| 1 |
<?php |
| 2 |
|
| 3 |
namespace SubstackImporter; |
| 4 |
|
| 5 |
use WP_Error; |
| 6 |
use WXR_Generator\Generator; |
| 7 |
use ZipArchive; |
| 8 |
use DOMDocument; |
| 9 |
use DomComment; |
| 10 |
use DomElement; |
| 11 |
use DOMText; |
| 12 |
|
| 13 |
|
| 14 |
/** |
| 15 |
* The Substack Converter is responsible for taking in a Substack export and providing |
| 16 |
* data to the WXR generator. |
| 17 |
* |
| 18 |
* @package SubstackImporter |
| 19 |
*/ |
| 20 |
class Converter { |
| 21 |
|
| 22 |
/** |
| 23 |
* @var string $export_file_path Path of the export file. |
| 24 |
*/ |
| 25 |
protected $export_file_path; |
| 26 |
|
| 27 |
/** |
| 28 |
* Instance of the WXR generator |
| 29 |
* @var Generator $generator |
| 30 |
*/ |
| 31 |
protected $generator; |
| 32 |
|
| 33 |
/** |
| 34 |
* Authors. |
| 35 |
* @var array |
| 36 |
*/ |
| 37 |
protected $authors = array(); |
| 38 |
|
| 39 |
/** |
| 40 |
* Categories. |
| 41 |
* |
| 42 |
* @var array |
| 43 |
*/ |
| 44 |
protected $categories = array(); |
| 45 |
|
| 46 |
/** |
| 47 |
* URL of the Substack Newsletter. |
| 48 |
* |
| 49 |
* @var string |
| 50 |
*/ |
| 51 |
protected $substack_url; |
| 52 |
|
| 53 |
/** |
| 54 |
* The classnames of all possible embed nodes in the Substack HTML. |
| 55 |
* |
| 56 |
* @var string[] |
| 57 |
*/ |
| 58 |
protected $supported_embeds = array( |
| 59 |
'tweet', |
| 60 |
'instagram', // No longer supported. |
| 61 |
'youtube-wrap', |
| 62 |
'spotify-wrap', |
| 63 |
'soundcloud-wrap', |
| 64 |
'vimeo-wrap', |
| 65 |
'bandcamp-wrap', // Shortcode embed, |
| 66 |
'github-gist', // Not supported in core, using shortcode embed |
| 67 |
); |
| 68 |
|
| 69 |
|
| 70 |
/** |
| 71 |
* Converter constructor. |
| 72 |
* |
| 73 |
* @param Generator $generator Instance of the WXR Generator. |
| 74 |
* @param string $export_file_path Path to the Substack export zip file. |
| 75 |
* @param null $substack_url URL of the Substack newsletter. |
| 76 |
*/ |
| 77 |
public function __construct( Generator $generator, $export_file_path, $substack_url = null ) { |
| 78 |
$this->generator = $generator; |
| 79 |
$this->export_file_path = $export_file_path; |
| 80 |
$this->substack_url = $substack_url; |
| 81 |
} |
| 82 |
|
| 83 |
/** |
| 84 |
* Convert the Substack export to a WXR. |
| 85 |
* |
| 86 |
* @returns WP_Error|void |
| 87 |
* |
| 88 |
* @throws \OxymelException |
| 89 |
*/ |
| 90 |
public function convert() { |
| 91 |
|
| 92 |
if ( ! $this->export_file_path || ! file_exists( $this->export_file_path ) ) { |
| 93 |
return new WP_Error( 'export_file_not_exist', 'The export file does not exist' ); |
| 94 |
} |
| 95 |
|
| 96 |
$this->generator->initialize(); |
| 97 |
|
| 98 |
// Add posts. |
| 99 |
$out = $this->add_posts(); |
| 100 |
|
| 101 |
if ( is_wp_error( $out ) ) { |
| 102 |
return $out; |
| 103 |
} |
| 104 |
|
| 105 |
// Add Authors. |
| 106 |
foreach ( $this->authors as $author ) { |
| 107 |
$this->generator->add_author( $author ); |
| 108 |
} |
| 109 |
|
| 110 |
// Add categories. |
| 111 |
foreach ( $this->categories as $category ) { |
| 112 |
$this->generator->add_category( $category ); |
| 113 |
} |
| 114 |
|
| 115 |
$this->generator->finalize(); |
| 116 |
} |
| 117 |
|
| 118 |
/** |
| 119 |
* Load additional information retrieved through the Substack API into the export zip file. |
| 120 |
* |
| 121 |
* @param int $offset 0-indexed starting offset for the post to start with. |
| 122 |
* @param int $limit Number of posts to process. |
| 123 |
* |
| 124 |
* @return array|WP_Error |
| 125 |
*/ |
| 126 |
public function load_meta_data( $offset = 0, $limit = 1 ) { |
| 127 |
|
| 128 |
$zip = $this->get_export_zip(); |
| 129 |
|
| 130 |
if ( is_wp_error( $zip ) ) { |
| 131 |
return $zip; |
| 132 |
} |
| 133 |
|
| 134 |
$total_count = 0; |
| 135 |
|
| 136 |
foreach ( $this->get_posts() as $idx => $post ) { |
| 137 |
$total_count++; |
| 138 |
|
| 139 |
if ( $idx < $offset || $idx >= $offset + $limit ) { |
| 140 |
continue; |
| 141 |
} |
| 142 |
|
| 143 |
list($id, $slug) = explode( '.', $post['post_id'], 2 ); |
| 144 |
$meta = $this->fetch_post_meta( $slug ); |
| 145 |
|
| 146 |
if ( $meta ) { |
| 147 |
$zip->addFromString( sprintf( 'meta/%s.json', $id ), $meta ); |
| 148 |
} |
| 149 |
} |
| 150 |
|
| 151 |
return array( |
| 152 |
'total' => $total_count, |
| 153 |
'processed' => min( $offset + $limit, $total_count ), |
| 154 |
); |
| 155 |
|
| 156 |
} |
| 157 |
|
| 158 |
/** |
| 159 |
* Convert each Substack post to a WordPress post and add it to the WXR. |
| 160 |
* |
| 161 |
* @return void|WP_Error |
| 162 |
* |
| 163 |
* @throws \OxymelException |
| 164 |
*/ |
| 165 |
protected function add_posts() { |
| 166 |
|
| 167 |
