/**
 * Normalize relative and legacy asset URLs within HTML / HTML内の相対パスと旧URLを正規化する
 *
 * 処理対象:
 *  - src, href, poster, srcset
 *  - data-*, data-src, data-srcset, data-bg, data-background 等
 *  - style 属性内の url(...)
 *  - <source> (picture/Video) の srcset, src
 *
 * @param string $html Raw HTML string to process.
 * @return string Normalized HTML.
 */
function etude_normalize_product_html( string $html ) : string {
    if ( trim( $html ) === '' ) {
        return $html;
    }

    $previous = libxml_use_internal_errors( true );

    $dom        = new DOMDocument( '1.0', 'UTF-8' );
    $html_utf8  = mb_convert_encoding( $html, 'HTML-ENTITIES', 'UTF-8' );
    $dom->loadHTML(
        '<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />' . $html_utf8,
        LIBXML_HTML_NOIMPLIED | LIBXML_HTML_NODEFDTD
    );

    $xpath = new DOMXPath( $dom );

    // 1) 代表的なURL属性（単一URL）
    $single_url_attrs = [ 'src', 'href', 'poster' ];
    foreach ( $single_url_attrs as $attr ) {
        foreach ( $xpath->query( '//*[@' . $attr . ']' ) as $el ) {
            /** @var DOMElement $el */
            $el->setAttribute( $attr, etude_map_asset_url( $el->getAttribute( $attr ) ) );
        }
    }

    // 2) srcset
    foreach ( $xpath->query( '//*[@srcset]' ) as $el ) {
        $el->setAttribute( 'srcset', etude_normalize_attribute_urls( $el->getAttribute( 'srcset' ), 'srcset' ) );
    }

    // 3) style (background: url(...))
    foreach ( $xpath->query( '//*[@style]' ) as $el ) {
        $el->setAttribute( 'style', etude_normalize_attribute_urls( $el->getAttribute( 'style' ), 'style' ) );
    }

    // 4) data-* 系（data-src, data-srcset, data-bg, data-background など）を一括
    foreach ( $xpath->query( '//*' ) as $el ) {
        /** @var DOMElement $el */
        if ( ! $el->hasAttributes() ) continue;
        $to_set = [];
        foreach ( $el->attributes as $attr ) {
            $name  = strtolower( $attr->name );
            if ( $name === 'src' || $name === 'href' || $name === 'poster' || $name === 'style' || $name === 'srcset' ) {
                continue; // 既に処理済み
            }
            if ( strpos( $name, 'data-' ) === 0 || preg_match( '/(?:background|bg|src|srcset)/i', $name ) ) {
                $to_set[ $attr->name ] = etude_normalize_attribute_urls( $attr->value, $attr->name );
            }
        }
        foreach ( $to_set as $n => $v ) {
            $el->setAttribute( $n, $v );
        }
    }

    // 5) <source>（video/picture）で src も拾う
    foreach ( $xpath->query( '//source[@src]' ) as $el ) {
        $el->setAttribute( 'src', etude_map_asset_url( $el->getAttribute( 'src' ) ) );
    }
    foreach ( $xpath->query( '//source[@srcset]' ) as $el ) {
        $el->setAttribute( 'srcset', etude_normalize_attribute_urls( $el->getAttribute( 'srcset' ), 'srcset' ) );
    }

    $output = $dom->saveHTML();
    $output = preg_replace( '#^<meta[^>]+?>#', '', $output ?? '' );

    libxml_clear_errors();
    libxml_use_internal_errors( $previous );

    return $output !== '' ? $output : $html;
}

/**
 * Normalize URLs embedded in arbitrary attribute values / 任意属性値内のURLを正規化する
 *
 * @param string $value Attribute value string.
 * @param string $attr_name (optional) 現在処理中の属性名ヒント（srcset/style等）
 * @return string Normalized attribute value.
 */
function etude_normalize_attribute_urls( string $value, string $attr_name = '' ) : string {
    if ( trim( $value ) === '' ) {
        return $value;
    }

    // srcset は「url<空白>descriptor」のカンマ区切り
    if ( strtolower( $attr_name ) === 'srcset' ) {
        $parts = array_map( 'trim', explode( ',', $value ) );
        foreach ( $parts as &$part ) {
            if ( $part === '' ) continue;
            $segments   = preg_split( '/\s+/', $part, 2 );
            $url_part   = etude_map_asset_url( $segments[0] ?? '' );
            $descriptor = $segments[1] ?? '';
            $part       = trim( $url_part . ( $descriptor ? ' ' . $descriptor : '' ) );
        }
        unset( $part );
        return implode( ', ', $parts );
    }

    // style="background: url(...)" 等： url(...) をすべて拾う
    if ( strtolower( $attr_name ) === 'style' ) {
        return preg_replace_callback(
            // 区切りを ~ にすることで / や # を多用しても衝突しない
            '~url\(\s*(["\']?)([^)]+?)\1\s*\)~i',
            function( $m ) {
                $mapped = etude_map_asset_url( trim( $m[2] ) );
                return 'url(' . $mapped . ')';
            },
            $value
        );
    }

    // data-* など任意属性：複合文字列中の URL トークンだけを置換
    //  - http(s)://... もしくは //...（プロトコル相対）
    //  - / から始まるルート相対
    return preg_replace_callback(
        // 1 個の安全なパターンで抽出（否定の後読みは \w と @ のみ）
        // () 全体が URL 本体。) や ] は文字クラス外に置き、デリミタは ~ を使用
        '~(?<![\w@])((?:https?:)?//[^\s"\'\)]*|/(?:[^\s"\'\)]*))~u',
        function( $m ) {
            return etude_map_asset_url( $m[1] );
        },
        $value
    );
}
<br />
<b>Warning</b>:  Cannot modify header information - headers already sent by (output started at /home/etudejapan/etude.jp/public_html/wp-content/themes/etudehouse/inc/product-importer.php:1) in <b>/home/etudejapan/etude.jp/public_html/wp-includes/sitemaps/class-wp-sitemaps-renderer.php</b> on line <b>190</b><br />
<?xml version="1.0" encoding="UTF-8"?>
<?xml-stylesheet type="text/xsl" href="https://etude.jp/wp-sitemap.xsl" ?>
<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9"><url><loc>https://etude.jp/category/uncategorized/</loc></url></urlset>