frootbox / migrate-frootbox1

v0.1.0 2026-11-04 11:59 UTC

Requires

• php: >=8.5
• doctrine/dbal: ^4.5
• frootbox/core: ^0.1
• symfony/console: ^8.1
• symfony/yaml: ^8.1

MIT

• cms
• migration
• frootbox

Published on Packagist.org on 2026-11-04 16:36 UTC

README

Moves a website from Frootbox 1 to Frootbox 2. The tool reads the database and the files/ folder of the old installation and changes nothing there. It is an extension for the development of a website and never part of the standard bundle.

Analyse

bin/frootbox fb1:analyze 'pdo-mysql://user:secret@localhost/old?charset=utf8mb4' \
    --files=/path/to/frootbox/files/site \
    --mapping=migration/frootbox1.yaml

The report lists the pages with their blocks and plugins, every block type with the fields found in its stored content, plugins with their layout and options, objects such as teasers, widgets inside texts, navigations and redirects. Problems are listed at the end: files missing on disk, widgets whose row is gone, redirects to pages that do not exist. The command exits with 1 if there are any.

With --mapping it writes a first mapping file, never over an existing one. Blocks and layouts of the same name in the Frootbox 2 project are filled in; everything else stays ~ to be chosen.

Frootbox 1 keeps texts and files by UID, e.g. Frootbox-Persistence-Content-Blocks-Repositories-Blocks:4:title, so the fields of a block are what is stored, not what the template declares. Numbered fields count as one (text-1, text-2 → text-#).

Mapping

The mapping says what becomes of each thing in Frootbox 1. false leaves it out; the report of the import lists it.

languages: { de-DE: de }
domains: [www.example.com]          # links to these count as internal
layouts:
  Default: acme/theme/Content       # or { layout: …, area: main }
blocks:
  Text:                             # the type of the block in Frootbox 1
    block: acme/theme/Text
    fields:
      title: title                  # field of Frootbox 1: field of Frootbox 2
      title.supertitle: eyebrow     # a value of the headline
      text-#: columns.#.body        # numbered fields into a list
    widgets: { Core/Images/Thumbnail: image }  # the first image widget in the text
  BreakerImage: false
plugins:
  Core/Navigation/Teaser@Index06: { block: acme/theme/Stage, fields: { items.1.background: image } }
  Core/ContactForms/FormTeaser@Index01: { block: frootbox/forms/Form, form: true }
  Core/FileManager/CopyrightFilesList@Index01: { block: frootbox/theme-base/ImageCredits }
items:
  Teaser: { fields: { title: title, text: text, link: { to: link, text: More } } }
navigations: { BannerMain: main }
site:
  fields: { brand: Example, contact: { target: '{page:5}', text: Contact } }
media:                              # by the file name in Frootbox 1
  team.jpg: { alt: 'The team in front of the practice', copyright: 'Jane Doe' }

{page:5} is the new page of page 5 in Frootbox 1, {config.x} a value in the configuration of the old block. Each value is converted by the type of the target field: HTML becomes rich text, links to old pages and aliases become links to the new pages, files become media. Under media the mapping fills in an alt text or a copyright that a file lacks in Frootbox 1.

Import

bin/frootbox fb1:import 'pdo-mysql://user:secret@localhost/old?charset=utf8mb4' \
    --files=/path/to/frootbox/files/site \
    --mapping=migration/mapping.yaml --out=migration/seed --publish
bin/frootbox seed migration/seed/site.yaml

The import writes no database. It writes a seed: site.yaml with the steps (media, pages, content, forms, navigations, fields of the website, redirects for every old address), the images in media/ and report.md with what a person should look at. The seed runs on an empty website, so a project builds itself anew from the old database on every run, e.g. in a script that backs up, installs and seeds. Without --publish all of it stays a draft.

Hidden blocks, plugins and entries of navigations stay behind, as do redirect pages: their address becomes a redirect to the target.

Each page takes over its description for search engines (the one of its address wins, as in Frootbox 1, cut to 300 characters) and whether search engines or the site search leave it out. Frootbox 2 has no title of its own for search engines yet, so such a title goes into the report.

Source and issues

This package is developed in the Frootbox monorepo and mirrored here read-only. Questions, bugs and ideas go there, see the documentation.