jackrabbit-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Alexander Klimetschek" <aklim...@day.com>
Subject Re: Does Mirage support different file types( or it is possible only with Jackrabbit)
Date Mon, 13 Oct 2008 15:06:28 GMT
On Mon, Oct 13, 2008 at 4:17 PM, Nayak Vishal <VishalN@nsb.no> wrote:
> Hello everyone,
> In our project we were planning to use mirage with Alfresco.
> But as per my knowdedge mirage doesn't support different types of files
> except( txt and html). Plz update me if it supports.
> Is it possible to switch mirage with jackrabbit completely for accessing
> alfresco.

I think you might be misguided. Let me explain - please note that I
don't know the details of Alfresco or Mirage, just what they tell on
their front pages, so excuse if it is me who is wrong here ;-):

- JCR is a standard for a hierarchical content repository, defining a
Java API for it (comparable to JDBC); it is well-suited for CMS
- Jackrabbit is a complete JCR implementation, thus it is a content
repository; it provides some remote connections, eg. WebDAV, but it is
not a CMS
- Mirage seems to be both a CMS and a JCR implementation
- Don't know if you can plug in Jackrabbit as content repository (aka
the backend) for the Mirage CMS part (but if they use the JCR API it
should be possible)
- Alfresco is both a CMS and includes a JCR implementation
- Not sure if you can plug in Jackrabbit into Alfresco for the
backend, but again, I think it should be possible

So switching from Mirage to Jackrabbit for usage in Alfresco sounds
quite difficult to me, except if Mirage is rather a JCR implementation
itself that can be plugged into Alfresco than a CMS.

Jackrabbit is AFAIK the most feature-complete open source JCR implementation.

> Through wicket framework can we access jackrabbit API for file uploading

By using the JCR API (which is Java, thus usable with Wicket), storing
files in the repository is easy (using the built-in node types
nt:folder and nt:file). JCR is explicitly designed to handle binary
data, such as files, efficiently.

Note the difference between the JCR API and Jackrabbit API -
Jackrabbit itself only contains a small additional API for
non-standardized things - in general you should only use the JCR API.
That allows you to switch the JCR implementation.

Automatic text extraction for various file types (as discussed in this
thread before) and including them in the full text index, which can be
used through "jcr:contains" queries, is not exactly defined in the JCR
standard, but is available in Jackrabbit. Since it happens
automatically on saving binary properties, it doesn't require the
client to call any API, just the presence of a indexing configuration
(which by default includes standard file types as mentioned, thus it
works out of the box).

Hope this explaination helps,

Alexander Klimetschek

View raw message