$posts_generator = $this->get_posts(); |
| 168 |
|
| 169 |
if ( is_wp_error( $posts_generator ) ) { |
| 170 |
return $posts_generator; |
| 171 |
} |
| 172 |
|
| 173 |
foreach ( $posts_generator as $post ) { |
| 174 |
|
| 175 |
$id = (int) $post['post_id']; |
| 176 |
$post_meta = $this->get_post_meta_from_export( $id ); |
| 177 |
$post_data = array( |
| 178 |
'id' => $id, |
| 179 |
'title' => $post['title'], |
| 180 |
'content' => $this->convert_html_to_gutenberg( $post['html_body'] ), |
| 181 |
'date' => 'true' === $post['is_published'] ? $post['post_date'] : '', |
| 182 |
'status' => 'true' === $post['is_published'] ? 'publish' : 'draft', |
| 183 |
'post_date_gmt' => $post['post_date'], |
| 184 |
'post_date' => $post['post_date'], |
| 185 |
'post_taxonomies' => array(), |
| 186 |
'metas' => array(), |
| 187 |
); |
| 188 |
|
| 189 |
// If we were able to retrieve more information through the Substack API, we might have |
| 190 |
// author information and comments. |
| 191 |
$post_data['author'] = $post_meta ? $this->get_post_author( $post_meta ) : $this->get_default_author(); |
| 192 |
$post_data['comments'] = $post_meta ? $this->get_post_comments( $post_meta ) : array(); |
| 193 |
|
| 194 |
// Handle podcast posts - prepend an Gutenberg audio block to the post content. |
| 195 |
if ( 'podcast' === $post['type'] && ! empty( $post['podcast_url'] ) ) { |
| 196 |
$post_data = $this->handle_podcast_post( $post_data, $post ); |
| 197 |
} |
| 198 |
|
| 199 |
$this->generator->add_post( $post_data ); |
| 200 |
} |
| 201 |
} |
| 202 |
|
| 203 |
protected function handle_podcast_post( $post_data, $post ) { |
| 204 |
$post_data['content'] = $this->get_audio_block( $post['podcast_url'] ) . $post_data['content']; |
| 205 |
|
| 206 |
// Create a new attachment |
| 207 |
$this->generator->add_post( |
| 208 |
array( |
| 209 |
'title' => urldecode( basename( $post['podcast_url'] ) ), |
| 210 |
'link' => $post['podcast_url'], |
| 211 |
'post_date' => $post_data['post_date'], |
| 212 |
'type' => 'attachment', |
| 213 |
'attachment_url' => $post['podcast_url'], |
| 214 |
'metas' => array( |
| 215 |
array( |
| 216 |
'key' => '_wp_original_image_link', |
| 217 |
'value' => $post['podcast_url'], |
| 218 |
), |
| 219 |
), |
| 220 |
) |
| 221 |
); |
| 222 |
|
| 223 |
// Add this post to the podcast category. |
| 224 |
$this->categories['podcast'] = array( |
| 225 |
'slug' => 'podcast', |
| 226 |
'name' => 'Podcast', |
| 227 |
); |
| 228 |
$post_data['post_taxonomies'][] = array( |
| 229 |
'name' => 'Podcast', |
| 230 |
'slug' => 'podcast', |
| 231 |
'domain' => 'category', |
| 232 |
); |
| 233 |
|
| 234 |
return $post_data; |
| 235 |
} |
| 236 |
|
| 237 |
/** |
| 238 |
* Get a Gutenberg Audio block for the podcast |
| 239 |
* @param $audio_url |
| 240 |
* |
| 241 |
* @return string |
| 242 |
*/ |
| 243 |
protected function get_audio_block( $audio_url ) { |
| 244 |
$code = '<!-- wp:audio --><figure class="wp-block-audio"><audio controls src="%s"></audio><figcaption>Podcast</figcaption></figure><!-- /wp:audio -->'; |
| 245 |
return sprintf( $code, $audio_url ); |
| 246 |
} |
| 247 |
|
| 248 |
protected function get_post_author( $post_meta ) { |
| 249 |
// If we can't get the author information, return a default author. |
| 250 |
if ( empty( $post_meta['publishedBylines'] ) ) { |
| 251 |
return $this->get_default_author(); |
| 252 |
} |
| 253 |
|
| 254 |
$byline = $post_meta['publishedBylines'][0]; |
| 255 |
$this->authors[ $byline['id'] ] = array( |
| 256 |
'login' => $byline['name'], |
| 257 |
'display_name' => $byline['name'], |
| 258 |
'id' => $byline['id'], |
| 259 |
); |
| 260 |
|
| 261 |
return $byline['name']; |
| 262 |
} |
| 263 |
|
| 264 |
protected function get_default_author() { |
| 265 |
$this->authors['unknown'] = array( |
| 266 |
'login' => 'Unknown', |
| 267 |
'display_name' => 'Unknown', |
| 268 |
'id' => 1, |
| 269 |
); |
| 270 |
|
| 271 |
return 'unknown'; |
| 272 |
} |
| 273 |
|
| 274 |
/** |
| 275 |
* Get post comments retrieved through the Substack api. |
| 276 |
* |
| 277 |
* @param array $post_data Additional data about a post retrieved through the Substack Post API. |
| 278 |
* |
| 279 |
* @return mixed |
| 280 |
*/ |
| 281 |
protected function get_post_comments( $post_meta ) { |
| 282 |
if ( empty( $post_meta['comments'] ) ) { |
| 283 |
return array(); |
| 284 |
} |
| 285 |
|
| 286 |
return $this->parse_comments( $post_meta['comments'] ); |
| 287 |
} |
| 288 |
|
| 289 |
/** |
| 290 |
* Recursively parse the comments and prepare the data required for the WXR output. |
| 291 |
* |
| 292 |
* @param array $comments An array of comments provided by the Substack posts API endpoint. |
| 293 |
* @param array $out Output that is ready to be passed to the WXR generator. |
