Metadata Tools and Related Questions

[anandshankar]

I have some queries regarding file metadata.

Every file on a filesystem automatically has some metadata, filename, uid, gid, Date and Time of Creation, Date and Time of Modification,Date and Time of last access etc etc. Then depending on the application whether its a text file or a media file or an OpenDocument file, there are other metadata which can be seen from inside the application (?). Some of these metadata conform to standards such as the Dublin Core.

Q1: Are there some standalone application(s) using which one can create custom metadata fields for an arbitrary file and some way to assign value to this field from outside the application which created this file? What I am trying to achieve is to :

(i) Assign a DocumentID to the document, either a coded ID or a random number, so that I can use this metadata field value in manipulating the way I store this document in many virtual directories, so that the same document can be seen by various departmental groups according to the functional view and functional hierarchy of that department

(ii) Thinking of attaching a metadata text which is the GPG signature of the author(s), so that the document is seen as a digitally signed document and I do'nt have to look for or manage the .sig file as well along with the main file.

Q2: Is there a filesystem attribute which is unique to every file or directory on that filesystem, in that case I can think of that attribute for my question above.

Any insights deeply appreciated.

anand

[invictus9]

There is a NodeRef value unique to every node in the system; by extension, every document and folder have a NodeRef. Between the NodeRef and the store in which the document is contained, you will be able to find the document.

[swithun]

Every file on a filesystem automatically has some metadata, filename, uid, gid, Date and Time of Creation, Date and Time of Modification,Date and Time of last access etc etc.

A lot of this metadata will get lost when you ingest the documents into Alfresco. So, you would need to extract it prior to ingest. There are tools like FITS, which can extract a lot of technical metadata, using a selection of other tools. A simplistic approach might be to ingest the metadata documents first, then the actual files, and have an automated rule which creates a new aspect for your files and populates the metadata fields with information taken from the corresponding metadata document.