#!/bin/bash
# ==============================================================================================
# =========================== Playback-Aware Lidarr Discovery ==================================
# ==============================================================================================
#
# PURPOSE
# ─────────────────────────────────────────────────────────────────────────────
# Two-stage behavior-driven music discovery.
#
# Stage 1 — Score what you actually played this week. The top artists become
#            high-quality seeds, not just anything that hit the minimum play count.
#
# Stage 2 — Run Last.fm artist.getSimilar on those seeds. Score the recommendations
#            and add the top artists to Lidarr.
#
# Goal: 0–5 meaningful Lidarr adds per week, not bulk imports.
#
# ==============================================================================================
# FLOW
# ==============================================================================================
#
#   1. Fetch play completions from Emby activity log (last LOOKBACK_DAYS days)
#   2. Aggregate by artist — apply per-user influence cap to play weights
#
#   ── Stage 1 ─────────────────────────────────────────────────────────────────
#   3. Score each played artist: user signal + recency + Last.fm popularity/quality
#   4. Take top MAX_ADDS by score → discovery seeds
#
#   ── Stage 2 ─────────────────────────────────────────────────────────────────
#   5. For each seed, call Last.fm artist.getSimilar → collect candidates
#   6. Aggregate: affinity (weighted similarity × seed score) + breadth (distinct seeds)
#   7. Filter: already in Lidarr, already in Emby, placeholder artists, cooldown
#   8. Score candidates: affinity + breadth + popularity + quality
#   9. Take top MAX_ADDS above threshold → add to Lidarr
#
# Step 7's "already in Lidarr" check reads the shared tracked-data cache via
# arr_get_tracked_data() (cache-first, live fallback, 2026-07-17) instead of a live fetch —
# this runs weekly right after arr_full_rescan.sh, so it's reading the genuine post-rescan
# snapshot arr_full_rescan.sh just wrote.
#
# ==============================================================================================
# SCORING MODEL
# ==============================================================================================
#
#   Stage 1 — seed selection (max 100)
#     user_score       (0-40) — effective plays × 5, cap 40
#     recency_score    (0-30) — days since last play; today→30, older→less
#     popularity_score (0-20) — Last.fm listeners
#     quality_score    (0-10) — Last.fm global playcount
#
#   Stage 2 — candidate scoring (max 100)
#     affinity_score   (0-40) — seed_score × similarity, normalized; raw/20, cap 40
#     breadth_score    (0-30) — 1 seed→5, 2 seeds→18, 3+seeds→30
#     popularity_score (0-20) — Last.fm listeners
#     quality_score    (0-10) — Last.fm global playcount
#
#   Threshold: LIDARR_DISCOVERY_THRESHOLD (default 70) applied to both stages
#   Max adds:  LIDARR_DISCOVERY_MAX_ADDS (default 5) caps each stage
#
# ==============================================================================================
# REQUIREMENTS
# ==============================================================================================
#
#   Last.fm API key — required for both stages
#   Configure HOST*_LASTFM_API_KEY in host*.conf
#
# ==============================================================================================
# DESIGN PRINCIPLES
# ==============================================================================================
#
# Playback as Intent Signal
#   What users actually listen to is a stronger signal than what they follow or
#   own. The scoring model weights demonstrated listening behaviour — recency,
#   play count, user breadth — over passive library membership.
#
# Selective by Design
#   0–5 adds per week is the target, not bulk imports. A high score threshold
#   combined with MAX_ADDS ensures only high-confidence recommendations are
#   acted on. Volume is not the goal — meaningful discovery is.
#
# Two-Stage Filtering
#   Stage 1 rejects weak seeds before they drive Stage 2. Low-quality seeds
#   produce low-quality similar-artist recommendations. Filtering at the seed
#   stage improves the entire output, not just the top of the list.
#
# ==============================================================================================
# OPERATIONAL SAFEGUARDS
# ==============================================================================================
#
# Root Required
#   acquire_lock and Lidarr API writes require root. Script exits cleanly if not root.
#
# Dry-Run Mode
#   --dry-run scores and ranks all candidates but makes no Lidarr API calls and does not
#   write to the history file. Safe to run at any time to preview what would be added.
#
# Add-Only
#   Only adds artists to Lidarr. Never deletes or modifies existing entries.
#
# Cooldown Guard
#   Candidates rejected this run are recorded in the history file and not re-evaluated
#   until LIDARR_DISCOVERY_REJECT_COOLDOWN days have elapsed.
#
# ==============================================================================================
# STATE FILES
# ==============================================================================================
#
#   LIDARR_DISCOVERY_HISTORY (default: $DATA_DIR/lidarr_discovery_history.db)
#     Tracks added artists and rejected candidates with timestamps. Written after every
#     non-dry-run. Enforces reject cooldown and prevents re-adding items added by previous
#     runs. Safe to delete — next run starts fresh with no memory.
#
# ==============================================================================================
# CONFIGURATION (master.conf)
# ==============================================================================================
#
#   LIDARR_DISCOVERY_THRESHOLD       — minimum score for Stage 1 seeds and Stage 2 adds (default: 70)
#   LIDARR_DISCOVERY_LOOKBACK_DAYS   — Emby play history window in days (default: 7)
#   LIDARR_DISCOVERY_MIN_PLAYS       — min plays to be evaluated in Stage 1 (default: 3)
#   LIDARR_DISCOVERY_MAX_ADDS        — max seeds (Stage 1) and max adds (Stage 2) (default: 5)
#   LIDARR_DISCOVERY_USER_CAP_PCT    — max % any one user contributes to play weight (default: 35)
#   LIDARR_DISCOVERY_REJECT_COOLDOWN — days before re-evaluating a Stage 2 reject (default: 30)
#   LIDARR_DISCOVERY_HISTORY         — history/state file path
#
# ==============================================================================================
# RUNTIME MODES
# ==============================================================================================
#
#   playback_aware_lidarr_discovery.sh            — normal run
#   playback_aware_lidarr_discovery.sh --dry-run  — score and rank, no Lidarr changes
#   playback_aware_lidarr_discovery.sh --log      — verbose output
#   playback_aware_lidarr_discovery.sh --status   — show config and exit
#
# Recommended schedule: weekly (WEEKLY_MAINTENANCE_SCRIPTS in master.conf)
#
# ==============================================================================================

SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"

source "$SCRIPT_DIR/../load_config.sh"
source "$SCRIPT_DIR/../Kernel/decision_engine.sh"

parse_args "$@"

# ==============================================================================================
# ━━━ Setup ━━━
# ==============================================================================================
if [[ "$EUID" -ne 0 ]]; then
    error "Must be run as root"
    exit 1
fi

acquire_lock
detect_hosts

if ! command -v jq >/dev/null 2>&1; then
    error "jq not installed — required for API JSON parsing"
    exit 1
fi

if [[ -z "${EMBY_URL:-}" || -z "${EMBY_API_KEY:-}" ]]; then
    error "EMBY_URL / EMBY_API_KEY not configured — check host*.conf"
    exit 1
fi

if [[ -z "${LIDARR_URL:-}" || -z "${LIDARR_API_KEY:-}" ]]; then
    error "LIDARR_URL / LIDARR_API_KEY not configured — check host*.conf"
    exit 1
fi

if [[ -z "${LASTFM_API_KEY:-}" ]]; then
    error "LASTFM_API_KEY not configured — required for discovery"
    error "Configure HOST*_LASTFM_API_KEY in host*.conf"
    exit 1
fi

THRESHOLD="${LIDARR_DISCOVERY_THRESHOLD:-70}"
LOOKBACK_DAYS="${LIDARR_DISCOVERY_LOOKBACK_DAYS:-7}"
MIN_PLAYS="${LIDARR_DISCOVERY_MIN_PLAYS:-3}"
MAX_ADDS="${LIDARR_DISCOVERY_MAX_ADDS:-5}"
USER_CAP_PCT="${LIDARR_DISCOVERY_USER_CAP_PCT:-35}"
REJECT_COOLDOWN="${LIDARR_DISCOVERY_REJECT_COOLDOWN:-30}"
HISTORY_FILE="${LIDARR_DISCOVERY_HISTORY:-${DATA_DIR}/lidarr_discovery_history.db}"

log "$ICON_GEAR Config: threshold=${THRESHOLD} lookback=${LOOKBACK_DAYS}d min-plays=${MIN_PLAYS} max-adds=${MAX_ADDS} user-cap=${USER_CAP_PCT}% reject-cooldown=${REJECT_COOLDOWN}d"

[[ "$DRY_RUN" == true ]] && warn "DRY RUN — no artists will be added to Lidarr"

