xCP accelerator: import data

Dear all,

I'm not an xCP expert: I'm just starting studying xCP to understand if it is useful to my scope.

I need to import data in an EMC Documentum repository. I would like to automate the process.

Example:

from a user interface I select a file (containing the metadata) and a folder (with the content files), and the system import the data and the metadata in the reposiotry.

Do you know if there are some available xCP accelerators?

xCP is suitable to this scope or I have to change technology?

Many thanks,

bulbapeppe

Find more posts tagged with

Note:

If there are Accepted Answers, those will be shown by default. You can switch to 'All Replies' by selecting the tab below.

Accepted answers

aflowers001

Have you considerdda custom UI on the client, such as a .NET (or Java) desktop application? This would do what you're after.

The problem you will likely come across with web based solutions is permissions. If you upload the CSV and then expect the web site to go back to the client and retrieve a set of files you will come across issues, which is why submitting files to a server is a 'person' activity, i.e. a person has to select the files to upload and then request that the action be carried out.

All comments

aflowers001

Importing files is out of the box for Webtop and for TaskSpace.

When you say 'process' do you have a business process that needs to be executed around the imported content? If so then using TaskSpace and forms as the front end to the process would be the way to go.

If you are just importing content into known locations, or having the BOF layer determine target locations based on metadata, the Webtop could well suffice.

bulbapeppe

Yes, I have to import data in a known location.

So you suggest webtop: I guess I have to use the BOF to customize the webtop and import data (from folder for example) and metadata (for examplefrom a structured file). Am I right?

Many thanks

aflowers001

Ah, sounds like you have a different requirement.

What I thought you weer after was the standard import functionality which can handle multiple input files and allows assignment of metadata on a file by file basis when importing.

What I believe you wish to do is import a number of files AND a metadata definition (as you mention).

This could be done in Webtop via a UI customisation to add a new 'component' that understands what it is your are doing, you can inherit from the Import component to achieve this.

Alternatively you can use TaskSpace to do this. A form that can take the input files and separate content import for the driver file that then passed into a small process that did the filing could also work.

Unfortunately what you are after doing is a little more than a simple import, and needs a bit more smarts in the client UI to know that you have content files and a driving metadata file. BOF TBO won't help with this but you could still use a BOF SBO to package the logic you need.

Perhaps if you gave more details about what you are trying to achieve better alternatives may pop up.

bulbapeppe

Hi,

many thanks for your answer.

I try,as you suggests, to explain better what I'm going to do:

suppose I have a predefined structured set of metadata and correspondent data. I mean, I have in a cvs file all the metadata regarding a set of documents and the correspondent content files.

What I want is to have a UI that reads the cvs, gets the metadata, reads the correspondent content file and then stores in a repository (in a known location) this files (metadata+content file). All automatically.

I hope I have been clear.

Bye

aflowers001

Have you considerdda custom UI on the client, such as a .NET (or Java) desktop application? This would do what you're after.

bulbapeppe

Ok. It is clear. Yes, I can create a desktop UI, so to avoid any permission problem.

Now I'm looking for some existing classes, but it seems not to be so hard.

Many thanks for your answers.

aflowers001

If .NET then look at DFS if you want to 'official' web service way, or you can still use the PIA for a while.

For Java you can just use raw DFC, although DFS may be better.

Or you could use a custom UI, custom web service and have the web service be the one that knows about Documentum.

The choice is all yours now...

lastnitescurry

listen to Johnny

http://johnnygee.wordpress.com/2007/04/06/you-installed-documentum-now-what/

what you are trying to do is a bulk load... Have you tried googling "Bulk load documentum"? there are free tools out there also

http://www.google.ie/search?q=bulk+load+documentum