"""Dashboard expander and section components."""

import html
import streamlit as st

from core.charts import EMOTION_KO
from core.utils import get_confidence_tier, truncate_text


# Content type labels for citations
CONTENT_TYPE_LABELS = {
    "EDITORIAL": "📰 에디토리얼",
    "TUTORIAL_REVIEW": "📝 리뷰/튜토리얼",
    "COMPARISON": "⚖️ 비교 분석",
    "RANKED_LIST": "📊 순위 목록",
    "FORUM_THREAD": "💬 포럼/커뮤니티",
    "HOMEPAGE": "🏠 홈페이지",
    "CATALOG": "📦 카탈로그",
    "DOCUMENTATION": "📚 문서",
    "FAQ": "❓ FAQ",
    "WHITEPAPER": "📄 백서",
    "PRESS_RELEASE": "📢 보도자료",
    "CASE_STUDY": "💼 사례연구",
    "PRICING": "💰 가격정보",
    "DETAIL": "🔍 상세페이지",
    "DIRECTORY_ENTRY": "📋 디렉토리",
    "SUBSTITUTE": "🔄 대체제",
    "OTHERS": "📎 기타",
}


def render_citation(cit: dict) -> None:
    """Render a single citation item.

    Args:
        cit: Citation dict with source_url, content_type, page_title
    """
    url = cit.get("source_url", "")
    ctype = cit.get("content_type") or "OTHERS"
    title = cit.get("page_title") or ""
    type_label = CONTENT_TYPE_LABELS.get(ctype, f"📎 {ctype}")
    display_url = url[:50] + "..." if len(url) > 50 else url
    display_title = f' "{title[:30]}..."' if title and len(title) > 30 else f' "{title}"' if title else ""
    st.markdown(
        f'<span style="background: #E0E7FF; color: #3730A3; padding: 2px 6px; '
        f'border-radius: 4px; font-size: 11px; margin-right: 4px;">{type_label}</span> '
        f'<a href="{url}" target="_blank">{display_url}</a>{display_title}',
        unsafe_allow_html=True
    )


def render_nudge_expander(
    item: dict,
    answer_id: int | None,
    index: int,
    fetch_full_answer_fn,
    fetch_citations_fn,
) -> None:
    """Render nudge candidate expander with full details.

    Args:
        item: Nudge candidate data dict
        answer_id: Answer ID for Athena fetch
        index: Item index for display
        fetch_full_answer_fn: Function to fetch full answer from Athena
        fetch_citations_fn: Function to fetch citations (Supabase fallback)
    """
    confidence = item.get("overall_confidence", 0) or 0
    tier, _, _ = get_confidence_tier(confidence)
    emotion = item.get("dominant_emotion", "N/A")
    emotion_ko = EMOTION_KO.get(emotion, emotion) if emotion else "N/A"
    answer = item.get("answer_preview", "")
    brand_detail = item.get("brand_sentiment_detail", {})

    with st.expander(f"📖 상세 보기 (답변 #{answer_id or index+1})"):
        # Analysis explanation box
        st.markdown(f"""
<div style="background: #FFF7ED; border-left: 4px solid #F59E0B; padding: 12px; margin-bottom: 12px; border-radius: 0 8px 8px 0; font-size: 13px;">
<strong>📊 분석 결과 해석</strong><br><br>
<strong>📄 답변 전체 부정 확신도: {confidence:.0%} ({tier})</strong><br>
답변 전체가 부정적인 톤인지 판단한 점수입니다. (여러 브랜드가 언급되면 혼합됨)<br><br>
<strong>🔍 브랜드별 부정 확신도</strong> (아래 ABSA 참조)<br>
특정 브랜드에 대한 언급만 추출하여 그 언급이 부정적인지 판단한 점수입니다.<br>
<em style="color: #9CA3AF;">예: 답변 전체는 64%(LOW)여도, 특정 브랜드 언급은 91%(HIGH)일 수 있음</em><br><br>
<strong>답변 톤: {emotion_ko}</strong><br>
답변 전체의 감정적 분위기입니다.
</div>
""", unsafe_allow_html=True)

        # Full answer from Athena
        st.markdown("**🤖 AI 답변 전문**")

        if answer_id:
            full_answer_key = f"full_answer_{answer_id}"
            load_full_key = f"load_full_{answer_id}"
            if full_answer_key not in st.session_state:
                st.session_state[full_answer_key] = None

            load_full = st.checkbox(
                "📥 전체 답변 불러오기",
                key=load_full_key,
                value=st.session_state.get(full_answer_key) is not None
            )

            if load_full and st.session_state.get(full_answer_key) is None:
                with st.spinner("전체 답변을 가져오는 중..."):
                    full_content = fetch_full_answer_fn(answer_id)
                    if isinstance(full_content, str) and len(full_content) > 0:
                        st.session_state[full_answer_key] = full_content
                        st.rerun()
                    else:
                        # Store empty string to prevent infinite re-fetch loop
                        st.session_state[full_answer_key] = ""

            cached = st.session_state.get(full_answer_key)
            display_answer = cached if (isinstance(cached, str) and len(cached) > 0) else answer or "N/A"
            is_full = isinstance(cached, str) and len(cached) > 0
            label = "✅ 전체 답변 로드됨" if is_full else f"📄 미리보기 ({len(answer or '')}자)"
            st.caption(label)
        else:
            display_answer = answer or "N/A"

        st.markdown(
            f'<div style="background: #FEF2F2; padding: 12px; border-radius: 8px; '
            f'font-size: 14px; white-space: pre-wrap; word-break: break-word; '
            f'max-height: 400px; overflow-y: auto;">{html.escape(display_answer)}</div>',
            unsafe_allow_html=True
        )