| 294 |
* @param null $parent If we are in a recursive call, the parent must be provided. |
| 295 |
* |
| 296 |
* @return array|mixed |
| 297 |
*/ |
| 298 |
protected function parse_comments( $comments, $out = array(), $parent = null ) { |
| 299 |
foreach ( $comments as $comment ) { |
| 300 |
$out[] = array( |
| 301 |
'id' => $comment['id'], |
| 302 |
'author' => $comment['name'], |
| 303 |
'date' => $comment['date'], |
| 304 |
'date_gmt' => $comment['date'], |
| 305 |
'content' => $comment['body'], |
| 306 |
'parent' => $parent, |
| 307 |
'metas' => array(), |
| 308 |
); |
| 309 |
|
| 310 |
if ( ! empty( $comment['children'] ) ) { |
| 311 |
$out = $this->parse_comments( $comment['children'], $out, (int) $comment['id'] ); |
| 312 |
} |
| 313 |
} |
| 314 |
|
| 315 |
return $out; |
| 316 |
} |
| 317 |
|
| 318 |
/** |
| 319 |
* Convert the content HTML to Gutenberg blocks and return the result. |
| 320 |
* |
| 321 |
* @param string $content The HTML provided by Substack. |
| 322 |
* |
| 323 |
* @return string|string[]|null |
| 324 |
* |
| 325 |
* @todo Load the content as XML to prevent errors from loadHTML. |
| 326 |
*/ |
| 327 |
protected function convert_html_to_gutenberg( $content ) { |
| 328 |
|
| 329 |
$dom = new DOMDocument(); |
| 330 |
|
| 331 |
// By inserting a meta tag with utf-8 encoding we make sure the content is converted to utf-8 |
| 332 |
$content = '<meta http-equiv="Content-Type" content="text/html; charset=utf-8">' . $content; |
| 333 |
@$dom->loadHTML( $content ); //phpcs:ignore WordPress.PHP.NoSilencedErrors.Discouraged |
| 334 |
|
| 335 |
$body = $dom->getElementsByTagName( 'body' )->item( 0 ); |
| 336 |
|
| 337 |
// We don't want to use the DomNodeList because it will change while we are iterating over the nodes. |
| 338 |
$nodes = array(); |
| 339 |
foreach ( $body->childNodes as $node ) { //phpcs:ignore WordPress.NamingConventions.ValidVariableName.UsedPropertyNotSnakeCase |
| 340 |
if ( ! $node instanceof DomElement ) { |
| 341 |
continue; |
| 342 |
} |
| 343 |
$nodes[] = $node; |
| 344 |
} |
| 345 |
|
| 346 |
// We go through the top-level nodes and handle each of them. |
| 347 |
foreach ( $nodes as $idx => $node ) { |
| 348 |
$next_sibling = count( $nodes ) - 1 > $idx ? $nodes[ $idx + 1 ] : null; |
| 349 |
$this->convert_node( $node, $body, $next_sibling ); |
| 350 |
} |
| 351 |
|
| 352 |
// Save as XML otherwise we don't get HTMl5 elements correctly. |
| 353 |
$content = $dom->saveXML( $body ); |
| 354 |
|
| 355 |
// Strip the body tag. |
| 356 |
$content = preg_replace( '/<body>(.+)<\/body>/s', '$1', $content ); |
| 357 |
|
| 358 |
return $content; |
| 359 |
} |
| 360 |
|
| 361 |
|
| 362 |
/** |
| 363 |
* Convert a single node to a Gutenberg block. |
| 364 |
* |
| 365 |
* Tries to convert a given HTML node into a Gutenberg block. |
| 366 |
* |
| 367 |
* @param DomElement $node The node to be converted. |
| 368 |
* @param DomElement $parent The parent of the node to be converted. |
| 369 |
* @param DomElement|null $next_sibling The next sibling of the node to be converted, if it exists. |
| 370 |
* |
| 371 |
*/ |
| 372 |
protected function convert_node( DOMElement $node, DomElement $parent, DomElement $next_sibling = null ) { |
| 373 |
|
| 374 |
$block_name = null; |
| 375 |
$block_attributes = array(); |
| 376 |
|
| 377 |
$node_name = $node->nodeName; //phpcs:ignore WordPress.NamingConventions.ValidVariableName.UsedPropertyNotSnakeCase |
| 378 |
switch ( $node_name ) { |
| 379 |
|
| 380 |
case 'p': |
| 381 |
$block_name = 'wp:paragraph'; |
| 382 |
$class = $node->getAttribute( 'class' ); |
| 383 |
|
| 384 |
// remove empty paragraphs. |
| 385 |
/** @todo Perhaps we can remove all empty nodes, not just paragraphs? */ |
| 386 |
if ( ! $node->childNodes->length ) { //phpcs:ignore WordPress.NamingConventions.ValidVariableName.UsedPropertyNotSnakeCase |
| 387 |
$parent->removeChild( $node ); |
| 388 |
$node = null; |
| 389 |
} |
| 390 |
|
| 391 |
// Button |
| 392 |
if ( 'button-wrapper' === $class ) { |
| 393 |
$node = $this->convert_button_node( $node, $parent ); |
| 394 |
$block_name = 'wp:button'; |
| 395 |
} |
| 396 |
|
| 397 |
break; |
| 398 |
|
| 399 |
case 'blockquote': |
| 400 |
$block_name = 'wp:quote'; |
| 401 |
$node->setAttribute( 'class', 'wp-block-quote' ); |
| 402 |
break; |
| 403 |
|
| 404 |
case 'div': |
| 405 |
case 'iframe': |
| 406 |
$class = $node->getAttribute( 'class' ); |
| 407 |
|
| 408 |
// Preformatted text |
| 409 |
if ( 'preformatted-block' === $class ) { |
| 410 |
$node = $this->convert_preformatted_node( $node, $parent ); |
| 411 |
$block_name = 'wp:preformatted'; |
| 412 |
} |
| 413 |
|
| 414 |
// Images |
| 415 |