# ==============================================================================================
# ━━━ Status ━━━
# ==============================================================================================
if [[ "$SHOW_STATUS" == true ]]; then
    echo ""
    echo "━━━━━ $ICON_SUMMARY LIDARR DISCOVERY STATUS ━━━━━"
    echo "$ICON_HOST    Identity:      $MY_ID ($LOCAL_SERVER_NAME)"
    echo "$ICON_SYNC    Emby:          ${EMBY_URL}"
    echo "$ICON_SYNC    Lidarr:        ${LIDARR_URL}"
    echo "$ICON_GEAR    Threshold:     ${THRESHOLD} / 100 (both stages)"
    echo "$ICON_TIME    Lookback:      ${LOOKBACK_DAYS} days"
    echo "$ICON_GEAR    Min plays:     ${MIN_PLAYS}"
    echo "$ICON_GEAR    Max adds:      ${MAX_ADDS} (seeds in Stage 1, adds in Stage 2)"
    echo "$ICON_GEAR    User cap:      ${USER_CAP_PCT}% max per user (floor: 3 plays)"
    echo "$ICON_GEAR    Reject TTL:    ${REJECT_COOLDOWN} days"
    echo "$ICON_GEAR    History file:  ${HISTORY_FILE}"
    echo "$ICON_GEAR    Last.fm:       configured"
    echo "$ICON_GEAR    Dry Run:       $DRY_RUN"
    echo "━━━━━━━━━━━━━━━━━━━━━━━"
    exit 0
fi

# ==============================================================================================
# ── API HELPERS ───────────────────────────────────────────────────────────────────────────────
# ==============================================================================================


_lidarr_get() {
    curl -sf --max-time 20 \
        -H "X-Api-Key: $LIDARR_API_KEY" \
        "${LIDARR_URL}/api/v1/${1}" 2>/dev/null
}

_lidarr_lookup() {
    curl -sf --max-time 20 --get \
        --data-urlencode "term=$1" \
        -H "X-Api-Key: $LIDARR_API_KEY" \
        "${LIDARR_URL}/api/v1/artist/lookup" 2>/dev/null
}

_lidarr_post() {
    curl -sf --max-time 20 -X POST \
        -H "X-Api-Key: $LIDARR_API_KEY" \
        -H "Content-Type: application/json" \
        -d "$1" \
        "${LIDARR_URL}/api/v1/artist" 2>/dev/null
}

_lastfm_info() {
    curl -sf --max-time 10 --get \
        --data-urlencode "method=artist.getinfo" \
        --data-urlencode "artist=$1" \
        --data-urlencode "api_key=$LASTFM_API_KEY" \
        --data-urlencode "format=json" \
        "http://ws.audioscrobbler.com/2.0/" 2>/dev/null
}

_lastfm_similar() {
    curl -sf --max-time 15 --get \
        --data-urlencode "method=artist.getSimilar" \
        --data-urlencode "artist=$1" \
        --data-urlencode "limit=10" \
        --data-urlencode "api_key=$LASTFM_API_KEY" \
        --data-urlencode "format=json" \
        "http://ws.audioscrobbler.com/2.0/" 2>/dev/null
}

# ==============================================================================================
# ── SCORING HELPERS ───────────────────────────────────────────────────────────────────────────
# ==============================================================================================

_user_score() {
    local s=$(( $1 * 5 ))
    (( s > 40 )) && s=40
    echo "$s"
}

_recency_score() {
    local days="$1"
    if   (( days == 0 )); then echo 30
    elif (( days == 1 )); then echo 25
    elif (( days == 2 )); then echo 20
    elif (( days == 3 )); then echo 15
    elif (( days == 4 )); then echo 10
    elif (( days == 5 )); then echo 8
    else                       echo 5
    fi
}

_affinity_score() {
    local s=$(( $1 / 20 ))
    (( s > 40 )) && s=40
    echo "$s"
}

_breadth_score() {
    local seeds="$1"
    if   (( seeds >= 3 )); then echo 30
    elif (( seeds == 2 )); then echo 18
    else                        echo 5
    fi
}

_popularity_score() {
    local listeners="$1"
    if   (( listeners >= 5000000 )); then echo 20
    elif (( listeners >= 1000000 )); then echo 15
    elif (( listeners >= 500000  )); then echo 10
    elif (( listeners >= 100000  )); then echo 5
    else                                  echo 2
    fi
}

_quality_score() {
    local playcount="$1"
    if   (( playcount >= 100000000 )); then echo 10
    elif (( playcount >= 10000000  )); then echo 7
    elif (( playcount >= 1000000   )); then echo 4
    else                                    echo 2
    fi
}

_is_placeholder_artist() {
    local a="${1,,}"
    [[ "$a" =~ ^(va|various|various artists|unknown artist|unknown|soundtrack|original soundtrack|ost)$ ]]
}

