@a83/orbiter-admin 0.2.0

package/src/wp-importer.js

@@ -0,0 +1,299 @@
+/**
+ * wp-importer.js
+ * WordPress WXR import logic for Orbiter.
+ *
+ * Exports:
+ *   parseWXR(xmlText)        → { site, postTypes, items, attachments }
+ *   buildImportPlan(parsed)  → summary for preview UI
+ *   executeImport(db, plan, options) → { imported, skipped, errors, mediaResults }
+ */
+
+import { XMLParser } from 'fast-xml-parser';
+import { createRequire } from 'node:module';
+import { randomUUID } from 'node:crypto';
+
+const _require = createRequire(import.meta.url);
+const TurndownService = _require('turndown');
+
+// ── Turndown (HTML → Markdown) ────────────────────────────────────────────
+const td = new TurndownService({
+  headingStyle:    'atx',
+  bulletListMarker: '-',
+  codeBlockStyle:  'fenced',
+  hr:              '---',
+});
+
+// Preserve WordPress shortcodes as inline code rather than dropping them
+td.addRule('shortcode', {
+  filter: (node) => node.nodeName === '#text' && /\[.+?\]/.test(node.nodeValue),
+  replacement: (content) => content,
+});
+
+// ── XML parser config ─────────────────────────────────────────────────────
+const xmlParser = new XMLParser({
+  ignoreAttributes:    false,
+  attributeNamePrefix: '@_',
+  cdataPropName:       '__cdata',
+  isArray: (name) => ['item', 'wp:postmeta', 'category', 'wp:author', 'wp:category', 'wp:tag'].includes(name),
+  parseAttributeValue: true,
+  trimValues:          true,
+});
+
+// Helper: unwrap CDATA or plain text value
+function v(val) {
+  if (val === undefined || val === null) return '';
+  if (typeof val === 'object' && val.__cdata !== undefined) return String(val.__cdata ?? '');
+  return String(val);
+}
+
+// Helper: safe slug from title
+function slugify(str) {
+  return str.toLowerCase()
+    .replace(/[äÄ]/g, 'ae').replace(/[öÖ]/g, 'oe').replace(/[üÜ]/g, 'ue').replace(/ß/g, 'ss')
+    .replace(/[^a-z0-9]+/g, '-').replace(/^-+|-+$/g, '') || 'untitled';
+}
+
+// WordPress post status → Orbiter status
+function mapStatus(wpStatus) {
+  return wpStatus === 'publish' ? 'published' : 'draft';
+}
+
+// Default schema per post type
+function schemaFor(postType) {
+  if (postType === 'page') {
+    return {
+      title:   { type: 'string',   required: true,  label: 'Title' },
+      body:    { type: 'richtext', required: false, label: 'Content' },
+      date:    { type: 'date',     required: false, label: 'Date' },
+    };
+  }
+  // posts + all custom post types
+  return {
+    title:   { type: 'string',   required: true,  label: 'Title' },
+    excerpt: { type: 'string',   required: false, label: 'Excerpt' },
+    body:    { type: 'richtext', required: false, label: 'Content' },
+    tags:    { type: 'array',    required: false, label: 'Tags & Categories' },
+    date:    { type: 'date',     required: false, label: 'Date' },
+    image:   { type: 'media',    required: false, label: 'Featured Image' },
+    author:  { type: 'string',   required: false, label: 'Author' },
+  };
+}
+
+// Collection ID from post type (safe slug)
+function collectionId(postType) {
+  if (postType === 'post') return 'posts';
+  if (postType === 'page') return 'pages';
+  return slugify(postType).replace(/-/g, '_');
+}
+
+// ── parseWXR ─────────────────────────────────────────────────────────────
+export function parseWXR(xmlText) {
+  const root    = xmlParser.parse(xmlText);
+  const channel = root?.rss?.channel ?? {};
+  const items   = Array.isArray(channel.item) ? channel.item : (channel.item ? [channel.item] : []);
+
+  const site = {
+    title: v(channel.title),
+    url:   v(channel.link),
+  };
+
+  // Build attachment map: wp:post_id → wp:attachment_url
+  const attachmentMap = {};
+  for (const item of items) {
+    const type = v(item['wp:post_type']);
