Lockie wrote:
> Hello all,
> I'm running my Xwiki as an information management Intranet alongside a
> networked drive that holds hundreds of documents. Based off the TODO
> tutorial, I made a system of inputting these documents as objects on
> individual pages that then appear on a filterable-sortable table. The link
> being a windowsPath macro. The idea behind this is that users can access
> these documents much easier.

So, you are using XWiki to "annotate" or add "metainformation" to files 
stored in your Windows file system, on that networked hard drive. I 
guess you have created a customized class where you fill in several 
properties for each file. One of that properties stored the mapped path 
for a given file. Provided the box where you are running the client 
accessing XWiki has the correct mapping, users will be able to recover 
each file by entering this path in an explorer window. What you want is 
to have "active links" to this mapped drive.

Concerning this development I think you must be aware of the current 
developments oriented to set an alternative store for attachments in the 
system file. Check devs list for "The future of attachments and 
attachment versioning" and related threads and Jira issues. Once this is 
working, you could have this networked drive set as alternative storage 
location. About having alternative methods to access those files... 
well, we'll see! It is always tricky to access a single repository with 
several, let's call them, agents (XWiki will be the primary agent, that 
storing files and being able to apply all its logic to add huge amounts 
of information and metainformation to each stored file).
> The problem with this system is if a new document is created, 

Do you mean a new file stored in the file system?

Perhaps this initiative could be useful for you by changing stored 
attachments with paths/filenames in the networked drive.

Lool for [xwiki-users] Create application similar to attach file 
application in the users list.

> the user then
> has to manually add it to this table, and if the document's title is
> changed, the link gets broken. So I'm after a way of automatically listing
> the documents and have working links, to make it require as little upkeep as
> possible.

To avoid to get this links broken, you need to reference the files with 
a UID independente from the name. By the end, even though I'm not a 
developer, you will end requiring the complexity needed to develop the 
alternative repository referenced above.

> So far I am able to list the name and the path of all the documents in the
> directory, but I need some help in integrating the sortable table html tags
> and turning the path into a working link, without breaking the groovy macro.
> This is what I have so far:
> ----
> {{groovy}}
> import groovy.io.FileType
> def list = []
> def dir = new File("R:/Quality/FORMS/")
> dir.eachFileRecurse (FileType.FILES) { file ->
>   list << file
> }
> list.each {
>   println it.name
>   println it.path
> }
> {{/groovy}}
> ----
> Any help would be appreciated,
> Thanks,
> -----
> ----
> Lockie

Finally, consider also this Jira issue about UNC paths...


Hope this helps!


Ricardo Rodríguez
Life Sciences, Data Modeling and Information Management Systems

users mailing list

Reply via email to