_lfm_scores() {
    local artist="$1"
    local LFM_JSON lfm_listeners lfm_playcount
    LFM_JSON=$(_lastfm_info "$artist")
    if [[ -n "$LFM_JSON" ]] && echo "$LFM_JSON" | jq -e '.artist' >/dev/null 2>&1; then
        lfm_listeners=$(echo "$LFM_JSON" | jq -r '.artist.stats.listeners // "0"' | tr -d ',')
        lfm_playcount=$(echo "$LFM_JSON" | jq -r '.artist.stats.playcount // "0"' | tr -d ',')
        lfm_listeners=${lfm_listeners//[^0-9]/}; lfm_listeners=${lfm_listeners:-0}
        lfm_playcount=${lfm_playcount//[^0-9]/}; lfm_playcount=${lfm_playcount:-0}
        echo "${lfm_listeners}|${lfm_playcount}"
    else
        echo "0|0"
    fi
}

# ==============================================================================================
# ━━━ Fetch Emby Play History ━━━
# ==============================================================================================
echo ""
echo "━━━━━ $ICON_SUMMARY Lidarr Discovery — $(date '+%Y-%m-%d %H:%M:%S') ━━━━━"
echo "$ICON_HOST $MY_ID ($LOCAL_SERVER_NAME) | lookback: ${LOOKBACK_DAYS}d | threshold: ${THRESHOLD}/100 | max: ${MAX_ADDS}"
[[ "$DRY_RUN" == true ]] && warn "DRY RUN — no artists will be added"

echo ""
echo "━━━ $ICON_SYNC Emby Play History ━━━"

CUTOFF_ISO=$(date -d "${LOOKBACK_DAYS} days ago" '+%Y-%m-%dT%H:%M:%SZ')
TODAY_EPOCH=$(date +%s)
TODAY=$(date +%Y-%m-%d)

ACTIVITY_JSON=$(emby_api "System/ActivityLog/Entries?MinDate=${CUTOFF_ISO}&Limit=5000") || {
    error "Could not fetch Emby activity log"
    exit 1
}

declare -A ARTIST_PLAYS
declare -A ARTIST_LAST_PLAY
declare -A ARTIST_USER_PLAYS
declare -A ARTIST_EFFECTIVE   # precomputed effective plays after user cap
TOTAL_PLAYS=0

while IFS='|' read -r play_date artist user_id; do
    [[ -z "$artist" || "$artist" == "null" ]] && continue
    _is_placeholder_artist "$artist" && continue

    (( TOTAL_PLAYS++ ))
    ARTIST_PLAYS["$artist"]=$(( ${ARTIST_PLAYS["$artist"]:-0} + 1 ))
    ARTIST_USER_PLAYS["${artist}|${user_id}"]=$(( ${ARTIST_USER_PLAYS["${artist}|${user_id}"]:-0} + 1 ))

    current="${ARTIST_LAST_PLAY["$artist"]:-}"
    if [[ -z "$current" || "$play_date" > "$current" ]]; then
        ARTIST_LAST_PLAY["$artist"]="$play_date"
    fi
done < <(echo "$ACTIVITY_JSON" | jq -r '
    .Items // [] | .[] |
    select(.Name | test("has finished playing"; "i")) |
    select(.Name | test(" - ")) |
    select(.Name | test(", Ep[0-9]") | not) |
    [
        .Date,
        (
            .Name |
            split(" has finished playing ")[1] |
            split(" on ") | .[0:-1] | join(" on ") |
            split(" - ")[0] | ltrimstr(" ") | rtrimstr(" ")
        ),
        (.UserId // "unknown")
    ] | join("|")
' 2>/dev/null)

# Precompute effective plays (user cap applied) for all qualifying artists
for _a in "${!ARTIST_PLAYS[@]}"; do
    _plays="${ARTIST_PLAYS["$_a"]}"
    (( _plays < MIN_PLAYS )) && continue
    _cap=$(( (_plays * USER_CAP_PCT + 99) / 100 ))
    (( _cap < 3 )) && _cap=3
    _eff=0
    for _ukey in "${!ARTIST_USER_PLAYS[@]}"; do
        [[ "$_ukey" == "${_a}|"* ]] || continue
        _uc="${ARTIST_USER_PLAYS["$_ukey"]}"
        (( _eff += _uc < _cap ? _uc : _cap ))
    done
    (( _eff == 0 )) && _eff="$_plays"
    ARTIST_EFFECTIVE["$_a"]="$_eff"
done

UNIQUE_ARTISTS=${#ARTIST_PLAYS[@]}
QUALIFIED=$(( $(echo "${!ARTIST_EFFECTIVE[@]}" | wc -w) ))
log "$TOTAL_PLAYS play completions | $UNIQUE_ARTISTS unique artists | $QUALIFIED qualify (${MIN_PLAYS}+ plays)"

if [[ "$QUALIFIED" -eq 0 ]]; then
    warn "No artists qualify this week — nothing to evaluate"
    exit 0
fi

