<?php
/**
 * Lightweight Server-Side DOM Scraper for Product URLs.
 *
 * @package AffiStyle
 */

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

/**
 * Class AffiStyle_Scraper
 *
 * Fetches and parses product details from retail URLs using WordPress HTTP API and native DOMDocument.
 */
class AffiStyle_Scraper {


	/**
	 * Extract and normalize product details from an Amazon product URL.
	 *
	 * @param string $url Target Amazon URL.
	 * @return array|WP_Error Array with product details or WP_Error on failure.
	 */
	public static function fetch_amazon_product( $url ) {
		$clean_url = esc_url_raw( trim( $url ) );

		if ( empty( $clean_url ) || false === wp_http_validate_url( $clean_url ) ) {
			return new WP_Error(
				'affistyle_invalid_url',
				__( 'Please provide a valid and complete web URL.', 'affistyle' )
			);
		}

		// Verify domain matches Amazon or amzn.to shortlink.
		if ( ! preg_match( '/https?:\/\/(?:[a-zA-Z0-9\-]+\.)?(?:amazon\.[a-z\.]{2,6}|amzn\.to)\/.*/i', $clean_url ) ) {
			return new WP_Error(
				'affistyle_unsupported_domain',
				__( 'The provided URL is not a supported Amazon product link.', 'affistyle' )
			);
		}

		// Handle shortlinks (amzn.to) by following redirects.
		if ( preg_match( '/https?:\/\/amzn\.to\/[a-zA-Z0-9]+/i', $clean_url ) ) {
			$head_response = wp_safe_remote_head(
				$clean_url,
				array(
					'redirection' => 5,
					'timeout'     => 10,
					'sslverify'   => true,
					'user-agent'  => self::get_user_agent(),
				)
			);

			if ( ! is_wp_error( $head_response ) ) {
				$effective_url = wp_remote_retrieve_header( $head_response, 'location' );
				if ( ! empty( $effective_url ) && is_string( $effective_url ) ) {
					$candidate_url = esc_url_raw( $effective_url );
					if ( false !== wp_http_validate_url( $candidate_url ) && preg_match( '/https?:\/\/(?:[a-zA-Z0-9\-]+\.)?amazon\.[a-z\.]{2,6}\/.*/i', $candidate_url ) ) {
						$clean_url = $candidate_url;
					}
				}
			}
		}

		// Extract ASIN (10 alphanumeric characters).
		$asin = self::extract_amazon_asin( $clean_url );

		// Check transient cache if ASIN is available.
		if ( ! empty( $asin ) ) {
			$transient_key = 'affistyle_az_' . md5( $asin );
			$cached_data   = get_transient( $transient_key );
			if ( false !== $cached_data && is_array( $cached_data ) ) {
				return $cached_data;
			}
		}

		// Fetch page contents via WordPress HTTP API.
		$args = array(
			'timeout'     => 12,
			'redirection' => 5,
			'sslverify'   => true,
			'user-agent'  => self::get_user_agent(),
			'headers'     => array(
				'Accept'          => 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8',
				'Accept-Language' => 'en-US,en;q=0.9',
				'Cache-Control'   => 'no-cache',
				'Pragma'          => 'no-cache',
			),
		);

		$response = wp_safe_remote_get( $clean_url, $args );

		if ( is_wp_error( $response ) ) {
			return new WP_Error(
				'affistyle_http_error',
				sprintf(
					/* translators: %s: error message */
					__( 'Failed to connect to Amazon: %s', 'affistyle' ),
					$response->get_error_message()
				)
			);
		}

		$response_code = wp_remote_retrieve_response_code( $response );
		if ( 200 !== (int) $response_code ) {
			return new WP_Error(
				'affistyle_http_status_' . $response_code,
				sprintf(
					/* translators: %d: HTTP status code */
					__( 'Amazon returned HTTP status code %d. The product may be unavailable or blocked.', 'affistyle' ),
					$response_code
				)
			);
		}

		$html = wp_remote_retrieve_body( $response );
		if ( empty( $html ) ) {
			return new WP_Error(
				'affistyle_empty_body',
				__( 'Amazon returned an empty response.', 'affistyle' )
			);
		}

