app.py

"""
Dispatch AI — Arabic Proverb Generator
Input: topic → Output: Arabic proverb in traditional style + English translation.
Uses Qwen2.5-7B via HF Inference API.
"""

import os
import json
import gradio as gr
from huggingface_hub import InferenceClient

# --- Configuration -----------------------------------------------------------
HF_TOKEN = os.environ.get("HF_TOKEN", None)
MODEL_ID = "Qwen/Qwen2.5-7B-Instruct"

client = InferenceClient(model=MODEL_ID, token=HF_TOKEN)

BG_COLOR = "#0A0F1A"
ACCENT = "#1FE0E6"

# Preset topics
PRESET_TOPICS = [
    "patience",
    "knowledge",
    "friendship",
    "honesty",
    "hard work",
    "wisdom",
    "family",
    "courage",
    "generosity",
    "time",
    "hope",
    "unity",
    "travel",
    "mother",
    "neighbor",
]


def generate_proverb(topic, style):
    """Generate an Arabic proverb using Qwen2.5-7B via HF Inference API."""
    if not topic or not topic.strip():
        topic = "wisdom"

    style_instruction = {
        "Classical": "in the style of classical Arabic literature, like ancient Bedouin wisdom",
        "Poetic": "in a poetic, rhyming style with rhythm (saja')",
        "Simple": "in simple, everyday Arabic that anyone can understand",
        "Bedouin": "in the style of Bedouin desert wisdom, referencing desert life and nature",
        "Royal": "in the style of royal court wisdom, grand and majestic",
    }.get(style, "in the style of classical Arabic literature")

    system_prompt = (
        f"You are an expert in Arabic culture and literature. "
        f"Generate a traditional Arabic proverb about '{topic}' {style_instruction}. "
        f"Respond ONLY in valid JSON format with these exact keys:\n"
        f'{{"arabic": "the proverb in Arabic", "english": "English translation", '
        f'"transliteration": "Arabic in Latin script", "explanation": "brief explanation of meaning"}}'
    )

    try:
        response = client.chat_completion(
            messages=[
                {"role": "system", "content": system_prompt},
                {"role": "user", "content": f"Generate a proverb about: {topic}"},
            ],
            max_tokens=300,
            temperature=0.8,
        )
        raw = response.choices[0].message.content.strip()

        # Try to parse JSON
        try:
            # Extract JSON from response (may have markdown code blocks)
            if "```json" in raw:
                raw = raw.split("```json")[1].split("```")[0].strip()
            elif "```" in raw:
                raw = raw.split("```")[1].split("```")[0].strip()
            data = json.loads(raw)
        except (json.JSONDecodeError, IndexError):
            # Fallback: use raw text as Arabic proverb
            data = {
                "arabic": raw,
                "english": "(Translation unavailable)",
                "transliteration": "",
                "explanation": "",
            }

        arabic = data.get("arabic", "—")
        english = data.get("english", "—")
        transliteration = data.get("transliteration", "—")
        explanation = data.get("explanation", "—")

        result = f"""
### 📜 Arabic Proverb

**{arabic}**

---

### 🌐 English Translation

*{english}*

---

### 🔤 Transliteration

{transliteration}

---

### 💡 Meaning

{explanation}

---

*Topic: {topic} · Style: {style} · Model: {MODEL_ID}*
"""
        return result, "✅ Proverb generated!"

    except Exception as e:
        return f"❌ Error: {str(e)}", f"❌ Error: {str(e)}"


def generate_multiple_proverbs(topic, style, count):
    """Generate multiple proverbs about a topic."""
    results = []
    n = int(count) if count else 3
    for i in range(min(n, 5)):
        result, status = generate_proverb(topic, style)
        results.append(f"### Proverb {i+1}\n\n{result}\n\n---\n")
    return "\n".join(results), "✅ Generated!"


# --- UI -----------------------------------------------------------------------
CSS = """
#dispatch-header h1 {
    color: #FFFFFF; font-size: 2.2rem; margin: 0;
    background: linear-gradient(90deg, #1FE0E6 0%, #FFFFFF 60%);
    -webkit-background-clip: text; -webkit-text-fill-color: transparent;
}
#dispatch-header p { color: #1FE0E6; font-size: 1.05rem; margin: 6px 0 0 0; }
.dispatch-footer { text-align: center; color: #8A8F9C; font-size: 0.9rem; padding-top: 8px; }
"""

with gr.Blocks(
    title="Dispatch AI — Arabic Proverb Generator",
    theme=gr.themes.Base(
        primary_hue="cyan", secondary_hue="cyan", neutral_hue="slate",
        font=[gr.themes.GoogleFont("Inter"), "ui-sans-serif", "system-ui"],
    ).set(
        body_background_fill="#0A0F1A", body_background_fill_dark="#0A0F1A",
        body_text_color="#FFFFFF", body_text_color_dark="#FFFFFF",
        block_background_fill="#0E1424", block_background_fill_dark="#0E1424",
        block_border_color="#1FE0E6", block_border_width="1px",
        block_label_text_color="#1FE0E6", block_title_text_color="#1FE0E6",
        button_primary_background_fill="#1FE0E6", button_primary_background_fill_dark="#1FE0E6",
        button_primary_text_color="#0A0F1A", button_primary_border_color="#1FE0E6",
        input_background_fill="#0E1424", input_background_fill_dark="#0E1424",
        input_border_color="#1FE0E6", input_border_width="1px",
    ),
    css=CSS,
) as demo:
    with gr.Column(elem_id="dispatch-header"):
        gr.Markdown(
            """
            # Dispatch AI — Arabic Proverb Generator
            Generate traditional Arabic proverbs + English translation · Qwen2.5-7B · Dispatch AI (FZE) · UAE
            """
        )

    with gr.Row():
        with gr.Column(scale=1):
            topic_input = gr.Textbox(
                label="Topic",
                placeholder="e.g. patience, friendship, knowledge...",
                value="patience",
                lines=1,
            )
            style_select = gr.Radio(
                ["Classical", "Poetic", "Simple", "Bedouin", "Royal"],
                label="Style", value="Classical",
            )
            generate_btn = gr.Button("📜 Generate Proverb", variant="primary")
            gr.Markdown("### Quick Topics")
            topic_buttons = gr.Dataset(
                label="Preset Topics",
                components=[topic_input],
                samples=[[t] for t in PRESET_TOPICS],
            )
            with gr.Accordion("Generate Multiple", open=False):
                count_slider = gr.Slider(1, 5, value=3, step=1, label="Number of Proverbs")
                multi_btn = gr.Button("📚 Generate Multiple Proverbs", variant="secondary")

        with gr.Column(scale=2):
            status_box = gr.Textbox(label="Status", interactive=False)
            output_md = gr.Markdown()

    # Events
    generate_btn.click(
        generate_proverb,
        inputs=[topic_input, style_select],
        outputs=[output_md, status_box],
    )
    multi_btn.click(
        generate_multiple_proverbs,
        inputs=[topic_input, style_select, count_slider],
        outputs=[output_md, status_box],
    )

    gr.Markdown(
        """
        <div class="dispatch-footer">
        © 2026 Dispatch AI (FZE) · UAE · License 10818 · Model: Qwen2.5-7B-Instruct via HF Inference API
        </div>
        """
    )

if __name__ == "__main__":
    demo.queue()
    demo.launch()