# ==============================================================================================
# ━━━ Stage 1: Score Played Artists → Select Seeds ━━━
# ==============================================================================================
echo ""
echo "━━━ $ICON_GEAR Stage 1: Scoring Played Artists ━━━"

S1_SCORED=()  # "score|artist"

for artist in "${!ARTIST_EFFECTIVE[@]}"; do
    effective="${ARTIST_EFFECTIVE["$artist"]}"
    last_played="${ARTIST_LAST_PLAY["$artist"]}"
    last_epoch=$(date -d "$last_played" +%s 2>/dev/null || echo "$TODAY_EPOCH")
    days_ago=$(( (TODAY_EPOCH - last_epoch) / 86400 ))

    user_s=$(_user_score "$effective")
    recency_s=$(_recency_score "$days_ago")

    IFS='|' read -r lfm_listeners lfm_playcount <<< "$(_lfm_scores "$artist")"
    popularity_s=$(_popularity_score "$lfm_listeners")
    quality_s=$(_quality_score "$lfm_playcount")

    total=$(score_candidate "$user_s" "$popularity_s" "$recency_s" "$quality_s")
    total=$(apply_temporal_decay "$total" "$days_ago")

    log "  [${total}] $artist (plays: ${ARTIST_PLAYS["$artist"]}→${effective} | ${days_ago}d ago)"
    S1_SCORED+=("${total}|${artist}")
done

IFS=$'\n' S1_SORTED=($(printf '%s\n' "${S1_SCORED[@]}" | sort -t'|' -k1 -rn 2>/dev/null))

echo ""
echo "━━━ $ICON_SUMMARY Stage 1 Results (top ${MAX_ADDS} seeds) ━━━"

SEEDS=()
SEED_SCORES=()  # parallel array: score for each seed (used as Stage 2 weight)

for _entry in "${S1_SORTED[@]}"; do
    (( ${#SEEDS[@]} >= MAX_ADDS )) && break
    _score="${_entry%%|*}"
    _artist="${_entry#*|}"
    (( _score < THRESHOLD )) && break  # sorted desc — below threshold means rest are too
    SEEDS+=("$_artist")
    SEED_SCORES+=("$_score")
    printf "  Seed [%3s] %s (plays: %s)\n" "$_score" "$_artist" "${ARTIST_PLAYS["$_artist"]}"
done

if [[ "${#SEEDS[@]}" -eq 0 ]]; then
    warn "No played artists scored above ${THRESHOLD} this week — no seeds for discovery"
    exit 0
fi

log "${#SEEDS[@]} seed(s) selected"

# ==============================================================================================
# ━━━ Stage 2: Last.fm Similarity Discovery ━━━
# ==============================================================================================
echo ""
echo "━━━ $ICON_SYNC Stage 2: Last.fm Similarity Discovery ━━━"

declare -A CANDIDATE_SCORE
declare -A CANDIDATE_SEED_COUNT
declare -A CANDIDATE_LAST_SEED

for (( _i=0; _i<${#SEEDS[@]}; _i++ )); do
    seed_artist="${SEEDS[$_i]}"
    seed_score="${SEED_SCORES[$_i]}"
    seed_last_play="${ARTIST_LAST_PLAY["$seed_artist"]}"

    log "  Seed: $seed_artist (score: $seed_score)"

    SIMILAR_JSON=$(_lastfm_similar "$seed_artist")
    if [[ -z "$SIMILAR_JSON" ]]; then
        warn "  Last.fm similar unavailable for: $seed_artist"
        continue
    fi

    while IFS='|' read -r sim_name sim_match; do
        [[ -z "$sim_name" || "$sim_name" == "null" ]] && continue
        _is_placeholder_artist "$sim_name" && continue
        [[ "${sim_name,,}" == "${seed_artist,,}" ]] && continue

        sim_pct=$(echo "$sim_match" | awk '{printf "%d", $1 * 100 + 0.5}')
        (( sim_pct < 1 )) && sim_pct=1

        # Weight contribution by seed's Stage 1 score, not raw play count
        contribution=$(( seed_score * sim_pct ))
        CANDIDATE_SCORE["$sim_name"]=$(( ${CANDIDATE_SCORE["$sim_name"]:-0} + contribution ))
        CANDIDATE_SEED_COUNT["$sim_name"]=$(( ${CANDIDATE_SEED_COUNT["$sim_name"]:-0} + 1 ))

        existing_date="${CANDIDATE_LAST_SEED["$sim_name"]:-}"
        if [[ -z "$existing_date" || "$seed_last_play" > "$existing_date" ]]; then
            CANDIDATE_LAST_SEED["$sim_name"]="$seed_last_play"
        fi
    done < <(echo "$SIMILAR_JSON" | jq -r '
        .similarartists.artist // [] | .[] |
        [.name, .match] | join("|")
    ' 2>/dev/null)
done