		// Detect CAPTCHA or anti-bot challenge pages.
		if ( false !== stripos( $html, 'validateCaptcha' ) || false !== stripos( $html, 'Robot Check' ) ) {
			return new WP_Error(
				'affistyle_captcha_blocked',
				__( 'Amazon presented a verification challenge (CAPTCHA). Please enter product details manually.', 'affistyle' )
			);
		}

		// Parse the HTML using native DOMDocument and DOMXPath.
		$data = self::parse_amazon_html( $html, $clean_url, $asin );

		if ( is_wp_error( $data ) ) {
			return $data;
		}

		// Cache successful result in transient for 12 hours.
		if ( ! empty( $asin ) ) {
			set_transient( 'affistyle_az_' . md5( $asin ), $data, 12 * HOUR_IN_SECONDS );
		}

		return $data;
	}

	/**
	 * Extract Amazon ASIN from URL.
	 *
	 * @param string $url Amazon URL.
	 * @return string Extracted ASIN or empty string.
	 */
	public static function extract_amazon_asin( $url ) {
		if ( preg_match( '/\/(?:dp|gp\/product|exec\/obidos\/asin|o\/asin|d)\/([a-z0-9]{10})/i', $url, $matches ) ) {
			return strtoupper( $matches[1] );
		}
		return '';
	}

	/**
	 * Parse Amazon HTML content using DOMDocument and DOMXPath.
	 *
	 * @param string $html      Response HTML.
	 * @param string $clean_url Normalized product URL.
	 * @param string $asin      Extracted ASIN.
	 * @return array|WP_Error Parsed product data array or WP_Error.
	 */
	private static function parse_amazon_html( $html, $clean_url, $asin ) {
		// Use libxml error suppression for non-strict HTML5 parsing.
		$previous_libxml_state = libxml_use_internal_errors( true );

		$dom = new DOMDocument();
		// Ensure UTF-8 decoding.
		$dom->loadHTML( '<?xml encoding="UTF-8">' . $html );
		libxml_clear_errors();
		libxml_use_internal_errors( $previous_libxml_state );

		$xpath = new DOMXPath( $dom );

		// 1. Extract Title.
		$title = self::extract_title( $xpath );

		// 2. Extract Image URL.
		$image_url = self::extract_image( $xpath );

		// 3. Extract Price.
		$price = self::extract_price( $xpath, $html );

		// If no title could be found, fail gracefully.
		if ( empty( $title ) ) {
			return new WP_Error(
				'affistyle_scrape_failed',
				__( 'Could not extract product details. Please enter them manually or check the URL.', 'affistyle' )
			);
		}

		return array(
			'title'     => $title,
			'image_url' => $image_url,
			'price'     => $price,
			'asin'      => $asin,
			'url'       => $clean_url,
		);
	}

	/**
	 * Extract product title from DOMXPath.
	 *
	 * @param DOMXPath $xpath XPath instance.
	 * @return string Extracted title.
	 */
	private static function extract_title( $xpath ) {
		$title_queries = array(
			'//*[@id="productTitle"]',
			'//h1[@id="title"]',
			'//span[@id="title"]',
			'//meta[@property="og:title"]/@content',
			'//meta[@name="title"]/@content',
			'//title',
		);

		foreach ( $title_queries as $query ) {
			$nodes = $xpath->query( $query );
			if ( $nodes && $nodes->length > 0 ) {
				$raw_title = trim( $nodes->item( 0 )->nodeValue );
				if ( ! empty( $raw_title ) ) {
					// Clean Amazon marketing suffix (e.g. " : Amazon.com: ...").
					$cleaned = preg_replace( '/\s*:\s*Amazon\.[a-z\.]+:.*$/i', '', $raw_title );
					$cleaned = preg_replace( '/\s*\|\s*Amazon\.[a-z\.]+/i', '', $cleaned );
					$cleaned = preg_replace( '/\s+/', ' ', $cleaned );
					return sanitize_text_field( trim( $cleaned ) );
				}
			}
		}

		return '';
	}

	/**
	 * Extract product image URL from DOMXPath and raw HTML attributes.
	 *
	 * @param DOMXPath $xpath XPath instance.
	 * @return string Extracted high-res image URL.
	 */
	private static function extract_image( $xpath ) {
		$img_queries = array(
			'//*[@id="landingImage"]',
			'//*[@id="imgBlkFront"]',
			'//img[@id="main-image"]',
			'//img[contains(@class, "a-dynamic-image")]',
		);

		foreach ( $img_queries as $query ) {
			$nodes = $xpath->query( $query );
			if ( $nodes && $nodes->length > 0 ) {
				$node = $nodes->item( 0 );