+    if (type === 'attachment') {
+      const wpId = v(item['wp:post_id']);
+      attachmentMap[wpId] = {
+        url:      v(item['wp:attachment_url']),
+        filename: v(item['wp:post_name']) || v(item.title),
+        mimeType: v(item['wp:attachment_metadata']?.mime_type) || '',
+      };
+    }
+  }
+
+  // Parse non-attachment items
+  const contentItems = [];
+  for (const item of items) {
+    const type = v(item['wp:post_type']);
+    if (type === 'attachment' || type === 'nav_menu_item' || type === 'revision') continue;
+    if (!type) continue;
+
+    const meta    = Array.isArray(item['wp:postmeta']) ? item['wp:postmeta'] : [];
+    const thumbId = v(meta.find(m => v(m['wp:meta_key']) === '_thumbnail_id')?.['wp:meta_value'] ?? '');
+
+    const categories = (Array.isArray(item.category) ? item.category : (item.category ? [item.category] : []))
+      .filter(c => c['@_domain'] === 'category')
+      .map(c => v(c));
+    const tags = (Array.isArray(item.category) ? item.category : (item.category ? [item.category] : []))
+      .filter(c => c['@_domain'] === 'post_tag')
+      .map(c => v(c));
+
+    const rawHtml   = v(item['content:encoded']);
+    const rawExcerpt = v(item['excerpt:encoded']);
+    const wpDate    = v(item['wp:post_date']).split(' ')[0] || '';
+
+    contentItems.push({
+      postType:    type,
+      wpId:        v(item['wp:post_id']),
+      title:       v(item.title),
+      slug:        slugify(v(item['wp:post_name']) || v(item.title)),
+      status:      mapStatus(v(item['wp:status'])),
+      rawHtml,
+      rawExcerpt,
+      tags:        [...new Set([...categories, ...tags])].filter(Boolean),
+      date:        wpDate,
+      author:      v(item['dc:creator']),
+      thumbId,     // WP attachment ID → resolve via attachmentMap
+      pubDate:     v(item.pubDate),
+    });
+  }
+
+  // Group by post type
+  const postTypes = {};
+  for (const item of contentItems) {
+    if (!postTypes[item.postType]) postTypes[item.postType] = [];
+    postTypes[item.postType].push(item);
+  }
+
+  return { site, postTypes, attachmentMap };
+}
+
+// ── buildImportPlan ───────────────────────────────────────────────────────
+export function buildImportPlan(parsed) {
+  const { site, postTypes, attachmentMap } = parsed;
+  const mediaCount = Object.keys(attachmentMap).length;
+
+  const types = Object.entries(postTypes).map(([type, items]) => ({
+    postType:     type,
+    collectionId: collectionId(type),
+    count:        items.length,
+    published:    items.filter(i => i.status === 'published').length,
+    drafts:       items.filter(i => i.status === 'draft').length,
+    schema:       schemaFor(type),
+  }));
+
+  return { site, types, mediaCount, attachmentCount: mediaCount };
+}
+
+// ── executeImport ─────────────────────────────────────────────────────────
+// options: { selectedTypes: string[], downloadMedia: boolean, onDuplicate: 'skip'|'overwrite', podPath: string }
+export async function executeImport(db, parsed, options) {
+  const { selectedTypes, downloadMedia, onDuplicate } = options;
+  const { postTypes, attachmentMap } = parsed;
+
+  const results = {
+    collections: [],
+    imported:    0,
+    skipped:     0,
+    overwritten: 0,
+    mediaOk:     0,
+    mediaFailed: 0,
+    errors:      [],
+  };
+
+  // ── 1. Ensure collections exist ─────────────────────────────────────────
+  for (const type of selectedTypes) {
+    const colId    = collectionId(type);
+    const existing = db.getCollection(colId);
+    if (!existing) {
+      const label = type === 'post' ? 'Posts' : type === 'page' ? 'Pages'
+        : type.charAt(0).toUpperCase() + type.slice(1).replace(/_/g, ' ');
+      db.db.prepare('INSERT INTO _collections (id, label, schema) VALUES (?, ?, ?)')