if ( 'captioned-image-container' === $class ) { |
| 416 |
$result = $this->convert_image_node( $node, $parent ); |
| 417 |
$node = $result['node']; |
| 418 |
$block_attributes = $result['block_attributes']; |
| 419 |
$block_name = 'wp:image'; |
| 420 |
} |
| 421 |
|
| 422 |
// Horizontal separator |
| 423 |
if ( $node && $node->getElementsByTagName( 'hr' )->length ) { |
| 424 |
$node = $this->convert_separator_node( $node, $parent ); |
| 425 |
$block_name = 'wp:separator'; |
| 426 |
} |
| 427 |
|
| 428 |
// Embeds |
| 429 |
$first_class = explode( ' ', $class ); |
| 430 |
if ( ! empty( $first_class ) && in_array( $first_class[0], $this->supported_embeds, true ) ) { |
| 431 |
$result = $this->convert_embed_node( $node, $parent ); |
| 432 |
$node = $result['node']; |
| 433 |
$block_attributes = $result['block_attributes']; |
| 434 |
$block_name = $result['block_name']; |
| 435 |
} |
| 436 |
|
| 437 |
break; |
| 438 |
|
| 439 |
case 'ol': |
| 440 |
case 'ul': |
| 441 |
$block_name = 'wp:list'; |
| 442 |
|
| 443 |
if ( 'ol' === $node_name ) { |
| 444 |
$block_attributes['ordered'] = true; |
| 445 |
} |
| 446 |
|
| 447 |
break; |
| 448 |
|
| 449 |
case 'pre': |
| 450 |
$block_name = 'wp:code'; |
| 451 |
$node->setAttribute( 'class', 'wp-block-code' ); |
| 452 |
break; |
| 453 |
|
| 454 |
case 'h1': |
| 455 |
case 'h2': |
| 456 |
case 'h3': |
| 457 |
case 'h4': |
| 458 |
case 'h5': |
| 459 |
case 'h6': |
| 460 |
$block_name = 'wp:heading'; |
| 461 |
$block_attributes['level'] = (int) substr( $node_name, 1, 1 ); |
| 462 |
break; |
| 463 |
|
| 464 |
case 'a': |
| 465 |
$class = $node->getAttribute( 'class' ); |
| 466 |
if ( 'image-link image2' === trim( $class ) ) { |
| 467 |
$result = $this->convert_image_node( $node, $parent ); |
| 468 |
$node = $result['node']; |
| 469 |
$block_attributes = $result['block_attributes']; |
| 470 |
$block_name = 'wp:image'; |
| 471 |
} |
| 472 |
|
| 473 |
break; |
| 474 |
|
| 475 |
} |
| 476 |
|
| 477 |
if ( ! $block_name || ! $node ) { |
| 478 |
return; |
| 479 |
} |
| 480 |
|
| 481 |
// Create the Gutenberg block code |
| 482 |
$attributes_part = ''; |
| 483 |
if ( count( $block_attributes ) ) { |
| 484 |
$attributes_part = ' ' . wp_json_encode( $block_attributes ); |
| 485 |
} |
| 486 |
$block_open = new DOMComment( ' ' . $block_name . $attributes_part . ' ' ); |
| 487 |
$block_close = new DOMComment( ' /' . $block_name . ' ' ); |
| 488 |
|
| 489 |
$parent->insertBefore( $block_open, $node ); |
| 490 |
|
| 491 |
$next_sibling |
| 492 |
? $parent->insertBefore( $block_close, $next_sibling ) |
| 493 |
: $parent->appendChild( $block_close ); |
| 494 |
} |
| 495 |
|
| 496 |
/** |
| 497 |
* Convert a preformatted text node to valid Gutenberg markup. |
| 498 |
* |
| 499 |
* @param DomElement $node The node to be converted. |
| 500 |
* @param DomElement $parent The parent of the node. |
| 501 |
* |
| 502 |
* @return DomElement The converted node. |
| 503 |
*/ |
| 504 |
protected function convert_preformatted_node( DomElement $node, DomElement $parent ) { |
| 505 |
|
| 506 |
$node_value = $node->getElementsByTagName( 'pre' )[0]->textContent; |
| 507 |
$new_node = new DomElement( 'pre', $node_value ); |
| 508 |
$parent->replaceChild( $new_node, $node ); |
| 509 |
$new_node->setAttribute( 'class', 'wp-block-preformatted' ); |
| 510 |
|
| 511 |
return $new_node; |
| 512 |
} |
| 513 |
|
| 514 |
/** |
| 515 |
* Handle a button node. |
| 516 |
* |
| 517 |
* @param DomElement $node The node to be converted. |
| 518 |
* @param DomElement $parent The parent of the node. |
| 519 |
* |
| 520 |
* @return DomElement |
| 521 |
* |
| 522 |
* @todo Support multiple types of buttons. For now buttons are removed. |
| 523 |
*/ |
| 524 |
protected function convert_button_node( DomElement $node, DomElement $parent ) { |
| 525 |
$parent->removeChild( $node ); |
| 526 |
return null; |
| 527 |
} |
| 528 |
|
| 529 |
/** |
| 530 |
* Convert an image node to a Gutenberg valid markup. |
| 531 |
* |
| 532 |
* @param DomElement $node The node to be converted. |
| 533 |
* @param DomElement $parent The parent of the node. |
| 534 |
* |
| 535 |
* @return array An array containing the Block attributes and the new node. |
| 536 |
* |
| 537 |
* @todo If the node is a (a) link we need to make this image a link as well. |
| 538 |
*/ |
| 539 |
protected function convert_image_node( DomElement $node, DomElement $parent ) { |
| 540 |
|
| 541 |
// Check if the image needs to be resized |
| 542 |
// Can we already upload the image here? |
| 543 |
/** @var DomElement $image */ |
| 544 |
$image = $node->getElementsByTagName( 'img' )[0]; |
| 545 |
|
| 546 |
// if there is no image we can't proceed. |
| 547 |
if ( ! $image ) { |
| 548 |
