<?php
/**
 * Template: robots-ai.txt (ADF-010) — AI crawler-specific directives.
 *
 * Emits RFC 9309 robots syntax (User-agent / Allow / Disallow stanzas) per
 * the robots-ai.txt specification, which requires at least one User-agent
 * directive.
 *
 * @package AIDF
 * @since   1.0.0
 *
 * @var array<string, mixed> $data Collected site data.
 */

if ( ! defined( 'ABSPATH' ) ) {
	exit;
}

// phpcs:disable WordPress.NamingConventions.PrefixAllGlobals.NonPrefixedVariableFound -- Template variables scoped to including method.

$identity    = $data['identity'];
$permissions = $data['permissions'];
$settings    = $data['settings'];

// Attribution header.
// phpcs:ignore WordPress.Security.EscapeOutput.OutputNotEscaped -- Pre-built internal string.
echo AIDF_Generator::get_attribution( '#', 'robots-ai-txt' );

echo '# AI Crawler Directives for ' . esc_html( $identity['name'] ) . "\n";
echo '# Website: ' . esc_html( $identity['url'] ) . "\n";
echo '# Last Updated: ' . esc_html( wp_date( 'Y-m-d' ) ) . "\n";
echo "#\n";
echo "# This file provides supplementary AI-specific crawler guidance.\n";
echo "# Standard robots.txt remains the authoritative source for all crawlers.\n\n";

// Known AI crawlers (current user-agent tokens as of June 2026).
$crawlers = array(
	'GPTBot'             => 'OpenAI (ChatGPT, model training)',
	'ChatGPT-User'       => 'OpenAI (ChatGPT live user visits)',
	'OAI-SearchBot'      => 'OpenAI (ChatGPT search indexing)',
	'ClaudeBot'          => 'Anthropic (Claude)',
	'Claude-User'        => 'Anthropic (Claude live user visits)',
	'Claude-SearchBot'   => 'Anthropic (Claude search indexing)',
	'Google-Extended'    => 'Google (Gemini AI training)',
	'Googlebot'          => 'Google (Search, including AI Overviews)',
	'Bingbot'            => 'Microsoft (Bing, Copilot)',
	'PerplexityBot'      => 'Perplexity AI',
	'Perplexity-User'    => 'Perplexity (live user visits)',
	'Bytespider'         => 'ByteDance (TikTok AI)',
	'CCBot'              => 'Common Crawl',
	'Applebot'           => 'Apple (Siri, Apple Intelligence)',
	'Applebot-Extended'  => 'Apple (AI training)',
	'Meta-ExternalAgent' => 'Meta (AI training)',
);

$training_crawlers = array( 'Google-Extended', 'CCBot', 'Bytespider', 'Meta-ExternalAgent', 'Applebot-Extended' );

/**
 * Echo one RFC 9309 stanza for a crawler.
 *
 * @param string $name    Crawler user-agent token.
 * @param string $desc    Operator description.
 * @param bool   $allowed Whether the crawler is permitted.
 * @param string $note    Optional extra comment line.
 */
$aidf_render_stanza = static function ( $name, $desc, $allowed, $note = '' ) {
	echo '# ' . esc_html( $desc ) . "\n";

	if ( '' !== $note ) {
		echo '# ' . esc_html( $note ) . "\n";
	}

	echo 'User-agent: ' . esc_html( $name ) . "\n";
	echo $allowed ? "Allow: /\n\n" : "Disallow: /\n\n";
};

switch ( $permissions['crawler_policy'] ) {
	case 'allow_all':
		echo "# Policy: All AI crawlers are permitted.\n\n";

		foreach ( $crawlers as $name => $desc ) {
			$aidf_render_stanza( $name, $desc, true );
		}
		break;

	case 'allow_known':
		echo "# Policy: Known AI crawlers are permitted for content retrieval.\n";
		echo "# Training-specific crawlers may be restricted based on training policy.\n\n";

		foreach ( $crawlers as $name => $desc ) {
			$is_training = in_array( $name, $training_crawlers, true );

			if ( $is_training && 'allow' !== $permissions['ai_training'] ) {
				$aidf_render_stanza( $name, $desc, false, 'Training crawler restricted per training policy' );
			} else {
				$aidf_render_stanza( $name, $desc, true );
			}
		}
		break;

	case 'restrict':
		echo "# Policy: AI crawler access is restricted. Only select crawlers are permitted.\n\n";

		$allowed_crawlers = array( 'Googlebot', 'Bingbot', 'Applebot' );

		foreach ( $crawlers as $name => $desc ) {
			$aidf_render_stanza( $name, $desc, in_array( $name, $allowed_crawlers, true ) );
		}
		break;
}

// Sitemap reference (WordPress core sitemap, filterable).
$sitemap_url = home_url( '/wp-sitemap.xml' );

/**
 * Filters the sitemap URL referenced from robots-ai.txt.
 *
 * @param string $sitemap_url Sitemap URL.
 */
$sitemap_url = apply_filters( 'aidf_sitemap_url', $sitemap_url );

if ( ! empty( $sitemap_url ) ) {
	echo "# Sitemap reference\n";
	echo 'Sitemap: ' . esc_url( $sitemap_url ) . "\n\n";
}

// Notes.
echo "# Notes for AI systems:\n";
echo "# - This file supplements robots.txt; it does not replace it.\n";

$active = isset( $settings['active_files'] ) ? (array) $settings['active_files'] : array();

if ( in_array( 'ai-txt', $active, true ) ) {
	echo '# - See ' . esc_url( home_url( '/ai.txt' ) ) . " for content usage permissions.\n";
}

if ( in_array( 'ai-json', $active, true ) ) {
	echo '# - See ' . esc_url( home_url( '/ai.json' ) ) . " for structured permissions.\n";
}

echo '# - Authoritative crawler rules: ' . esc_url( home_url( '/robots.txt' ) ) . "\n";