Retrieving Multiple Assets via webservice

Hi,

This really is a simple question about process implementation for retrieving multiple assets via the MB webservices.

The basic steps I am going to implement are:

1) Perform a search to retrieve a list of multiple asset ids (I hear the search functionality has been improved somewhat with the addition of the quicksearch)
2) Retrieve the assets
3) Zip them up and give link to user

My assumption is that I have to call the webservice multiple times (ms.RetrieveFile(...)) for each asset as I cannot see a "batch" call.

However, I am a little unclear on how the job process works, in that, whilst the operation is effectively one job, I believe the process works by creating a job for each retrieval.

The problem to me with multiple jobs is that the MB Server does not know how many images to process, so the completion state is client side based on whether the current job succeeds, and when all images in the array have been processed.

From what I have seen of previous posts, the zipping is for me to implement using XP shell extensions rather than done on the MB server.

If someone can confirm my suspicions, or suggest the best way for multiple file retrieval, then I would be most appreciative.

I currently use MediaBin 4.5.3 and C# to implement the webservices

Kind Regards

Sky Rumsey

Find more posts tagged with

Comments

lyman

I can confirm that by the time jobs reach the server they are treated one at a time. The server does not have a concept of a batch of jobs.

The web service does not have a call to aggregate multiple jobs as well.

Cheers,

Lyman Hurd

sample-works.rptdesign

sample-fails.rptdesign

Sky_Rumsey

Thanks for the prompt reply, I will continue as planned in my code.

Forgive me for my staggered posts regarding the same topic, however I just need confirmation in that becuase they are unique jobs, then the call to the Retrieve*_URL files, or anything that creates a temporary folder in the Transfer directory means, it's one temporary folder per image. Is this correct?

If not, can someome please advise on a way to have the images in the same directory (mainly for compression purposes) as I cannot produce this behaviour. I noticed that the MB Web Client puts multiple image downloads in one transfer directory, which I assume it's because it's referencing the DLL directly?

Finally, regarding the new quick search, 1) is it exposed via the webservice in 4.5.3 as I can't seem to find it and 2) is there any documentation as my colleague Jason Copsey has discovered that you can do operator searchs e.g AND, OR?

Regards

Sky Rumsey
Software Developer
Brandbank

aggregate_error.txt

lyman

I have asked our web services expert to step in on this one! I expect a reply soon.

Cheers,

Lyman Hurd

msnider

When using any of the retrieval or insertion methods that generate a temporary file, the files may or may not exist in the same temporary directory...if the temp file would create a filename collision then a new temp directory is created, otherwise it will be written to the same directory. If you need to create them all within the same folder, use the UNC retrieval methods so that you can control the output filename.

Yes, QuickSearch is available in 4.5.3. Use the web method IsQuickSearchAvailable() to programmatically determine if QuickSearch has been enabled on the MediaBin server. If it has been enabled then the QuickSearch metadata ( id = {F24FC776-00D2-4ED7-ADA3-9D5288B149A3})will be returned within the searchable metadata list retrieved from GetSearchableMetadata().

Mark