Rich Filesystem

Question

My company downloads numerous files on a daily basis via ftp. Currently these files are stored in a fairly large file hierarchy, which is cumbersome to navigate. In addition the files often have meaningless names making them difficult to identify.

We are considering building an application that presents a rich "filesystem" to the users, whereby a file can be tagged with meta-data and then presented to the user via a GUI. This would allow users to search for files quickly using various criteria (e.g. "Give me all files with today's date.", "Give me all files for account 04F12345."). It would also allow us to store multiple versions of a file and record metrics about who accesses a file and when.

Has anyone had any experience in building such a system, and if so can they recommend a suitable approach?

Better still, Is there an off-the-shelf solution we could use?

Thanks in advance.

Answer 1

Use a database, where each row contains all the data you need and a link to the actual file.

Answer 2

Lots of COTS for that. LiveLink and SharePoint come to my mind right off.

Answer 3

Basically you're looking at a kind of CMS, or digital Asset management system.

I would take a look at something like Alfresco, they have a free and commercial version.

It's likely overkill for what you want, but it should support what you need.

One nice feature of Alfresco is that the data store can be accessed to Java Content Repository API, as well as being mounted as a shared drives on systems that support mounting Windows shares (which means pretty much all of them).

I would start there as a possible implementation, or for ideas of implementing your own if it doesn't fit.

Answer 4

Yes, any document management system should handle this stuff with ease. Not all DMSs have nice tagging features by user but do allow for attribute searches.

All that said, I've seen lots of customized document systems to match business requirements and doc workflows.