+        .run(colId, label, JSON.stringify(schemaFor(type)));
+      results.collections.push(colId);
+    }
+  }
+
+  // ── 2. Download media (optional) ─────────────────────────────────────────
+  // orbiterMediaId map: wpAttachmentId → orbiter media UUID
+  const orbiterMediaId = {};
+  if (downloadMedia) {
+    for (const [wpId, att] of Object.entries(attachmentMap)) {
+      if (!att.url) continue;
+      try {
+        const res  = await fetch(att.url, { signal: AbortSignal.timeout(15000) });
+        if (!res.ok) throw new Error(`HTTP ${res.status}`);
+        const buf  = Buffer.from(await res.arrayBuffer());
+        const mime = res.headers.get('content-type')?.split(';')[0]?.trim()
+                     || att.mimeType || 'application/octet-stream';
+        const filename = att.url.split('/').pop().split('?')[0] || `media-${wpId}`;
+        const now  = new Date().toISOString().replace('T', ' ').replace(/\.\d{3}Z$/, '');
+        const id   = randomUUID();
+        db.db.prepare(
+          'INSERT OR IGNORE INTO _media (id, filename, mime_type, size, data, alt, created_at) VALUES (?, ?, ?, ?, ?, ?, ?)'
+        ).run(id, filename, mime, buf.length, buf, att.filename || filename, now);
+        orbiterMediaId[wpId] = id;
+        results.mediaOk++;
+      } catch (err) {
+        results.mediaFailed++;
+        results.errors.push(`Media ${att.url}: ${err.message}`);
+      }
+    }
+  }
+
+  // ── 3. Import entries ────────────────────────────────────────────────────
+  const now = new Date().toISOString().replace('T', ' ').replace(/\.\d{3}Z$/, '');
+
+  for (const type of selectedTypes) {
+    const colId = collectionId(type);
+    const items = postTypes[type] ?? [];
+
+    for (const item of items) {
+      try {
+        // Check for duplicate slug
+        const existing = db.db
+          .prepare('SELECT id FROM _entries WHERE collection_id = ? AND slug = ?')
+          .get(colId, item.slug);
+
+        if (existing && onDuplicate === 'skip') {
+          results.skipped++;
+          continue;
+        }
+
+        // Convert HTML → Markdown
+        const bodyMd   = item.rawHtml   ? td.turndown(item.rawHtml)   : '';
+        const excerptMd = item.rawExcerpt ? td.turndown(item.rawExcerpt) : '';
+
+        // Resolve featured image
+        const imageId = item.thumbId ? (orbiterMediaId[item.thumbId] ?? null) : null;
+
+        const data = {
+          title:   item.title || 'Untitled',
+          body:    bodyMd,
+          excerpt: excerptMd,
+          tags:    item.tags,
+          date:    item.date,
+          author:  item.author,
+          image:   imageId,
+        };
+
+        // Remove page-irrelevant fields
+        if (type === 'page') {
+          delete data.excerpt;
+          delete data.tags;
+          delete data.author;
+          delete data.image;
+        }
+
+        const created = item.date || now;
+
+        if (existing && onDuplicate === 'overwrite') {
+          db.db.prepare(
+            'UPDATE _entries SET data = ?, status = ?, updated_at = ? WHERE id = ?'
+          ).run(JSON.stringify(data), item.status, now, existing.id);
+          results.overwritten++;
+        } else {
+          const id = randomUUID();
+          db.db.prepare(
+            'INSERT INTO _entries (id, collection_id, slug, data, status, created_at, updated_at) VALUES (?, ?, ?, ?, ?, ?, ?)'
+          ).run(id, colId, item.slug, JSON.stringify(data), item.status, created, now);
+          results.imported++;
+        }
+      } catch (err) {
+        results.errors.push(`Entry "${item.slug}": ${err.message}`);
+      }
+    }
+  }
+
+  return results;
+}