$parent->removeChild( $node ); |
| 549 |
return array( |
| 550 |
'block_attributes' => array(), |
| 551 |
'node' => null, |
| 552 |
); |
| 553 |
} |
| 554 |
|
| 555 |
$block_attributes = array(); |
| 556 |
|
| 557 |
$new_node = new DomElement( 'figure' ); |
| 558 |
|
| 559 |
$parent->replaceChild( $new_node, $node ); |
| 560 |
|
| 561 |
$classes = array( 'wp-block-image', 'size-large' ); |
| 562 |
|
| 563 |
// The data we need is set as json data attribute on the img node. |
| 564 |
$image_data = json_decode( $image->getAttribute( 'data-attrs' ), true ); |
| 565 |
|
| 566 |
// Add the image as an attachement post to the WXR. |
| 567 |
$this->generator->add_post( |
| 568 |
array( |
| 569 |
'title' => urldecode( basename( $image_data['src'] ) ), |
| 570 |
'link' => $image_data['src'], |
| 571 |
'type' => 'attachment', |
| 572 |
'attachment_url' => $image_data['src'], |
| 573 |
'metas' => array( |
| 574 |
array( |
| 575 |
'key' => '_wp_original_image_link', |
| 576 |
'value' => $image_data['src'], |
| 577 |
), |
| 578 |
), |
| 579 |
) |
| 580 |
); |
| 581 |
|
| 582 |
// Create the new image element. |
| 583 |
$new_image = new DomElement( 'img' ); |
| 584 |
$new_node->appendChild( $new_image ); |
| 585 |
$new_image->setAttribute( 'src', $image_data['src'] ); |
| 586 |
$new_image->setAttribute( 'alt', $image_data['alt'] ); |
| 587 |
|
| 588 |
// Deal with resizing. |
| 589 |
if ( $image_data['resizeWidth'] ) { |
| 590 |
$classes[] = 'is-resized'; |
| 591 |
$new_image->setAttribute( 'width', $image_data['resizeWidth'] ); |
| 592 |
$block_attributes['width'] = $image_data['resizeWidth']; |
| 593 |
} |
| 594 |
|
| 595 |
// Set the classes on the figure element. |
| 596 |
$new_node->setAttribute( 'class', implode( ' ', $classes ) ); |
| 597 |
|
| 598 |
$block_attributes['sizeSlug'] = 'large'; |
| 599 |
$block_attributes['linkDestination'] = 'none'; |
| 600 |
|
| 601 |
return array( |
| 602 |
'block_attributes' => $block_attributes, |
| 603 |
'node' => $new_node, |
| 604 |
); |
| 605 |
} |
| 606 |
|
| 607 |
/** |
| 608 |
* Convert the node to a valid Gutenberg separator block. |
| 609 |
* |
| 610 |
* @param DomElement $node The node to be converted. |
| 611 |
* @param DomElement $parent The parent of the node to be converted. |
| 612 |
* |
| 613 |
* @return DomElement The new node. |
| 614 |
*/ |
| 615 |
protected function convert_separator_node( DomElement $node, DomElement $parent ) { |
| 616 |
|
| 617 |
$new_node = new DomElement( 'hr' ); |
| 618 |
$parent->replaceChild( $new_node, $node ); |
| 619 |
$new_node->setAttribute( 'class', 'wp-block-separator' ); |
| 620 |
|
| 621 |
return $new_node; |
| 622 |
} |
| 623 |
|
| 624 |
/** |
| 625 |
* Convert a node that represents an embed to valid Gutenberg embed block markup. |
| 626 |
* |
| 627 |
* @param DomElement $node The node to be converted. |
| 628 |
* @param DomElement $parent The parent of the node to be coverted. |
| 629 |
* |
| 630 |
* @return array Containing the block_name, block_attributes and node. |
| 631 |
*/ |
| 632 |
protected function convert_embed_node( DomElement $node, DomElement $parent ) { |
| 633 |
|
| 634 |
$first_class = explode( ' ', $node->getAttribute( 'class' ) )[0]; |
| 635 |
|
| 636 |
switch ( $first_class ) { |
| 637 |
|
| 638 |
case 'youtube-wrap': |
| 639 |
$output = $this->convert_youtube_embed( $node, $parent ); |
| 640 |
break; |
| 641 |
|
| 642 |
case 'vimeo-wrap': |
| 643 |
$output = $this->convert_vimeo_embed( $node, $parent ); |
| 644 |
break; |
| 645 |
|
| 646 |
case 'soundcloud-wrap': |
| 647 |
$output = $this->convert_soundcloud_embed( $node, $parent ); |
| 648 |
break; |
| 649 |
|
| 650 |
case 'tweet': |
| 651 |
$output = $this->convert_tweet_embed( $node, $parent ); |
| 652 |
break; |
| 653 |
|
| 654 |
case 'spotify-wrap': |
| 655 |
$output = $this->convert_spotify_embed( $node, $parent ); |
| 656 |
break; |
| 657 |
|
| 658 |
case 'bandcamp-wrap': |
| 659 |
$output = $this->convert_bandcamp_embed( $node, $parent ); |
| 660 |
break; |
| 661 |
|
| 662 |
case 'github-gist': |
| 663 |
$output = $this->convert_gist_embed( $node, $parent ); |
| 664 |
break; |
| 665 |
|
| 666 |
default: |
| 667 |
$parent->removeChild( $node ); |
| 668 |
$output = array( |
| 669 |
'node' => null, |
| 670 |
'block_attributes' => array(), |
| 671 |
'block_name' => null, |
| 672 |
); |
| 673 |
|
| 674 |
} |
| 675 |
|
| 676 |
return $output; |
| 677 |
} |
| 678 |
|
| 679 |
/** |
| 680 |
* Convert the embed node into Gutenberg markup for a Youtube embed. |
| 681 |
* |
| 682 |
* @param DomElement $node The node to be converted. |
| 683 |
* @param DomElement $parent The parent of the node to be coverted. |
| 684 |
* |
| 685 |
* @return array Containing the block_name, block_attributes and node. |