				// Check data-old-hires attribute first for full resolution.
				$old_hires = $node->getAttribute( 'data-old-hires' );
				if ( ! empty( $old_hires ) && false !== wp_http_validate_url( $old_hires ) ) {
					return esc_url_raw( $old_hires );
				}

				// Check data-a-dynamic-image attribute (JSON map of url => [width, height]).
				$dynamic_image_json = $node->getAttribute( 'data-a-dynamic-image' );
				if ( ! empty( $dynamic_image_json ) ) {
					$img_map = json_decode( $dynamic_image_json, true );
					if ( is_array( $img_map ) && ! empty( $img_map ) ) {
						// Pick highest resolution image based on width.
						$best_url  = '';
						$max_width = 0;
						foreach ( $img_map as $u => $dims ) {
							$w = ( is_array( $dims ) && isset( $dims[0] ) ) ? (int) $dims[0] : 0;
							if ( $w >= $max_width ) {
								$max_width = $w;
								$best_url  = $u;
							}
						}
						if ( ! empty( $best_url ) && false !== wp_http_validate_url( $best_url ) ) {
							return esc_url_raw( $best_url );
						}
					}
				}

				// Check standard src attribute.
				$src = $node->getAttribute( 'src' );
				if ( ! empty( $src ) && false !== wp_http_validate_url( $src ) && false === stripos( $src, 'data:image' ) ) {
					return esc_url_raw( $src );
				}
			}
		}

		// Fallback to Open Graph image tag.
		$og_nodes = $xpath->query( '//meta[@property="og:image"]/@content' );
		if ( $og_nodes && $og_nodes->length > 0 ) {
			$og_img = trim( $og_nodes->item( 0 )->nodeValue );
			if ( ! empty( $og_img ) && false !== wp_http_validate_url( $og_img ) ) {
				return esc_url_raw( $og_img );
			}
		}

		return '';
	}

	/**
	 * Extract product price from DOMXPath.
	 *
	 * @param DOMXPath $xpath XPath instance.
	 * @param string   $html  Raw HTML fallback.
	 * @return string Extracted price string (e.g. "$199.99").
	 */
	private static function extract_price( $xpath, $html ) {
		$price_queries = array(
			'//span[contains(@class, "apexPriceToPay")]//span[contains(@class, "a-offscreen")]',
			'//span[contains(@class, "priceToPay")]//span[contains(@class, "a-offscreen")]',
			'//*[@id="corePrice_desktop"]//span[contains(@class, "a-offscreen")]',
			'//*[@id="corePriceDisplay_desktop_feature_div"]//span[contains(@class, "a-offscreen")]',
			'//*[@id="priceblock_ourprice"]',
			'//*[@id="priceblock_dealprice"]',
			'//span[contains(@class, "a-price")]//span[contains(@class, "a-offscreen")]',
		);

		foreach ( $price_queries as $query ) {
			$nodes = $xpath->query( $query );
			if ( $nodes && $nodes->length > 0 ) {
				for ( $i = 0; $i < $nodes->length; $i++ ) {
					$val = trim( $nodes->item( $i )->nodeValue );
					if ( ! empty( $val ) && preg_match( '/[\$\£\€\¥\₹]?\s*\d+(?:[\.,]\d{2})?/', $val ) ) {
						return sanitize_text_field( $val );
					}
				}
			}
		}

		// Fallback regex matching in HTML if DOM extraction failed.
		if ( preg_match( '/"priceAmount"\s*:\s*"?([0-9\.]+)"?/i', $html, $matches ) ) {
			return '$' . sanitize_text_field( $matches[1] );
		}

		return '';
	}

	/**
	 * Modern realistic desktop User-Agent string.
	 *
	 * @return string
	 */
	private static function get_user_agent() {
		return 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/122.0.0.0 Safari/537.36';
	}
}