        # Brand sentiment detail
        if brand_detail and isinstance(brand_detail, dict):
            st.markdown("**🔍 브랜드별 감성 분석 (ABSA) - 브랜드별 부정 확신도**")
            _render_brand_absa(brand_detail)

        # Citations
        st.markdown("**🔗 인용 출처 (Citation Sources)**")
        _render_citations_section(answer_id, item.get("citation_urls", []), fetch_citations_fn)


def _render_brand_absa(brand_detail: dict) -> None:
    """Render brand ABSA results."""
    in_house_data = brand_detail.get("in_house", {})
    in_house_absa = in_house_data.get("absa_results", [])
    for absa in in_house_absa:
        if isinstance(absa, dict):
            brand_name = absa.get("brand", "Unknown")
            sentiment = absa.get("sentiment", "N/A")
            conf = absa.get("confidence", 0)
            absa_tier, absa_emoji, _ = get_confidence_tier(conf)
            sent_color = "#10B981" if sentiment == "positive" else "#EF4444" if sentiment == "negative" else "#6B7280"
            st.markdown(
                f'<span style="background: {sent_color}; color: white; padding: 2px 8px; '
                f'border-radius: 4px; font-size: 12px; margin-right: 8px;">{sentiment}</span> '
                f'<strong>{brand_name}</strong> (🏠 자사) - {absa_emoji} 브랜드 확신도 {conf:.0%} ({absa_tier})',
                unsafe_allow_html=True
            )

    competitor_data = brand_detail.get("competitor", {})
    competitor_brands = competitor_data.get("brands", [])
    competitor_absa = competitor_data.get("absa_results", [])

    if competitor_absa:
        for absa in competitor_absa:
            if isinstance(absa, dict):
                brand_name = absa.get("brand", "Unknown")
                sentiment = absa.get("sentiment", "N/A")
                conf = absa.get("confidence", 0)
                absa_tier, absa_emoji, _ = get_confidence_tier(conf)
                sent_color = "#10B981" if sentiment == "positive" else "#EF4444" if sentiment == "negative" else "#6B7280"
                st.markdown(
                    f'<span style="background: {sent_color}; color: white; padding: 2px 8px; '
                    f'border-radius: 4px; font-size: 12px; margin-right: 8px;">{sentiment}</span> '
                    f'<strong>{brand_name}</strong> (🏢 경쟁사) - {absa_emoji} 브랜드 확신도 {conf:.0%} ({absa_tier})',
                    unsafe_allow_html=True
                )
    elif competitor_brands:
        st.markdown(
            f'<span style="background: #6B7280; color: white; padding: 2px 8px; '
            f'border-radius: 4px; font-size: 12px;">언급됨</span> '
            f'<strong>{", ".join(competitor_brands)}</strong> (🏢 경쟁사)',
            unsafe_allow_html=True
        )


def _render_citations_section(answer_id: int | None, citation_urls: list, fetch_citations_fn) -> None:
    """Render citations section."""
    citations_key = f"citations_{answer_id}"
    if citations_key not in st.session_state:
        st.session_state[citations_key] = None

    if st.session_state.get(citations_key) is None and answer_id:
        citations = fetch_citations_fn(answer_id)
        st.session_state[citations_key] = citations if citations else []

    citations = st.session_state.get(citations_key, [])
    if citations:
        if len(citations) <= 5:
            for cit in citations:
                render_citation(cit)
        else:
            for cit in citations[:5]:
                render_citation(cit)
            with st.expander(f"📂 나머지 {len(citations) - 5}개 더 보기"):
                for cit in citations[5:]:
                    render_citation(cit)
    elif citation_urls:
        if len(citation_urls) <= 5:
            for url in citation_urls:
                st.markdown(f"• [{url[:60]}...]({url})" if len(url) > 60 else f"• [{url}]({url})")
        else:
            for url in citation_urls[:5]:
                st.markdown(f"• [{url[:60]}...]({url})" if len(url) > 60 else f"• [{url}]({url})")
            with st.expander(f"📂 나머지 {len(citation_urls) - 5}개 더 보기"):
                for url in citation_urls[5:]:
                    st.markdown(f"• [{url[:60]}...]({url})" if len(url) > 60 else f"• [{url}]({url})")
    else:
        st.caption("인용 소스 없음")


def render_feedback_section(feedback_stats: dict) -> None:
    """Render feedback statistics expander section.

    Args:
        feedback_stats: Dict with feedback counts and accuracy
    """
    from .metrics import render_feedback_stats

    fb_total = feedback_stats.get("total_feedback", 0)
    if fb_total > 0:
        with st.expander("📝 **피드백 분석** - 사용자 검증 현황", expanded=False):
            render_feedback_stats(feedback_stats)


def render_llm_verification_section(item: dict, is_false_positive: bool = True) -> None:
    """Render LLM verification item section (used inside expander).

    This is a wrapper that calls render_verification_item from cards module.

    Args:
        item: Verification result dict
        is_false_positive: True for FP, False for TN
    """
    from .cards import render_verification_item
    render_verification_item(item, is_false_positive)