Migrating from another CMS

Total CMS ships importers for four other sources, under Utilities in the sidebar’s Import group: Import Alloy, Import RSS Feed, Import Total CMS 1 and Import WordPress. Each maps its source onto object properties and queues the result into a collection you choose, leaving the source data untouched. For a spreadsheet or JSON export, see Importing Data instead.

They only ever create objects. The RSS importer skips an entry whose id already exists, so re-polling a feed is safe; the others queue it anyway and the job fails, because Total CMS rejects a duplicate id rather than overwriting. The “update existing objects” rule in Importing Data covers the CSV and JSON forms, not these.

WordPress

Utilities → Import WordPress. Requires the Standard edition or higher (the route is gated on the rss_import feature).

Generate the file in WordPress under Tools → Export and upload the .xml (WXR). Analyze it first: that step reports the post count, date range, every category, tag and author found, and the first ten posts. The import step takes the same file plus a target collection.

Only items whose wp:post_type is post are imported. Fields map like this:

WXR field	Object property
wp:post_name, else a title slug	id
title	title
content:encoded	content
excerpt:encoded	summary
wp:post_date	date, as ISO 8601
dc:creator	author
category with domain="category"	category, comma separated, Uncategorized dropped
category with domain="post_tag"	tags, comma separated
link	url
wp:status	draft — anything but publish is a draft

The featured image is followed: _thumbnail_id is resolved against the export’s attachment items, downloaded, and imported into image. Everything else is not imported — pages, custom post types, the rest of the media library, images embedded in post bodies (they keep pointing at the old site), comments, users, menus, and anything a plugin or shortcode produced. The author is a name string, not a user object. “Import as drafts” overrides the status row above. No CLI command.

Alloy

Utilities → Import Alloy. No edition requirement.

You supply four folder names relative to the document root, defaulting to posts, image-uploads, embeds and droplets. Analyze first for a count of what was found.

• Blog posts — each .md file is YAML frontmatter plus a Markdown body, landing in the blog collection (created with the blog schema if missing). A filename of YYYY-MM-DD_slug.md supplies date and id; otherwise the whole filename is the id. title, author, category, tags, draft and summary come from the frontmatter, and the body becomes HTML in content. topper resolves against the image uploads folder into image, with topperalt as imageAlt.
• Embeds — converted from Markdown to HTML and stored as styledtext in the styledtext collection.
• Droplets — type: text becomes an object in the text collection; type: image becomes one in the image collection, resolved against the image uploads folder. Any other type is skipped and logged.

No CLI command.

Total CMS 1

Utilities → Import Total CMS 1. No edition requirement. The page auto-detects a cms-data folder in your document root, or you can type an absolute path.

Each v1 folder becomes a collection: blogs get one each on the blog-legacy schema (a .posturl file sets the collection URL and pretty-URL flag; if a non-empty collection with that id exists, the import goes to <id>-one), date, depot, feed, file, gallery, image and text map to collections of the same name, and videos become url objects. Templates are the other half of a v1 migration: the folder-by-folder table and the full macro → Twig mapping live in Migrating from Total CMS v1.

RSS, Atom and JSON feeds

Utilities → Import RSS Feed. Requires the Standard edition or higher.

Give it any RSS, Atom or JSON feed URL and analyze it to see the feed title, description, entry count and a per-entry preview. The object id is a slug of the entry title.

The default mapping is title → title, content → content, summary → summary, date → date, author → author, categories → categories and link → media. Change any of it in the field mapping panel; an empty target skips that field. An entry with no content falls back to its description (or JSON summary). An image is taken from the entry’s enclosure, media:content or media:thumbnail — or image / banner_image in a JSON Feed — downloaded, and written to image.

The same import runs from the CLI, which is what you schedule with cron:

tcms rss:import https://example.com/feed.xml blog --no-draft \
  --map title=heading --map image=featured

--map accepts title, content, summary, date, author, categories, link and image. Use --user-agent when a host blocks the default (TotalCMS/<version> (+https://totalcms.co)). The admin page’s “Schedule with cron” panel builds this command line for you.

How the import runs

Every importer above writes into the job queue rather than saving immediately, so “queued N items” counts jobs, not saved objects. Drain the queue with the scheduled job processor:

php <install_dir>/resources/bin/tcms jobs:process

Watch progress under Utilities → Job Queue Manager, which also builds that cron line. A bad item does not stop an import: the failure is logged and the importer moves on. Those messages go to importer.log, readable in Utilities → Log Analyzer, with warnings and errors mirrored into totalcms.log. Each imported object fires import.created rather than object.created — see Events if listeners or automations watch for new content.

After importing

Check a few objects — dates, images, and anything that was a category or tag list — and publish if you imported as drafts.

Ids come from the source slug wherever one exists (wp:post_name, the Alloy filename, the v1 permalink), so old URLs are recoverable: set the collection’s URL to match and redirect old paths at the web server. Feed entries have no source slug, so their ids are slugs of their titles.

Finally, rebuild your search index. With indexOnSave enabled, every imported object is pushed to the search provider as its job runs — slow, and hard on an external provider. Turn it off for a large migration, then run:

tcms search:reindex

See Search Backends for the provider settings.