| 686 |
*/ |
| 687 |
protected function convert_youtube_embed( DomElement $node, DomElement $parent ) { |
| 688 |
|
| 689 |
$data_attributes = json_decode( $node->getAttribute( 'data-attrs' ), true ); |
| 690 |
|
| 691 |
$block_attributes = array( |
| 692 |
'url' => 'https://youtu.be/' . $data_attributes['videoId'], |
| 693 |
'type' => 'video', |
| 694 |
'providerNameSlug' => 'youtube', |
| 695 |
'responsive' => true, |
| 696 |
'className' => 'wp-embed-aspect-16-9 wp-has-aspect-ratio', |
| 697 |
); |
| 698 |
|
| 699 |
$node = $this->replace_embed_node( $node, $parent, $block_attributes['url'] ); |
| 700 |
$classes = 'wp-block-embed is-type-video is-provider-youtube wp-block-embed-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio'; |
| 701 |
$node->setAttribute( 'class', $classes ); |
| 702 |
|
| 703 |
return array( |
| 704 |
'block_name' => 'wp:embed', |
| 705 |
'block_attributes' => $block_attributes, |
| 706 |
'node' => $node, |
| 707 |
); |
| 708 |
} |
| 709 |
|
| 710 |
/** |
| 711 |
* Convert the embed node into Gutenberg markup for a Vimeo embed. |
| 712 |
* |
| 713 |
* @param DomElement $node The node to be converted. |
| 714 |
* @param DomElement $parent The parent of the node to be coverted. |
| 715 |
* |
| 716 |
* @return array Containing the block_name, block_attributes and node. |
| 717 |
*/ |
| 718 |
protected function convert_vimeo_embed( DomElement $node, DomElement $parent ) { |
| 719 |
|
| 720 |
$data_attributes = json_decode( $node->getAttribute( 'data-attrs' ), true ); |
| 721 |
|
| 722 |
$block_attributes = array( |
| 723 |
'url' => 'https://vimeo.com/' . $data_attributes['videoId'], |
| 724 |
'type' => 'video', |
| 725 |
'providerNameSlug' => 'vimeo', |
| 726 |
'responsive' => true, |
| 727 |
'className' => 'wp-embed-aspect-16-9 wp-has-aspect-ratio', |
| 728 |
); |
| 729 |
|
| 730 |
$node = $this->replace_embed_node( $node, $parent, $block_attributes['url'] ); |
| 731 |
$classes = 'wp-block-embed is-type-video is-provider-vimeo wp-block-embed-vimeo wp-embed-aspect-16-9 wp-has-aspect-ratio'; |
| 732 |
$node->setAttribute( 'class', $classes ); |
| 733 |
|
| 734 |
return array( |
| 735 |
'block_name' => 'wp:embed', |
| 736 |
'block_attributes' => $block_attributes, |
| 737 |
'node' => $node, |
| 738 |
); |
| 739 |
} |
| 740 |
|
| 741 |
/** |
| 742 |
* Convert the embed node into Gutenberg markup for a Soundcloud embed. |
| 743 |
* |
| 744 |
* @param DomElement $node The node to be converted. |
| 745 |
* @param DomElement $parent The parent of the node to be coverted. |
| 746 |
* |
| 747 |
* @return array Containing the block_name, block_attributes and node. |
| 748 |
*/ |
| 749 |
protected function convert_soundcloud_embed( DomElement $node, DomElement $parent ) { |
| 750 |
|
| 751 |
$data_attributes = json_decode( $node->getAttribute( 'data-attrs' ), true ); |
| 752 |
|
| 753 |
$block_attributes = array( |
| 754 |
'url' => $data_attributes['url'], |
| 755 |
'type' => 'rich', |
| 756 |
'providerNameSlug' => 'soundcloud', |
| 757 |
'responsive' => true, |
| 758 |
'className' => 'wp-embed-aspect-4-3 wp-has-aspect-ratio', |
| 759 |
); |
| 760 |
|
| 761 |
$node = $this->replace_embed_node( $node, $parent, $block_attributes['url'] ); |
| 762 |
$classes = 'wp-block-embed is-type-rich is-provider-soundcloud wp-block-embed-soundcloud wp-embed-aspect-4-3 wp-has-aspect-ratio'; |
| 763 |
$node->setAttribute( 'class', $classes ); |
| 764 |
|
| 765 |
return array( |
| 766 |
'block_name' => 'wp:embed', |
| 767 |
'block_attributes' => $block_attributes, |
| 768 |
'node' => $node, |
| 769 |
); |
| 770 |
} |
| 771 |
|
| 772 |
/** |
| 773 |
* Convert the embed node into Gutenberg markup for a Tweet embed. |
| 774 |
* |
| 775 |
* @param DomElement $node The node to be converted. |
| 776 |
* @param DomElement $parent The parent of the node to be coverted. |
| 777 |
* |
| 778 |
* @return array Containing the block_name, block_attributes and node. |
| 779 |
*/ |
| 780 |
protected function convert_tweet_embed( DomElement $node, DomElement $parent ) { |
| 781 |
|
| 782 |
$data_attributes = json_decode( $node->getAttribute( 'data-attrs' ), true ); |
| 783 |
|
| 784 |
$block_attributes = array( |
| 785 |
'url' => $data_attributes['url'], |
| 786 |
'type' => 'rich', |
| 787 |
'providerNameSlug' => 'twitter', |
| 788 |
'responsive' => true, |
| 789 |
); |
| 790 |
|
| 791 |
$node = $this->replace_embed_node( $node, $parent, $block_attributes['url'] ); |
| 792 |
$classes = 'wp-block-embed is-type-rich is-provider-twitter wp-block-embed-twitter'; |
| 793 |
$node->setAttribute( 'class', $classes ); |
| 794 |
|
| 795 |
return array( |
| 796 |
'block_name' => 'wp:embed', |
| 797 |
