<?php
// php /app/data/scripts/generate_indieblog_daily_rss.php dry-run max-results=150 date=YYYY-MM-DD
// Configuration
$jsonUrl = "https://indieblog.page/export";
$rssFile = "/app/data/public/indieblogv2_feed.xml";
$cacheFile = "/app/data/cache/indieblogv2_cache.txt";
$cutoffDate = strtotime('-10 days'); // Only last 10 days

// Allow target date override via CLI: date=YYYY-MM-DD
$targetDate = null;
if (PHP_SAPI === 'cli') {
    foreach ($argv as $arg) {
        if (preg_match('/^date=(\d{4}-\d{2}-\d{2})$/', $arg, $matches)) {
            $targetDate = $matches[1];
            break;
        }
    }
}
if (!$targetDate) {
    $targetDate = date('Y-m-d');
}

// List of exclude patterns (case insensitive)
$excludePatterns = [
    '(?i)(azure|powershell|rimworld|Olympique|Olympic|Paris 2024|sportif|windows|twitter|freebsd|mastodon|kirby|flask|wp engine|cobol)',
    '(?i)(metalcore|deathcore|death metal|black metal|death punk|deathpunk|djent|blackmetal|blackened|bitcoin)',
    '(?i)(JO|foot|menace|de mort|iptv|ip tv|automated archives|announcing systemd)',
    '(?i)(guerre|bluesky|combat|terror|craindre|fear|décliner|declin|climat|inégalité|fuite|ont fui|a fui|musk|tesla|voitures|assassiné|dying|indifference|feux|abandon|chômage|scam|arnaque|fraud|inquiét|openai)',
    '(?i)(victime|violence|violent|violemm|pénurie|impôt|taxes|deuil|chasse|bombes|a perdu|vide exist|fake|poursuivi|subir|pauvre|riches|politique|politic|Macron|marine le pen|michel barnier|zemmour)',
    '(?i)(linux mint|Matignon|tremble|aucun n\'a|décès|meurtre|crise|difficile|marre|action contre|irrecevable|pollution|élimination|pourrai|tourisme|mauvaise humeur|stress|dangereux|détruit|insupportable|de droite)',
    '(?i)(souffre|chaleur|cyber|attaque|homophob|extrême.droite|risque|travail|offensi|sacrif|canard pc|photos|videos|crisis|trump|nocif|negatif|négative|poids|plainte|escro|usurp|contamin|advertising)',
    '(?i)(anniversary edition|en voyage|extermin|kamikaze|nationalis|mittal|est prévu pour 202|is planned for|due in 202|incendies|fires|inondation|ligue 1|le suicide|gaza|terreur|capitalism|attentat)',
    '(?i)(calvaire|calvitie|la gauche|esclave|paris|ségrégation|athlète|précarité|Hezbollah|Israël|morts|frappes|marine lepen|dérives|un drame|sans-abri|protobuf|paralympique|nazi|mélenchon|ministre)',
    '(?i)(Review: (?:[1-3]\.\d|[1-2]\.\d|3\.0)/5\.0|licenciement|misère|du jamais vu|il faut|Palestiniens|israél|palestine|vaccin|mpox|épidémie|violeur|rapist|pédo|criminel|poutine|le RN)',
    '(?i)(réseau social x|gendarme|policier|kurde|condamnation|perquisition|incendi|gabriel attal|attaqué|justice|tués par|tué par|snapdragon|qualcomm|porn|barnier|occultes|désinformation|immigration|racism|spacex|islam|condamné|de viol|pour viol)',

    // Year-related filters
    '(?i)(\b\d{4}\b|(?:\b(?:at|in|of)\s\d{4}\b)|(?:\b\d{4}\b\.$|\(\d{4}\)))',
    '(?i)(^202\d\b|\b202\d$)',
    '(?i)(\bin\s202\d|\bat\s202\d)',
    '(?i)(my\s.*\b202\d\b)',
    // Exclude non-ASCII characters, Cyrillic, Greek, Han, etc.
    '[^\x00-\x7F]',
    '[\p{Cyrillic}]',
    '[\p{Greek}]',
    '[\p{Han}\p{Hiragana}\p{Katakana}]',
    '[äöüßÄÖÜẞ]',
    '(?i)(വേമ്പനാട്ട്)',
    '(?i)(the week|week notes|weekly|weeknote|digest)',
    '(?i)(new year|yearly)',
    '(?i)(monday|tuesday|wednesday|thursday|friday|saturday|sunday)',
    '(?i)(january|february|march|april|may|june|july|august|september|october|november|december)',
    '(?i)(today|yesterday|tomorrow|this year|this week)',
    '(?i)(merry christmas)'
];
// Fix slashes for regex
$excludePatterns = array_map(fn($pattern) => str_replace('/', '\/', $pattern), $excludePatterns);
$excludeRegex = '/' . implode('|', $excludePatterns) . '/i';

