Listing and working with Plone content objects using plone.app.contentlisting

This is valid for Plone 4.1 upwards.

Many of the operations for customizations, templates, views and portlets in Plone are related to lists of content objects. Their sources can be different, although usually they are some sort of catalog search, the contents of a particular folder or a list of objects from a relation.

To make it simpler to work with these, we have made plone.app.contentlisting, which ensures that lists of content objects always behave in the same way and according to predefined interfaces, regardless of what the source of the objects are. The integrator shouldn't have to care whether the list of objects came from the catalog, an ORM or they are the actual objects.

Making or getting a contentListing

The typical way to get a contentlisting is to call one of two built-in views:

Listing the contents of a Folder or Collection

In Page templates getting the contents of a folder or the results of a collection is as simple as this:

context/@@contentlisting

Every template-writer's dream ;)

Note

In previous versions there was only support to list the contents of a folder with context/@@folderListing. There was no collection support. The @@folderListing view is kept for compatibility, but we encourage you to use the @@contentlisting instead.

A real example of listing the titles of the content objects of a folder:

<ul>
  <li tal:repeat="item context/@@contentlisting" tal:content="item/Title"/>
</ul>

The context in which it is called defines which folder is listed or which collection results are queried.

You can also use Python expressions to be able to pass parameters, like which content type or review state you want to use:

<li tal:repeat="item python:context.restrictedTraverse('@@contentlisting')(portal_type='Document')">

Batching can be done like this:

<ul tal:define="
    Batch python:modules['Products.CMFPlone'].Batch;
    b_size python:int(request.get('b_size', 20));
    b_start python:int(request.get('b_start', 0));
    results python:context.restrictedTraverse('@@contentlisting')(batch=True, b_size=b_size, b_start=b_start);
    batch python:Batch(results, b_size, b_start);">
  <li tal:repeat="item results"
      tal:content="item/Title" />
  <div metal:use-macro="context/batch_macros/macros/navigation" />
</ul>

Note that you iterate directly over the results that you get from @@contentlisting. The definition of batch is only used in the batch_macros call.

In Python a ContentListing of a particular folder's contents can be fetched by using:

>>> path.to.your.folder.restrictedTraverse('@@contentlisting')()

Exactly the same for collections:

>>> path.to.your.collection.restrictedTraverse('@@contentlisting')()

The contentlisting view called above implements all the logic the old getFolderContents script in Plone used to do. The old script has been left in place to not break compatibility for customizations and add-ons that might depend on its particular return values.

Rolling your own with adaption

At the time of writing, all parts of Plone do not yet return 'contentlistings' when asked for lists of content. It was impossible to change this everywhere without breaking backwards compatibility. Therefore you may have to convert your sequence of stuff to a contentlisting manually.

To do this, you need to import and adapt:

>>> from plone.app.contentlisting.interfaces import IContentListing
>>> catalog = getToolByName(self.portal, 'portal_catalog')
>>> results = catalog.searchResults()
>>> contentlist = IContentListing(results)
>>> print(contentlist)
<plone.app.contentlisting.contentlisting.ContentListing object ...>

The contentListing, its properties and behaviors

Now, you no longer need to worry whether you have a bunch of catalog brains or the actual objects (or fake objects for that sake). As long as you have a contentlisting, you know what you can expect from it. You also know what you can expect from each item within it - a content listing object.

The content listing is a normal iterator that we can loop over and do all sorts of stuff you normally can do with sequences.

contentListingObjects, the items inside the sequence

The contentListingObjects are wrapper objects, each representing a content object in the site. Their intention is to be predictable so you can always call at least a common base set of methods on the objects listed.

You do not have to be aware whether the object originates from a brain, a full object or something else. If you try to call a method or access an attribute of the object and the wrapper is not aware of it, it will silently fetch the real object and delegate the call to it. This means you can treat your objects as you would any other -- even writing to it.

Methods of contentlistingObjects

getId() -

Returns the object id in its container for example my-example-page.

getObject() -

Returns the real object

def getDataOrigin() -

The origin of the data for the object.

getPath() -

Path to the object, relative to the site root for example /artifacts/my-example-page

getURL()-

Full url to the object, including the site root for example http://my.site.com/artifacts/my-example-page

uuid() -

Unique content identifier for example an uuid from plone.uuid The only real point of it is to be unique. It can for example look like this b0e80776-d41d-4f48-bf9e-7cb1aebabad5.

getSize() -

Size in bytes for example 24.

review_state() -

Workflow review state for example published.

ContentTypeClass() -

A normalized type name that identifies the object in listings. Used for CSS styling, for example content-type-page.

Title() -

Return a single string, the DCMI Title element (resource name). For example My example page.

Description() -

Return the DCMI Description element (resource summary). Result is a natural language description of this object. Description is a plain text string describing the object. It should not contain HTML or similar.

Type() -

Return the DCMI Type element (resource type). Result is a human-readable message id for the resource (typically the Title of its type info object). For example u'Page' from the plone domain.

listCreators() -

Return a sequence of DCMI Creator elements (resource authors). Depending on the implementation, this returns the full name(s) of the author(s) of the content object or their ids. For example Jane Smith.

Creator() -

Return the first DCMI Creator element, or an empty string. For example Jane Smith.

Subject() -

Return a sequence of DCMI Subject elements (resource keywords). Result is zero or more keywords associated with the content object. These are the tags in Plone. For example ['Ecology', 'Sustainability'].

Publisher() -

Return the DCMI Publisher element (resource publisher). Result is the full formal name of the entity or person responsible for publishing the resource. For example Plone Foundation.

listContributors() -

Return a sequence of DCMI Contributor elements (resource collaborators). Return zero or more collaborators (beyond those returned by listCreators).

Contributors() -

Deprecated alias for listContributors.

Date(zone=None) -

Return the DCMI Date element (default resource date). Result is a string, formatted 'YYYY-MM-DD H24:MN:SS TZ'. The zone keyword is not yet supported (but is part of the DublinCore interface and has to stay)

CreationDate(zone=None) -

Return the DCMI Date element (date resource created). Result is a string, formatted 'YYYY-MM-DD H24:MN:SS TZ'. The zone keyword is not yet supported (but is part of the DublinCore interface and has to stay)

EffectiveDate(zone=None) -

Return the DCMI Date element (date resource becomes effective). Result is a string, formatted 'YYYY-MM-DD H24:MN:SS TZ', or None. The zone keyword is not yet supported (but is part of the DublinCore interface and has to stay)

ExpirationDate(zone=None) -

Return the DCMI Date element (date resource expires). Result is a string, formatted 'YYYY-MM-DD H24:MN:SS TZ', or None. The zone keyword is not yet supported (but is part of the DublinCore interface and has to stay)

ModificationDate(zone=None) -

DCMI Date element - date resource last modified. Result is a string, formatted 'YYYY-MM-DD H24:MN:SS TZ'. The zone keyword is not yet supported (but is part of the DublinCore interface and has to stay)

Format() -

Return the DCMI Format element (resource format). Result is the resource's MIME type (e.g. text/html, image/png, etc.).

Identifier() -

Return the DCMI Identifier element (resource ID). Result is a unique ID (a URL) for the resource.

Language() -

DCMI Language element (resource language). Result it the RFC language code (e.g. en-US, pt-BR) for the resource.

Rights() -

Return the DCMI Rights element (resource copyright). Return a string describing the intellectual property status, if any, of the resource.

isVisibleInNav() -

Return whether this object will be visible in a navigation view.

MimeTypeIcon():

Return mimetype icon from mimetype registry if contenttype is

File else None.