If we do not pass the content when building an essence,
`after_initialize` hooks do not have acc...

This adds an Alchemy::EssenceHeadline. It has the following features:
When used as a preview tex...

Touch Elements only after update

When an essence has a related object, such as a picture or an
attachment, we need to break cache...

Convert "Visit page" button to "Link to new tab"

The benefit of this action was that it would unlock the page before
allowing the user to visit i...

This allows an editor to stay in their Alchemy window while previewing
a page.

Allow creating an EssenceRichtext without a content

This worked before adding the sanitization. The problem is that if there
is no content, we pass ...

* Add "Sanitized Body" to Alchemy::EssenceRichtext

Often, we want HTML, but not all the thing...

Allow instantiating a content on an unpersisted valid element

Rails don't need a content, and neiter do we.

When copying an Essence, it needs to know its content - otherwise it
can't figure out what optio...

This reduces the number of dependencies we have to the ones
we actually need. Removing the need ...

Fix add nested element with multiple nestable elements

If an element allows multiple nestable elements the element select
dialog appears. That was miss...

So that it actually tests what it does:
Immediately creating a nested element.

Also add a pendi...

Destroy public version if public checkbox is unset

MySQL does not support milliseconds, so we need to adopt the tests

In order unpublish a page via the "Published" checkbox
in the page settings dialog we need to de...

Create new contents on demand

If new contents have been defined on the elements definition, but they are
not existing yet we c...

Instead of in the view

Fixes paste element and create element with autogenerated nested elements

Autogenerated nested elements need to have the same page_version as the parent.

We need the page_version_id not the page_id

Render site layout with block

Adds a repository class that mimics the ActiveRecord interface of elements, but uses the eager l...

Rufo uses the same default as Rubocop here.

By removing the override we use the dewfault of Rub...

The elements repository uses array methods to filter, order and limit
elements on a eager loaded...

This is way fast than using create and since we only use
array methods in the repository
class i...

These scopes are used on the ElementsFinder and needs to
be supported.

This mimics the ActiveRecord interface of elements, but
uses the eager loaded elements and POR m...

It is very common to chain ActiveRecord scopes to filter elements.

This makes it possible with ...

If a page has a published page version with nested elements
the nested elements got deleted bef...

NodeTree is being used for rendering the Menu in tree
structure. This Tree could be made generi...

Update rubocop config and stick version

New cops have been introduced.

We only support Ruby 2.5 and above.

We use hound ci to run linting. They only support some versions of rubocop.

Lets stick with the...

Fix factory loading

Our factories now get appended to the definition paths of factory bot.

Also FactoryBot.reload i...

Before the method returned a list of individual file paths.
FactoryBot is able to load files fro...

Add Page Versions

With this change it is possible to define something like

<%= render_site_layout do %>
<%= ren...

This reverts commit 9bddcd2883aa328841318e8b0f89041e268fb1c4.

Always use the draft version of the page set or passed into
render_elements if we are in preview...

Since we load the draft version elements in the elements window,
we need to load the pages draft...

In render_elements we use either a `@page_version` or the pages public
version to load elements ...

Only use the assigned page version if there is no page
given via the `:from_page` option, so tha...

A page is considered public? if a public version exists and this is published.

Instead of creating a new public version every time we publish a page,
we re-use the already pub...

Any page that does not have any published version is considered not public

These timestamps now live on the page version.

The values still exists on the table for histori...

Only elements that have a public version should be returned
by the API.

It turned out that using .find_each messed with the order of elements during copy.

Instead of the current page the ElementsFinder now uses
the given page version to load elements....

We want to make sure that we only have one published version
at a given time. Ideally we would a...

Instead of passing the page_id we use the page_version_id
to create, update and order elements i...

and use the page version instead.

The element trash has been removed recently.

While loading a page tree from the API include elements
from public page version.

Create a public version as well and copy the public elements over.

Points to the gem and allows to run webpacker in the dummy app

and associate with page

Tidy Rake task to remove duplicate legacy URLs

Change Factory loading mechanism to FactoryBots supported mechanism

Add rake task to remove trashed elements

We build these factories such that they could be independently loaded.
However, FactoryBot is bu...

Fix element re-ordering

Reordering elements does not need to change the elements page id anymore
since we removed the tr...

This reverts commit c9b1c6857ad08016d0115e817df8fa324ad1e140.

We removed the element trashed with Alchemy 6.0

This task helps to clean up the database with a...