CANDIDATE_COUNT=${#CANDIDATE_SCORE[@]}
log "$CANDIDATE_COUNT discovery candidates from Last.fm"

if [[ "$CANDIDATE_COUNT" -eq 0 ]]; then
    warn "No candidates returned from Last.fm — check API key or seed artist names"
    exit 0
fi

# ==============================================================================================
# ━━━ Fetch Existing Libraries ━━━
# ==============================================================================================
echo ""
echo "━━━ $ICON_SYNC Existing Libraries ━━━"

# Cache-first — arr_get_tracked_data() serves the shared cache when it's fresh (kept current
# every 30min by arr_cache_prefill.sh in CRITICAL_MAINTENANCE_SCRIPTS), falls back to a live
# fetch when it's stale, and waits out an active rescan before either.
LIDARR_ARTISTS_JSON=$(arr_get_tracked_data "lidarr" "$LIDARR_URL" "$LIDARR_API_KEY" "v1") || { error "Could not fetch Lidarr artists"; exit 1; }
LIDARR_NAMES=$(echo "$LIDARR_ARTISTS_JSON" | jq -r '.[].artistName' 2>/dev/null)
LIDARR_COUNT=$(echo "$LIDARR_NAMES" | grep -c . 2>/dev/null || echo 0)
log "$LIDARR_COUNT artists in Lidarr"

EMBY_LIBRARY_JSON=$(emby_api "Items?IncludeItemTypes=MusicAlbum&Recursive=true&Fields=AlbumArtists&Limit=10000") || {
    warn "Could not fetch Emby artist library — skipping Emby filter"
    EMBY_ARTIST_NAMES=""
}
EMBY_ARTIST_NAMES=$(echo "$EMBY_LIBRARY_JSON" | jq -r '.Items[] | .AlbumArtists[]?.Name' 2>/dev/null)
EMBY_ARTIST_COUNT=$(echo "$EMBY_ARTIST_NAMES" | grep -c . 2>/dev/null || echo 0)
log "$EMBY_ARTIST_COUNT album artists in Emby library"

# MusicBrainz's canonical name for some artists (e.g. "blink‐182") uses a Unicode
# hyphen/dash rather than plain ASCII "-". Last.fm's candidate names are plain ASCII,
# so an exact-string match against Lidarr/Emby's names silently misses these artists
# every time — they never register as "already known" and get retried (and rejected
# as duplicates) on every future run. Normalize both sides before comparing.
_normalize_dashes() {
    local n="$1"
    n="${n//‐/-}"  # U+2010 HYPHEN
    n="${n//‑/-}"  # U+2011 NON-BREAKING HYPHEN
    n="${n//‒/-}"  # U+2012 FIGURE DASH
    n="${n//–/-}"  # U+2013 EN DASH
    n="${n//—/-}"  # U+2014 EM DASH
    echo "$n"
}
LIDARR_NAMES=$(_normalize_dashes "$LIDARR_NAMES")
EMBY_ARTIST_NAMES=$(_normalize_dashes "$EMBY_ARTIST_NAMES")

_in_lidarr()      { echo "$LIDARR_NAMES"     | grep -iq "^$(_normalize_dashes "$1")$"; }
_in_emby_library(){ [[ -n "$EMBY_ARTIST_NAMES" ]] && echo "$EMBY_ARTIST_NAMES" | grep -iq "^$(_normalize_dashes "$1")$"; }

# ==============================================================================================
# ━━━ Score Stage 2 Candidates ━━━
# ==============================================================================================
echo ""
echo "━━━ $ICON_GEAR Stage 2: Scoring Candidates ━━━"

ACCEPT_LIST=()
REJECT_LIST=()
ALREADY_KNOWN=0
SKIP_COOLDOWN=0

for candidate in "${!CANDIDATE_SCORE[@]}"; do
    raw_score="${CANDIDATE_SCORE["$candidate"]}"
    seed_count="${CANDIDATE_SEED_COUNT["$candidate"]}"
    last_seed_date="${CANDIDATE_LAST_SEED["$candidate"]:-}"

    if _in_lidarr "$candidate"; then
        (( ALREADY_KNOWN++ ))
        log "  $ICON_SKIP In Lidarr: $candidate"
        continue
    fi

    if _in_emby_library "$candidate"; then
        (( ALREADY_KNOWN++ ))
        log "  $ICON_SKIP In Emby: $candidate"
        continue
    fi

