<?php
// api/debug/report.php — structured bug-report intake (Part A)
// POST { title, body?, diagnostics{...} } -> { success, id, duplicate, url }
require_once __DIR__ . '/config.php';

if ($_SERVER['REQUEST_METHOD'] !== 'POST') {
    error_response('Method not allowed', 405);
}

define('REPORTS_DIR', __DIR__ . '/../../data/reports');
define('RATE_FILE', DEBUG_DATA_DIR . '/report_rate.json');
define('MAX_REPORT_KB', 512);
define('RATE_MAX', 5);        // reports per IP …
define('RATE_WINDOW', 600);   // … per 10 minutes

$raw = file_get_contents('php://input');
if (!$raw) error_response('Empty request body');
if (strlen($raw) > MAX_REPORT_KB * 1024) {
    error_response('Payload too large (max ' . MAX_REPORT_KB . 'KB)', 413);
}

$payload = json_decode($raw, true);
if (!is_array($payload)) error_response('Invalid JSON');

$title = trim((string)($payload['title'] ?? ''));
if ($title === '') error_response('title is required');
if (mb_strlen($title) > 160) $title = mb_substr($title, 0, 160);
$body = (string)($payload['body'] ?? '');
if (mb_strlen($body) > 4000) $body = mb_substr($body, 0, 4000);
$diag = (isset($payload['diagnostics']) && is_array($payload['diagnostics'])) ? $payload['diagnostics'] : [];

// Bound every stack: 20 errors x 10 frames of long URLs can otherwise approach the 512 KB ceiling
// and reject a legitimate report.
if (!empty($diag['errors']) && is_array($diag['errors'])) {
    foreach ($diag['errors'] as $k => $de) {
        if (!is_array($de)) { unset($diag['errors'][$k]); continue; }
        if (isset($de['stack'])) {
            $diag['errors'][$k]['stack'] = mb_substr((string)$de['stack'], 0, 2000);
        }
    }
    $diag['errors'] = array_values($diag['errors']);
}

// Maintenance thread: lets the in-app agent and the Hermes maintenance job hold a conversation.
// `thread` is a client-generated id (thr-xxxxxxxx); `kind` distinguishes agent feedback from
// operator bug reports so triage can weight them differently.
$thread = preg_replace('/[^a-zA-Z0-9_-]/', '', (string)($payload['thread'] ?? ''));
if (mb_strlen($thread) > 40) $thread = substr($thread, 0, 40);
$kind = (($payload['kind'] ?? '') === 'maintenance') ? 'maintenance' : 'user';

// Storage guard: data/ is not writable by php-fpm, so never pretend we wrote a report.
if (!is_dir(REPORTS_DIR)) @mkdir(REPORTS_DIR, 0775, true);
if (!is_dir(REPORTS_DIR) || !is_writable(REPORTS_DIR)) {
    error_response('Report storage unavailable', 503);
}

// Per-IP rate limit
$ip = $_SERVER['REMOTE_ADDR'] ?? 'unknown';
$now = time();
$rate = load_json(RATE_FILE);
$hits = array_values(array_filter($rate[$ip] ?? [], function ($t) use ($now) {
    return $t > $now - RATE_WINDOW;
}));
if (count($hits) >= RATE_MAX) {
    header('Retry-After: ' . RATE_WINDOW);
    error_response('Rate limit exceeded (max ' . RATE_MAX . ' reports per 10 minutes)', 429);
}
$hits[] = $now;
$rate[$ip] = $hits;
foreach ($rate as $k => $v) {
    $v = array_values(array_filter($v, function ($t) use ($now) { return $t > $now - 86400; }));
    if ($v) $rate[$k] = $v; else unset($rate[$k]);
}
save_json(RATE_FILE, $rate);

// ── Network-class filter (2026-10-26) ────────────────────────────────────────────────────────
// The debug supervisor files an auto report for *any* failed fetch, so a flaky phone connection
// or a retired endpoint becomes a review card. One cluster alone held 14 400 occurrences and
// produced 3 cards. A `Fetch failed:` line is a retry, not a defect: count it and stop.
// Only auto reports whose every error is network-class are filtered — a stack trace, TypeError
// or user-filed report is never dropped here.
define('NET_FILE', DEBUG_DATA_DIR . '/network_failures.json');
define('NET_KEEP_SOURCES', 40);   // bounded: keep the noisiest paths only

