Skip to main content

Feeds protocol implementation and search integration

Project description


Plone integration for Google Search Appliance.

Compatible with Plone 4.


Use Plone’s setup tool to install the profile: “Integration to Google Search Appliance”.

Then visit the control panel to configure your appliance. Also in the control panel are various settings to customize the site integration.


This package exists to provide a light-weight integration between Plone and a Google Search Appliance service. Its philosophy is to let the appliance do the heavy lifting instead of Plone. The “64” postfix corresponds to the version of the search appliance software at the time of writing: 6.4.

There’s an alternative in Matous Hora’s collective.gsa package which takes a different approach and is in return a much more complex piece of software. The package also integrates more closely with Plone by patching into the default portal catalog.

The present package – collective.gsa64 – is about site search and just that.

Feed protocol implementation

The package comes with an implementation for the Feed Protocol API. When a content item for which search is enabled (use Plone’s search control panel to configure) is created, modified or deleted, a feed is sent to the search appliance to update the status.

There are two modes of operation: web (implicitly enabled for a feed with the name "web") and url and metadata mode (any other name).

The metadata for the second mode is drawn directly from HTML using the plone.htmlhead content provider. As such this corresponds directly to the metadata which would be indexed using crawling.

You can enable include the standard Dublin Core metadata using the site control panel.

An example of a viewlet definition in ZCML that renders additional metadata:


Note that feed requests are issued after the browser request has ended and will not delay server response.

Search view

The setup script adds a Python Script object to the portal root with the id "search". This takes over the normal skin lookup and redirects the traversal to the @@gsa-search view.

You can enable or disable the search appliance results view (and operation in general) using the control panel.

The results view is by default generated using an included browser view with a template that generates markup similar to Plone’s own search results view. However, it’s also possible to have an XSLT transform applied directly to the XML result document and show this instead.

For custom projects, it may be necessary to subclass the default search view implementation. Here’s an example of a modified query:

from collective.gsa64.browser import search

class SearchView(search.SearchView):
    def build_query(self):
        query = super(SearchView, self).build_query()

        # always include hits for "plone"
        query['q'] = '(%s) OR plone' % query['q']

        return query

It may be desirable to change the display of the individual search results. You can accomplish this by providing your own search templates and fill out the "result" METAL slot:

class SearchView(search.SearchView):
    template = ViewPageTemplateFile("templates/")

    def master_search_template(self):
        return super(SearchView, self).template

The templates/ template:

<html xmlns="" xml:lang="en"

  <metal:result fill-slot="result">

    <h4 tal:content="structure result/heading" />



Each result object contains the following keys:

  • description: HTML document description
  • heading: HTML document title
  • url: Document’s display URL
  • creator: DC creator value
  • normalized_content_type: Normalized Plone content type id
  • metadata: Dictionary that maps metadata keys to values

Content types filter

Plone’s standard types not searched setting is applied to HTML output via a browser viewlet "meta-robots" which renders a noarchive meta tag for content types excluded from search.


This package was designed and developed by the following authors.

  • Malthe Borch
  • Mustapha Benali

Development was funded by Headnet.


1.0-beta4 (2011-11-30)


  • Added settings for client features:

    KeyMatch, related queries, filters, etc.

    These are available in the control panel. [mustapha]


  • Handle the case where the product is not installed. [mustapha]
  • Handle the case of empty description or heading. [mustapha]
  • Added browser layer to enable browser components through the setup. [mustapha]

1.0-beta3 (2011-01-10)

  • Recover from component lookup errors that result from trying to render HTML meta-tags for use as metadata.
  • Disable search result filtering (at least until we can provide an interface that presents the normal ‘we have filtered your results’ dialogue).
  • Ignore event handler if registry is unavailable (missing setup of
  • Fixed an issue where a checkPermission call would raise an exception because no interaction was established yet (for instance in a test setup).

1.0-beta2 (2010-12-20)

  • If an HTML meta-tag has no content attribute, drop it from the feed.

1.0-beta1 (2010-12-17)

  • Initial public release.

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

collective.gsa64-1.0-beta4.tar.gz (37.0 kB view hashes)

Uploaded source

Supported by

AWS AWS Cloud computing Datadog Datadog Monitoring Facebook / Instagram Facebook / Instagram PSF Sponsor Fastly Fastly CDN Google Google Object Storage and Download Analytics Huawei Huawei PSF Sponsor Microsoft Microsoft PSF Sponsor NVIDIA NVIDIA PSF Sponsor Pingdom Pingdom Monitoring Salesforce Salesforce PSF Sponsor Sentry Sentry Error logging StatusPage StatusPage Status page