    if [[ -f "$HISTORY_FILE" ]]; then
        last_rejection=$(grep -i "^REJECT|${candidate}|" "$HISTORY_FILE" 2>/dev/null | tail -1 | cut -d'|' -f3)
        if [[ -n "$last_rejection" ]]; then
            last_epoch=$(date -d "$last_rejection" +%s 2>/dev/null || echo 0)
            days_since=$(( (TODAY_EPOCH - last_epoch) / 86400 ))
            if (( days_since < REJECT_COOLDOWN )); then
                (( SKIP_COOLDOWN++ ))
                log "  $ICON_SKIP Cooldown (rejected ${days_since}d ago): $candidate"
                continue
            fi
        fi
    fi

    if [[ -n "$last_seed_date" ]]; then
        last_epoch=$(date -d "$last_seed_date" +%s 2>/dev/null || echo "$TODAY_EPOCH")
        days_ago=$(( (TODAY_EPOCH - last_epoch) / 86400 ))
    else
        days_ago=0
    fi

    affinity_s=$(_affinity_score "$raw_score")
    breadth_s=$(_breadth_score "$seed_count")

    IFS='|' read -r lfm_listeners lfm_playcount <<< "$(_lfm_scores "$candidate")"
    if (( lfm_listeners > 0 || lfm_playcount > 0 )); then
        lfm_label="${lfm_listeners} lfm listeners"
        popularity_s=$(_popularity_score "$lfm_listeners")
        quality_s=$(_quality_score "$lfm_playcount")
    else
        lfm_label="not on Last.fm"
        popularity_s=0
        quality_s=0
    fi

    total=$(score_candidate "$affinity_s" "$popularity_s" "$breadth_s" "$quality_s")
    total=$(apply_temporal_decay "$total" "$days_ago")
    decision=$(make_decision "$total" "$THRESHOLD")

    entry="${total}|${candidate}|${seed_count}|${affinity_s}|${lfm_label}"
    [[ "$decision" == "ACCEPT" ]] && ACCEPT_LIST+=("$entry") || REJECT_LIST+=("$entry")
done

# ==============================================================================================
# ━━━ Results ━━━
# ==============================================================================================
echo ""
echo "━━━ $ICON_SUMMARY Stage 2 Results (threshold: ${THRESHOLD}, max: ${MAX_ADDS}) ━━━"

IFS=$'\n' _ALL_ACCEPTS=($(printf '%s\n' "${ACCEPT_LIST[@]}" | sort -t'|' -k1 -rn 2>/dev/null))
IFS=$'\n' SORTED_REJECTS=($(printf '%s\n' "${REJECT_LIST[@]}" | sort -t'|' -k1 -rn 2>/dev/null))

# Cap to MAX_ADDS — overflow goes to reject history so they don't resurface for REJECT_COOLDOWN days
SORTED_ACCEPTS=("${_ALL_ACCEPTS[@]:0:$MAX_ADDS}")
for (( _i=MAX_ADDS; _i<${#_ALL_ACCEPTS[@]}; _i++ )); do
    SORTED_REJECTS+=("${_ALL_ACCEPTS[$_i]}")
done

_print_row() {
    local label="$1" entry="$2"
    IFS='|' read -r score name seeds affinity lfm <<< "$entry"
    printf "  %-8s [%3s] %-40s seeds: %s | affinity: %2s | %s\n" \
        "$label" "$score" "$name" "$seeds" "$affinity" "$lfm"
}

for entry in "${SORTED_ACCEPTS[@]}"; do [[ -n "$entry" ]] && _print_row "ACCEPT" "$entry"; done
for entry in "${SORTED_REJECTS[@]}"; do [[ -n "$entry" ]] && _print_row "REJECT" "$entry"; done

echo ""
echo "  Already in library: $ALREADY_KNOWN | Cooldown: $SKIP_COOLDOWN"
echo "  ACCEPT: ${#SORTED_ACCEPTS[@]} | REJECT: ${#SORTED_REJECTS[@]}"

if [[ "$DRY_RUN" == true ]]; then
    warn "DRY RUN complete — run without --dry-run to add accepted artists"
    exit 0
fi

if [[ "${#SORTED_ACCEPTS[@]}" -eq 0 ]]; then
    log "No artists above threshold — nothing to add"
    for entry in "${SORTED_REJECTS[@]}"; do
        [[ -z "$entry" ]] && continue
        IFS='|' read -r score name _ <<< "$entry"
        echo "REJECT|${name}|${TODAY}" >> "$HISTORY_FILE" 2>/dev/null
    done
    exit 0
fi