$isNetworkMsg = function ($m) {
    return (bool)preg_match(
        '/fetch failed|failed to fetch|networkerror|network error|load failed'
        . '|the user aborted a request|signal is aborted|err_internet_disconnected|err_network_changed/i',
        (string)$m
    );
};

if (!empty($diag['auto']) && !empty($diag['errors']) && is_array($diag['errors'])) {
    $msgs = [];
    foreach ($diag['errors'] as $de) {
        if (is_array($de)) $msgs[] = (string)($de['message'] ?? '');
    }
    $netCount = count(array_filter($msgs, $isNetworkMsg));
    if ($msgs && $netCount === count($msgs)) {
        $day = gmdate('Y-m-d');
        $net = load_json(NET_FILE);
        if (!is_array($net) || ($net['day'] ?? '') !== $day) {
            $net = ['day' => $day, 'total' => 0, 'filtered_reports' => 0, 'by_source' => []];
        }
        $net['total'] = (int)($net['total'] ?? 0) + 1;
        $net['filtered_reports'] = (int)($net['filtered_reports'] ?? 0) + 1;
        $src = 'unknown';
        foreach (array_merge($msgs, [$title]) as $candidate) {
            if (preg_match('~(/[A-Za-z0-9_\-./]+\.(?:php|json))~', (string)$candidate, $m)) {
                $src = $m[1];
                break;
            }
            if (preg_match('~(https?://[^\s]+)~', (string)$candidate, $m)) {
                $src = preg_replace('~^https?://[^/]+~', '', $m[1]);
                break;
            }
        }
        $net['by_source'][$src] = (int)($net['by_source'][$src] ?? 0) + 1;
        if (count($net['by_source']) > NET_KEEP_SOURCES) {
            arsort($net['by_source']);
            $net['by_source'] = array_slice($net['by_source'], 0, NET_KEEP_SOURCES, true);
        }
        save_json(NET_FILE, $net);
        json_response([
            'success' => true,
            'filtered' => 'network',
            'source' => $src,
            'day' => $day,
            'count_today' => $net['total'],
            'url' => null,
        ], 202);
    }
}

// Dedupe: same title + error set + version within 24h collapses onto the existing report.
// Hash only the fields the pre-2026-10-23 payload had. Adding `stack` to json_encode($diag['errors'])
// would change every fingerprint, so an identical error filed today would not dedupe against the
// same error filed yesterday.
$errKey = [];
foreach (is_array($diag['errors'] ?? null) ? $diag['errors'] : [] as $de) {
    if (!is_array($de)) continue;
    $errKey[] = ['message' => (string)($de['message'] ?? ''),
                 'source'  => (string)($de['source'] ?? ''),
                 'count'   => (int)($de['count'] ?? 1)];
}
$fingerprint = sha1($title . '|' . json_encode($errKey) . '|' . (string)($diag['version'] ?? ''));
$existing = glob(REPORTS_DIR . '/RPT-*.json') ?: [];
foreach (array_slice(array_reverse($existing), 0, 100) as $f) {
    $d = json_decode(@file_get_contents($f), true);
    if (!is_array($d)) continue;
    if (($d['fingerprint'] ?? '') === $fingerprint && strtotime($d['created_at'] ?? '') > $now - 86400) {
        // Don't discard user content on a duplicate: keep it as an appended note.
        $d['duplicate_count'] = (int)($d['duplicate_count'] ?? 0) + 1;
        if ($body !== '' && ($d['body'] ?? '') !== $body) {
            $d['body'] = trim((string)($d['body'] ?? '') . "\n---\n[duplicate " . gmdate('c') . "] " . $body);
            $json = json_encode($d, JSON_PRETTY_PRINT | JSON_UNESCAPED_UNICODE | JSON_UNESCAPED_SLASHES);
            if ($json !== false) { @file_put_contents($f, $json); @chmod($f, 0664); }
        }
        json_response([
            'success' => true,
            'id' => $d['id'],
            'duplicate' => true,
            'url' => '/spaces/newos4/downloads/bugs/BUG-REPORT.html',
        ]);
    }
}

$id = 'RPT-' . gmdate('Ymd-His') . '-' . substr(bin2hex(random_bytes(4)), 0, 8);
$record = [
    'id' => $id,
    'created_at' => gmdate('c'),
    'fingerprint' => $fingerprint,
    'title' => $title,
    'body' => $body,
    'diagnostics' => $diag,
    'thread' => $thread,
    'kind' => $kind,
    'meta' => [
        'ip' => $ip,
        'ua' => $_SERVER['HTTP_USER_AGENT'] ?? '',
    ],
];

$file = REPORTS_DIR . '/' . $id . '.json';
$json = json_encode($record, JSON_PRETTY_PRINT | JSON_UNESCAPED_UNICODE | JSON_UNESCAPED_SLASHES);
if ($json === false || file_put_contents($file, $json) === false) {
    error_response('Failed to persist report', 500);
}
@chmod($file, 0664);

// Regenerate the public page immediately so the URL we print is never stale.
$generator = realpath(__DIR__ . '/../../scripts/gen-bug-report.py');
if ($generator) {
    @exec('/usr/bin/python3 ' . escapeshellarg($generator) . ' 2>&1', $genOut, $genRc);
}

json_response([
    'success' => true,
    'id' => $id,
    'duplicate' => false,
    'url' => '/spaces/newos4/downloads/bugs/BUG-REPORT.html',
], 201);