'block_attributes' => $block_attributes, |
| 798 |
'node' => $node, |
| 799 |
); |
| 800 |
} |
| 801 |
|
| 802 |
/** |
| 803 |
* Convert the embed node into Gutenberg markup for a Spotify embed. |
| 804 |
* |
| 805 |
* @param DomElement $node The node to be converted. |
| 806 |
* @param DomElement $parent The parent of the node to be coverted. |
| 807 |
* |
| 808 |
* @return array Containing the block_name, block_attributes and node. |
| 809 |
*/ |
| 810 |
protected function convert_spotify_embed( DomElement $node, DomElement $parent ) { |
| 811 |
|
| 812 |
$data_attributes = json_decode( $node->getAttribute( 'data-attrs' ), true ); |
| 813 |
|
| 814 |
$block_attributes = array( |
| 815 |
'url' => $data_attributes['url'], |
| 816 |
'type' => 'rich', |
| 817 |
'providerNameSlug' => 'spotify', |
| 818 |
'responsive' => true, |
| 819 |
'className' => 'wp-embed-aspect-9-16 wp-has-aspect-ratio', |
| 820 |
); |
| 821 |
|
| 822 |
$node = $this->replace_embed_node( $node, $parent, $block_attributes['url'] ); |
| 823 |
$classes = 'wp-block-embed is-type-rich is-provider-spotify wp-block-embed-spotify wp-embed-aspect-9-16 wp-has-aspect-ratio'; |
| 824 |
$node->setAttribute( 'class', $classes ); |
| 825 |
|
| 826 |
return array( |
| 827 |
'block_name' => 'wp:embed', |
| 828 |
'block_attributes' => $block_attributes, |
| 829 |
'node' => $node, |
| 830 |
); |
| 831 |
} |
| 832 |
|
| 833 |
/** |
| 834 |
* Converts the node into a shortcode for Bandcamp. |
| 835 |
* |
| 836 |
* The shortcode is currently not supported in Core but is available by enabling the embeds module |
| 837 |
* of the Jetpack plugin. |
| 838 |
* |
| 839 |
* @example [bandcamp width=350 height=470 album=473417827 size=large bgcol=ffffff linkcol=0687f5 tracklist=false] |
| 840 |
* |
| 841 |
* @param DomElement $node The node to be converted. |
| 842 |
* @param DomElement $parent The parent of the node to be coverted. |
| 843 |
* |
| 844 |
* @return array |
| 845 |
*/ |
| 846 |
protected function convert_bandcamp_embed( DomElement $node, DomElement $parent ) { |
| 847 |
|
| 848 |
$data_attributes = json_decode( $node->getAttribute( 'data-attrs' ), true ); |
| 849 |
|
| 850 |
// The embed URL contains the attributes for the shortcode. Here we extract them and add them to the shortcode. |
| 851 |
preg_match_all( '/[a-z]+=[a-z0-9]+/', $data_attributes['embed_url'], $matches ); |
| 852 |
$shortcode = sprintf( '[bandcamp %s]', implode( ' ', $matches[0] ) ); |
| 853 |
|
| 854 |
$new_node = new DOMText( $shortcode ); |
| 855 |
$parent->replaceChild( $new_node, $node ); |
| 856 |
|
| 857 |
return array( |
| 858 |
'block_name' => 'wp:shortcode', |
| 859 |
'block_attributes' => array(), |
| 860 |
'node' => $new_node, |
| 861 |
); |
| 862 |
} |
| 863 |
|
| 864 |
/** |
| 865 |
* Convert a Github Gist node into a shortcode. |
| 866 |
* |
| 867 |
* Tries to get the Gist id from the raw link or removes the entire Gist if the ID can not be determined. |
| 868 |
* |
| 869 |
* @param DomElement $node The node to be converted. |
| 870 |
* @param DomElement $parent The parent of the node to be coverted. |
| 871 |
* |
| 872 |
* @return array |
| 873 |
*/ |
| 874 |
protected function convert_gist_embed( DomElement $node, DomElement $parent ) { |
| 875 |
|
| 876 |
$a_elements = $node->getElementsByTagName( 'a' ); |
| 877 |
|
| 878 |
$url = $a_elements->length > 0 |
| 879 |
? $a_elements[0]->getAttribute( 'href' ) |
| 880 |
: null; |
| 881 |
|
| 882 |
if ( ! $url || ! preg_match( '/\/([a-z0-9]+)\/raw/', $a_elements[0]->getAttribute( 'href' ), $matches ) ) { |
| 883 |
$parent->removeChild( $node ); |
| 884 |
return array( |
| 885 |
'node' => null, |
| 886 |
'block_attributes' => array(), |
| 887 |
'block_name' => null, |
| 888 |
); |
| 889 |
} |
| 890 |
|
| 891 |
$shortcode = sprintf( '[gist https://gist.github.com/%s]', $matches[1] ); |
| 892 |
|
| 893 |
$new_node = new DOMText( $shortcode ); |
| 894 |
$parent->replaceChild( $new_node, $node ); |
| 895 |
|
| 896 |
return array( |
| 897 |
'block_name' => 'wp:shortcode', |
| 898 |
'block_attributes' => array(), |
| 899 |
'node' => $new_node, |
| 900 |
); |
| 901 |
} |
| 902 |
|
| 903 |
/** |
| 904 |
* Replace the Substack Embed node with embed markup that is valid for Gutenberg. |
| 905 |
* |
| 906 |
* Returns the replacement node. |
| 907 |
* |
| 908 |
* @param DomElement $node |
| 909 |
* @param DomElement $parent |
| 910 |
* |
| 911 |
* @return DomElement |
| 912 |
*/ |
| 913 |
protected function replace_embed_node( DomElement $node, DomElement $parent, $content ) { |
| 914 |
$new_node = new DomElement( 'figure' ); |
| 915 |
$wrapper = new DomElement( 'div' ); |
| 916 |
|
| 917 |
$parent->replaceChild( $new_node, $node ); |
| 918 |
