import json
import re
import gradio as gr

# ----------------------------
# Load Kent rubric titles
# ----------------------------
with open("titles.json", "r", encoding="utf-8") as f:
    RAW_TITLES = json.load(f)

def normalize(text: str) -> str:
    text = text.lower()
    text = re.sub(r"[^a-z0-9\s]", " ", text)
    text = re.sub(r"\s+", " ", text).strip()
    return text

TITLES = [
    {
        "original": title,
        "norm": normalize(title)
    }
    for title in RAW_TITLES
]

# ----------------------------
# Search function
# ----------------------------
def search(query, top_k=5):
    q = normalize(query)

    if not q:
        return []

    scored = []
    for item in TITLES:
        score = 0.0

        # Strong signal: substring match
        if q in item["norm"]:
            score += 1.0

        # Weaker signal: token overlap
        q_tokens = set(q.split())
        t_tokens = set(item["norm"].split())
        overlap = len(q_tokens & t_tokens)
        score += overlap * 0.1

        if score > 0:
            scored.append((score, item["original"]))

    scored.sort(reverse=True)

    return [
        {
            "rubric": title,
            "score": round(score, 3)
        }
        for score, title in scored[:int(top_k)]
    ]

# ----------------------------
# Gradio interface (API host)
# ----------------------------
api = gr.Interface(
    fn=search,
    inputs=[
        gr.Textbox(label="query"),
        gr.Number(label="top_k", value=5)
    ],
    outputs="json",
    api_name="search"   # 🔴 THIS IS THE KEY LINE
)

if __name__ == "__main__":
    api.launch()