WordPress Headless Replace Images From External Source

wp_replace_images_from_external_source.php

<?php
/**
 ** Blogger 文章匯入使用內建匯入工具，不過圖片抓取機制鳥到爆，預設只會抓取文章封面圖，其他文章內的圖片都需要抓取的話就要跑這隻！
 **/
include 'wp-load.php';
set_time_limit(0);
ini_set('memory_limit', '256M');

add_action('after_setup_theme', function () {
	add_filter('intermediate_image_sizes', '__return_empty_array');
	add_filter('wp_get_attachment_image_src', function ($image, $attachment_id, $size, $icon) {
		// get a thumbnail or intermediate image if there is one
		$image = image_downsize($attachment_id, 'full');
		if (!$image) {
			$src = false;

			if ($icon && $src = wp_mime_type_icon($attachment_id)) {
				/** This filter is documented in wp-includes/post.php */
				$icon_dir = apply_filters('icon_dir', ABSPATH . WPINC . '/images/media');

				$src_file = $icon_dir . '/' . wp_basename($src);
				@list($width, $height) = getimagesize($src_file);
			}

			if ($src && $width && $height) {
				$image = array($src, $width, $height);
			}
		}
		return $image;
	}, 999, 4);
});

function logger($file, $data) {
	error_log($file . " => " . $data . PHP_EOL, 3, dirname(__FILE__) . "/replace.log");
}

function download_source($url, $dest) {
	$file_name = $dest;
	$buffer_size = 4096; // read 4kb at a time
	if (file_exists($file_name)) {
		return $file_name;
	}
	$fp = fopen($file_name, 'w');
	$ch = curl_init($url);
	curl_setopt($ch, CURLOPT_USERAGENT, 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_12_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/69.0.3497.100 Safari/537.36');
	curl_setopt($ch, CURLOPT_REFERER, $url);
	curl_setopt($ch, CURLOPT_FAILONERROR, true);
	curl_setopt($ch, CURLOPT_FOLLOWLOCATION, true);
	curl_setopt($ch, CURLOPT_FILE, $fp);
	$data = curl_exec($ch);
	if (curl_error($ch)) {
		$error_message = curl_error($ch);
		logger('request_download_source_error', print_r("LINK: {$url} | ERRMSG: $error_message", true));
	}
	curl_close($ch);
	fclose($fp);
	return $file_name;
}

function parsing_image($post_id, $content) {
	$content = preg_replace_callback('/<img\s+.*?src=[\"\'](http[s]{0,1}\:\/\/?[^\"\' >]*)[\"\']?[^>]*>/i',
		function ($res) {
			$pos = strpos($res[1], site_url());
			if ($pos === false) {
				$upload_dir = wp_upload_dir();
				$tmp_name = join('-', explode('/', $res[1]));
				$tmp_name = join('-', explode('.', $tmp_name));
				$tmp_name = join('-', explode('?', $tmp_name));
				$tmp_name = join('-', explode('&', $tmp_name));
				$tmp_name = join('-', explode('%', $tmp_name));
				$tmp_name = join('-', explode(';', $tmp_name));
				$tmp_name = join('-', explode('=', $tmp_name));
				$tmp_name = md5(explode(':', $tmp_name)[1]);
				$file_path = "{$upload_dir['basedir']}/{$tmp_name}";
				download_source($res[1], $file_path);
				$info = getimagesize($file_path);
				$type = explode('/', $info['mime']);
				$type = end($type);
				switch ($type) {
				case 'gif':
					rename($file_path, $file_path . ".gif");
					$tmp_name .= ".gif";
					break;
				case 'jpeg':
					rename($file_path, $file_path . ".jpg");
					$tmp_name .= ".jpg";
					break;
				case 'png':
					rename($file_path, $file_path . ".png");
					$tmp_name .= ".png";
					break;
				default:
					rename($file_path, $file_path . ".jpg");
					$tmp_name .= ".jpg";
					break;
				}
				$doc = new DOMDocument();
				$doc->loadHTML($res[0]);
				$tags = $doc->getElementsByTagName('img');
				if (count($tags) > 0) {
					$tag = $tags->item(0);
					$tag->setAttribute('src', "{$upload_dir['baseurl']}/{$tmp_name}");
					return $doc->saveXML($tag);
				}
			} else {
				return $res[0];
			}
		},
		$content);