$new_node->appendChild( $wrapper ); |
| 919 |
$wrapper->setAttribute( 'class', 'wp-block-embed__wrapper' ); |
| 920 |
|
| 921 |
$content = new DOMText( $content ); |
| 922 |
$new_node->getElementsByTagName( 'div' )[0]->appendChild( $content ); |
| 923 |
|
| 924 |
return $new_node; |
| 925 |
} |
| 926 |
|
| 927 |
|
| 928 |
/** |
| 929 |
* Retrieve additional post information through the Substack Post API. |
| 930 |
* |
| 931 |
* The most important data we are after includes author information and comments as this currently is not provided |
| 932 |
* in the export file. |
| 933 |
* |
| 934 |
* It is important to note that comments might not be included or might not contain any information |
| 935 |
* if the comments are only visible to paid users or if post itself is only accessible to paid users. |
| 936 |
* |
| 937 |
* The completeness of information in the response depends on the type of the post (paid vs. public). |
| 938 |
* |
| 939 |
* @param string $slug The slug of the post. |
| 940 |
* |
| 941 |
* @return string|null Returns a JSON string with post information or null if it could not be retrieved. |
| 942 |
*/ |
| 943 |
protected function fetch_post_meta( $slug ) { |
| 944 |
|
| 945 |
// If the substack url is not set, we skip this step. |
| 946 |
if ( ! $this->substack_url ) { |
| 947 |
return null; |
| 948 |
} |
| 949 |
|
| 950 |
$post_url = sprintf( '%s/api/v1/posts/%s?all_comments=true', $this->substack_url, $slug ); |
| 951 |
|
| 952 |
$response = wp_remote_get( $post_url, array( 'redirection' => 0 ) ); |
| 953 |
|
| 954 |
if ( is_wp_error( $response ) || 200 !== $response['response']['code'] ) { |
| 955 |
return null; |
| 956 |
} |
| 957 |
|
| 958 |
return wp_remote_retrieve_body( $response ); |
| 959 |
} |
| 960 |
|
| 961 |
/** |
| 962 |
* Get meta info from the substack export zip. Returns null if no meta was found. |
| 963 |
* |
| 964 |
* @param int $id Substack Post ID. |
| 965 |
* |
| 966 |
* @return array|null |
| 967 |
*/ |
| 968 |
protected function get_post_meta_from_export( $id ) { |
| 969 |
$zip = $this->get_export_zip(); |
| 970 |
|
| 971 |
if ( is_wp_error( $zip ) ) { |
| 972 |
return null; |
| 973 |
} |
| 974 |
|
| 975 |
$meta = $zip->getFromName( sprintf( 'meta/%s.json', $id ) ); |
| 976 |
|
| 977 |
return $meta |
| 978 |
? json_decode( $meta, true ) |
| 979 |
: null; |
| 980 |
} |
| 981 |
|
| 982 |
/** |
| 983 |
* Returns a generator yielding posts retrieved from the Substack export. |
| 984 |
* |
| 985 |
* If a there was a problem retrieving the Zip file, a WP_Error will be returned. |
| 986 |
* |
| 987 |
* @return \Generator|WP_Error |
| 988 |
*/ |
| 989 |
public function get_posts() { |
| 990 |
|
| 991 |
$zip = $this->get_export_zip(); |
| 992 |
|
| 993 |
if ( is_wp_error( $zip ) ) { |
| 994 |
return $zip; |
| 995 |
} |
| 996 |
|
| 997 |
return $this->get_posts_generator( $zip ); |
| 998 |
} |
| 999 |
|
| 1000 |
protected function get_posts_generator( ZipArchive $zip ) { |
| 1001 |
$post_csv = $zip->getFromName( 'posts.csv' ); |
| 1002 |
|
| 1003 |
$posts = explode( "\n", trim( $post_csv ) ); |
| 1004 |
$map = str_getcsv( array_shift( $posts ) ); |
| 1005 |
|
| 1006 |
foreach ( $posts as $post ) { |
| 1007 |
$post = str_getcsv( $post, ',' ); |
| 1008 |
$post = array_combine( $map, $post ); |
| 1009 |
$post['html_body'] = $zip->getFromName( sprintf( 'posts/%s.html', $post['post_id'] ) ); |
| 1010 |
yield $post; |
| 1011 |
} |
| 1012 |
} |
| 1013 |
|
| 1014 |
/** |
| 1015 |
* Get a ZipArchive instance of the export file or return an error if it failed. |
| 1016 |
* |
| 1017 |
* @return WP_Error|ZipArchive The zip archive or a WP_error instance on failure. |
| 1018 |
*/ |
| 1019 |
protected function get_export_zip() { |
| 1020 |
|
| 1021 |
if ( ! class_exists( 'ZipArchive' ) ) { |
| 1022 |
return new WP_Error( 'missing_zip_extension', __( 'Could not unzip the substack export file.' ) ); |
| 1023 |
} |
| 1024 |
|
| 1025 |
$zip = new ZipArchive(); |
| 1026 |
$success = $zip->open( $this->export_file_path ); |
| 1027 |
|
| 1028 |
if ( true !== $success || 0 === $zip->numFiles ) { //phpcs:ignore WordPress.NamingConventions.ValidVariableName.UsedPropertyNotSnakeCase -- ZipArcive property |
| 1029 |
return new WP_Error( 'invalid_export_file', __( 'Could not unzip the substack export file.' ) ); |
| 1030 |
} |
| 1031 |
|
| 1032 |
// If posts.csv was not found in the zip archive, the export is invalid. |
| 1033 |
if ( false === $zip->getFromName( 'posts.csv' ) ) { |
| 1034 |
return new WP_Error( 'no_posts_in_export_file', __( 'The export file is not a valid Substack export, no posts.csv was found in the archive. ' ) ); |
| 1035 |
} |
| 1036 |
|
| 1037 |
return $zip; |
| 1038 |
} |
| 1039 |
} |
| 1040 |
|