# ==============================================================================================
# ━━━ Add to Lidarr ━━━
# ==============================================================================================
echo ""
echo "━━━ $ICON_SYNC Adding to Lidarr ━━━"

LIDARR_ROOT=$(_lidarr_get "rootfolder" | jq -r 'first(.[] | select(.accessible == true)) | .path' 2>/dev/null)
if [[ -z "$LIDARR_ROOT" ]]; then error "Could not determine Lidarr root folder"; exit 1; fi

QUALITY_PROFILES=$(_lidarr_get "qualityprofile") || { error "Could not fetch quality profiles"; exit 1; }
METADATA_PROFILES=$(_lidarr_get "metadataprofile") || { error "Could not fetch metadata profiles"; exit 1; }

DEFAULT_QUALITY_ID=$(echo "$QUALITY_PROFILES" | jq -r '.[0].id' 2>/dev/null)
DEFAULT_METADATA_ID=$(echo "$METADATA_PROFILES" | jq -r '
    first(.[] | select(.name | test("Standard"; "i")) | .id) // .[0].id' 2>/dev/null)

log "Quality profile: $DEFAULT_QUALITY_ID | Metadata profile: $DEFAULT_METADATA_ID"

ADDED=0
FAILED=0

for entry in "${SORTED_ACCEPTS[@]}"; do
    [[ -z "$entry" ]] && continue
    IFS='|' read -r score name seeds affinity lfm <<< "$entry"

    LOOKUP=$(_lidarr_lookup "$name")

    if [[ -z "$LOOKUP" ]] || echo "$LOOKUP" | jq -e '. == [] or . == null' >/dev/null 2>&1; then
        warn "  $ICON_WARN No match in Lidarr lookup: $name"
        echo "FAIL|${name}|${TODAY}" >> "$HISTORY_FILE" 2>/dev/null
        (( FAILED++ ))
        continue
    fi

    ARTIST_DATA=$(echo "$LOOKUP" | jq '.[0]' 2>/dev/null)
    MBID=$(echo "$ARTIST_DATA" | jq -r '.foreignArtistId // ""' 2>/dev/null)
    LIDARR_NAME=$(echo "$ARTIST_DATA" | jq -r '.artistName // ""' 2>/dev/null)

    if [[ -z "$MBID" ]]; then
        warn "  $ICON_WARN No MusicBrainz ID for: $name"
        echo "FAIL|${name}|${TODAY}" >> "$HISTORY_FILE" 2>/dev/null
        (( FAILED++ ))
        continue
    fi

    PAYLOAD=$(echo "$ARTIST_DATA" | jq \
        --arg root "$LIDARR_ROOT" \
        --argjson qid "$DEFAULT_QUALITY_ID" \
        --argjson mid "$DEFAULT_METADATA_ID" \
        '. + {
            rootFolderPath: $root,
            qualityProfileId: $qid,
            metadataProfileId: $mid,
            monitored: true,
            addOptions: {
                monitor: "all",
                searchForMissingAlbums: true
            }
        }' 2>/dev/null)

    RESULT=$(_lidarr_post "$PAYLOAD")

    if echo "$RESULT" | jq -e '.id' >/dev/null 2>&1; then
        log "  $ICON_DONE Added: $LIDARR_NAME (score: $score | seeds: $seeds)"
        echo "ACCEPT|${name}|${TODAY}" >> "$HISTORY_FILE" 2>/dev/null
        (( ADDED++ ))
    else
        warn "  $ICON_WARN Failed to add: $name"
        log "  $(echo "$RESULT" | head -c 200)"
        echo "FAIL|${name}|${TODAY}" >> "$HISTORY_FILE" 2>/dev/null
        (( FAILED++ ))
    fi
done

for entry in "${SORTED_REJECTS[@]}"; do
    [[ -z "$entry" ]] && continue
    IFS='|' read -r score name _ <<< "$entry"
    echo "REJECT|${name}|${TODAY}" >> "$HISTORY_FILE" 2>/dev/null
done

# ==============================================================================================
# ━━━ Summary ━━━
# ==============================================================================================
echo ""
echo "━━━━━ $ICON_SUMMARY DISCOVERY COMPLETE ━━━━━"
echo "  $ICON_DONE  Added:   $ADDED"
[[ "$FAILED" -gt 0 ]] && echo "  $ICON_WARN  Failed:  $FAILED"
echo "  $ICON_SKIP  Rejected: ${#SORTED_REJECTS[@]}"
echo "━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━"

[[ "$ADDED" -gt 0 ]] && notify \
    "$ADDED artist(s) added to Lidarr via discovery on $(hostname)" \
    "Lidarr Discovery" "normal"

exit 0