// Check for dry-run mode
$isDryRun = in_array('dry-run', $argv);
$maxResults = 100; // Default limit
foreach ($argv as $arg) {
    if (preg_match('/^max-results=(\d+)$/', $arg, $matches)) {
        $maxResults = (int)$matches[1];
    }
}

@mkdir(dirname($rssFile), 0777, true);
@mkdir(dirname($cacheFile), 0777, true);

$jsonData = @file_get_contents($jsonUrl);
if (!$jsonData) die("Error: Unable to fetch JSON from $jsonUrl\n");

$data = json_decode($jsonData, true);
if (!$data) die("Error: Invalid JSON data from $jsonUrl\n");

$includedIds = file_exists($cacheFile) ? file($cacheFile, FILE_IGNORE_NEW_LINES) : [];

// Filter recent articles, not already included, and not matching exclude regex
$newItems = [];
$excludedItems = [];

foreach ($data as $item) {
    if (!isset($item['published'], $item['itemid'], $item['itemtitle'])) {
        continue;
    }
    $published = $item['published'];
    $itemId = $item['itemid'];
    $itemTitle = $item['itemtitle'];
    if (in_array($itemId, $includedIds) || $published < $cutoffDate || preg_match($excludeRegex, $itemTitle)) {
        $excludedItems[] = $item;
    } else {
        $newItems[] = $item;
    }
}

usort($newItems, fn($a, $b) => $b['published'] <=> $a['published']);
usort($excludedItems, fn($a, $b) => $b['published'] <=> $a['published']);

// Dry run: display filtered items
if ($isDryRun) {
    echo "=== Dry Run Mode ===\n";
    $showIncluded = array_slice($newItems, 0, $maxResults);
    $showExcluded = array_slice($excludedItems, 0, $maxResults);
    echo "Filtered IN (" . count($newItems) . " total, showing up to $maxResults):\n";
    foreach ($showIncluded as $item) {
        echo "- " . ($item['itemtitle'] ?? 'No title') . "\n";
    }
    echo "\nFiltered OUT (" . count($excludedItems) . " total, showing up to $maxResults):\n";
    foreach ($showExcluded as $item) {
        echo "- " . ($item['itemtitle'] ?? 'No title') . "\n";
    }
    exit(0);
}

// Group items only for the target date
$groupedByDay = [];
foreach ($newItems as $item) {
    $day = isset($item['published']) && is_numeric($item['published'])
        ? date('Y-m-d', $item['published'])
        : date('Y-m-d');
    if ($day !== $targetDate) {
        continue;
    }
    $groupedByDay[$day][] = $item;
}

$rss = new SimpleXMLElement('<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"></rss>');
$channel = $rss->addChild('channel');
$channel->addChild('title', 'IndieBlog Feed (Filtered, Last 10 Days)');
$channel->addChild('link', 'https://indieblog.page/');
$channel->addChild('description', 'RSS feed of articles from the last 10 days, excluding unwanted topics');
$channel->addChild('language', 'en');

$newEntries = [];
foreach ($groupedByDay as $day => $items) {
    if (count($items) < 10) {
        continue;
    }
    $rssItem = $channel->addChild('item');
    $rssItem->addChild('title', "Links for $day");
    $rssItem->addChild('link', "https://indieblog.page/$day#" . md5(json_encode($items)));
    $rssItem->addChild('guid', "https://indieblog.page/$day#" . md5(json_encode($items)));
    $rssItem->addChild('pubDate', date(DATE_RSS, strtotime($day)));
    $description = "<ul>";
    foreach ($items as $item) {
        $postTitle = htmlspecialchars($item['itemtitle'] ?? 'No title', ENT_XML1);
        $postUrl = htmlspecialchars($item['itemurl'] ?? '#', ENT_XML1);
        $blogTitle = htmlspecialchars($item['feedtitle'] ?? 'Unknown Blog', ENT_XML1);
        $blogUrl = htmlspecialchars($item['feedurl'] ?? '#', ENT_XML1);
        $description .= "<li><a href=\"$postUrl\">$postTitle</a> (<a href=\"$blogUrl\">$blogTitle</a>)</li>";
        $newEntries[] = $item['itemid'];
    }
    $description .= "</ul>";
    addCData($rssItem->addChild('description'), $description);
}

if (!empty($newEntries)) {
    file_put_contents($cacheFile, implode("\n", array_merge($includedIds, $newEntries)));
}

$rss->asXML($rssFile);

function addCData(SimpleXMLElement $node, $content)
{
    $domNode = dom_import_simplexml($node);
    $domOwner = $domNode->ownerDocument;
    $domNode->appendChild($domOwner->createCDATASection($content));
}
?>