except Exception as e:
+            st.error(f"Une erreur est survenue lors de la traduction : {e}")
+
+# ---------------------------------------------------------
+# Affichage du résultat
+if st.session_state.translated_text:
+    st.subheader("Traduction :")
+    st.text_area(
+        "Résultat",
+        value=st.session_state.translated_text,
+        height=150,
+        label_visibility="collapsed",
+    )
+
+    col_a, col_b = st.columns(2)
+
+    # --- Bouton copier (via un petit composant HTML/JS) ---
+    with col_a:
+        copy_html = f"""
+        <textarea id="toCopy" style="display:none;">{st.session_state.translated_text}</textarea>
+        <button onclick="navigator.clipboard.writeText(document.getElementById('toCopy').value)"
+            style="width:100%; padding:8px; border-radius:6px; border:1px solid #ccc; cursor:pointer;">
+            📋 Copier le texte
+        </button>
+        """
+        st.components.v1.html(copy_html, height=45)
+
+    # --- Bouton text-to-speech ---
+    with col_b:
+        if target_lang in TTS_SUPPORTED:
+            if st.button("🔊 Écouter la traduction", use_container_width=True):
+                try:
+                    tts = gTTS(text=st.session_state.translated_text, lang=target_lang)
+                    audio_bytes = io.BytesIO()
+                    tts.write_to_fp(audio_bytes)
+                    audio_bytes.seek(0)
+                    st.audio(audio_bytes, format="audio/mp3")
+                except Exception as e:
+                    st.error(f"Impossible de générer l'audio : {e}")
+        else:
+            st.info("🔇 Audio non disponible pour cette langue.")
 
-num_points = st.slider("Number of points in spiral", 1, 10000, 1100)
-num_turns = st.slider("Number of turns in spiral", 1, 300, 31)
-
-indices = np.linspace(0, 1, num_points)
-theta = 2 * np.pi * num_turns * indices
-radius = indices
-
-x = radius * np.cos(theta)
-y = radius * np.sin(theta)
-
-df = pd.DataFrame({
-    "x": x,
-    "y": y,
-    "idx": indices,
-    "rand": np.random.randn(num_points),
-})
-
-st.altair_chart(alt.Chart(df, height=700, width=700)
-    .mark_point(filled=True)
-    .encode(
-        x=alt.X("x", axis=None),
-        y=alt.Y("y", axis=None),
-        color=alt.Color("idx", legend=None, scale=alt.Scale()),
-        size=alt.Size("rand", legend=None, scale=alt.Scale(range=[1, 150])),
-    ))
+# ---------------------------------------------------------
+# Pied de page
+st.divider()
+st.caption("Projet réalisé dans le cadre du stage AI @CodeAlpha")