	preg_match_all('/<img\s+.*?src=[\"\'](http[s]{0,1}\:\/\/?[^\"\' >]*)[\"\']?[^>]*>/i', $content, $matches);
	$images = $matches[1];
	//logger('parsing_image', print_r($matches, true));
	mxp_import_image($post_id, $images);
	$content = preg_replace('/<p.*?>/s', "<p>", $content);
	$content = preg_replace('/<div.*?>/s', "<div>", $content);
	$remove_divs = array("<div>", "</div>");
	$content = str_replace($remove_divs, "", $content);
	return $content;
}

function mxp_import_image($post_id, $imgs) {
	require_once ABSPATH . 'wp-admin/includes/media.php';
	require_once ABSPATH . 'wp-admin/includes/file.php';
	require_once ABSPATH . 'wp-admin/includes/image.php';
	global $wpdb;
	$filename = array();
	$upload_file = array();
	$upload_dir = wp_upload_dir();
	for ($i = 0; $i < count($imgs); ++$i) {
		$img_filename = str_replace($upload_dir['baseurl'] . "/", "", $imgs[$i]);
		$img_dir_path = str_replace($upload_dir['baseurl'], $upload_dir['basedir'], $imgs[$i]);
		$pos = strpos($imgs[$i], site_url());
		if (!file_exists($img_dir_path) && $pos === false) {
			$filename[] = $img_filename;
			$upload_file[] = $img_dir_path;
		} else {
			echo "已有 {$img_filename} 此媒體。位在： {$img_dir_path}" . PHP_EOL;
			$attachment_id = $wpdb->get_results(
				$wpdb->prepare(
					"SELECT ID FROM $wpdb->posts WHERE guid LIKE %s", '%' . pathinfo(str_replace($upload_dir['baseurl'] . "/", "", $imgs[$i]), PATHINFO_FILENAME) . '%'
				),
				ARRAY_A
			);
			$media_post = wp_update_post(array(
				'ID' => $attachment_id[0]['ID'],
				'post_parent' => $post_id,
			), true);
			if (is_wp_error($media_post)) {
				$filename[] = $img_filename;
				$upload_file[] = $img_dir_path;
			}
		}
	}
	//logger('mxp_import_image', print_r($filename, true));
	// 如果上傳沒失敗，就附加到剛剛那篇文章
	$set_feature_image = true;
	for ($i = 0; $i < count($upload_file); ++$i) {
		if (isset($upload_file[$i]) && $upload_file[$i] != "" && isset($filename[$i]) && $filename[$i] != "") {
			$wp_filetype = wp_check_filetype($filename[$i], null);
			$attachment = array(
				'post_mime_type' => $wp_filetype['type'],
				'post_parent' => $post_id,
				'post_title' => preg_replace('/\.[^.]+$/', '', $filename[$i]),
				'post_content' => '',
				'post_status' => 'inherit',
			);
			$attachment_id = wp_insert_attachment($attachment, $upload_file[$i], $post_id);
			if (!is_wp_error($attachment_id)) {
				//產生附加檔案中繼資料
				$attachment_data = wp_generate_attachment_metadata($attachment_id, $upload_file[$i]);
				wp_update_attachment_metadata($attachment_id, $attachment_data);
				//將圖像的附加檔案設為特色圖片
				$type = explode("/", $wp_filetype['type']);
				if ($set_feature_image == true && $type[0] == 'image') {
					set_post_thumbnail($post_id, $attachment_id);
					$set_feature_image = false;
				}
				echo "匯入 {$filename[$i]} 媒體。位在： {$upload_file[$i]}" . PHP_EOL;
			}
		}
	}
}

global $wpdb;
$querystr = "SELECT ID,post_content FROM $wpdb->posts WHERE $wpdb->posts.post_status = 'publish' AND $wpdb->posts.post_type = 'post' AND $wpdb->posts.post_date < NOW() ORDER BY $wpdb->posts.post_date DESC";

$posts = $wpdb->get_results($querystr, ARRAY_A);

foreach ($posts as $key => $post) {
	$post_id = $post['ID'];
	$post_content = $post['post_content'];
	//匯入圖片
	$update_attachment_post = array(
		'ID' => $post_id,
		'post_content' => parsing_image($post_id, $post_content),
	);
	$update_attachment_post['post_excerpt'] = wp_trim_words($update_attachment_post['post_content'], 200, '...');
	$upid = wp_update_post($update_attachment_post);
	echo $post_id . " => 完成匯入！" . PHP_EOL;
}