Recognizing the format of a file selected by user (DFS)

In DFS, the FileContent class can be used to update the content of a document. The class has a LocalPath property that should hold the path of the local file, and a Format property that should hold the format of the file. In my implementation, I let the user choose a local file and upload it to the Documentum repository. The problem is I can't always say what's the format of the file based on its extension. For instance, Documentum has a jpeg format rather than jpg, so using the more common extension of jpg will fail. Similarly, a Word 2000 file with a doc extension will have a msw8 format.

I'd rather let Documentum figure out the type based on the name and content, if that's possible. If I leave the Format field black, the file format is just unknown. I can hard-code some extension-to-format conversion, but that'll be disgusting. Is there a way to move the task of format recognition to Documentum?

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

Hi,

You could cache the information documentum uses to map file extensions for formats using the query service, using a query such as

select name from dm_format where UPPER(dos_extension)='<upper case extension>'

If > 1 record is returned then they will, in theory, be in order of best match to worst, but that isn't always the case for certain formats such as .doc which has several formats for this extension.

or you could look to write a custom service that wraps up an implementation of the IDfFormatRecognizer interface & implementation, which boils down to wrappers around queries of dm_format

or you could get ugly and take the list of formats and work out what best suits your needs by 'hard coding' it into the application.

All comments

aflowers001

Hi,

You could cache the information documentum uses to map file extensions for formats using the query service, using a query such as

select name from dm_format where UPPER(dos_extension)='<upper case extension>'

If > 1 record is returned then they will, in theory, be in order of best match to worst, but that isn't always the case for certain formats such as .doc which has several formats for this extension.

or you could look to write a custom service that wraps up an implementation of the IDfFormatRecognizer interface & implementation, which boils down to wrappers around queries of dm_format

or you could get ugly and take the list of formats and work out what best suits your needs by 'hard coding' it into the application.

erangi

Thanks, Andy.

The chosen solution will be hybrid: reduce the hard-coding ugliness to only a few very common extensions such as doc, and rely on dm_format for the rest. Problem solved.

nitinsengar

Hi Erangi,

If you are looking for the best solution for your problem then you should register all the new format which is not understand by Documentum in the content server so that when you upload any document in the content repository, Documentum display the right format type for the document.

Following the steps to register new file formats.

1> Login into the Documentum Administrator

2> Login with the system admin username and password where the Documentum is installed.

3> Select the format option-> File Menu-> New -> Format

Defiantly this will solve your problem.

Thanks,

Nitin Sengar

erangi

Nitin, my problem is not with special file formats, but rather with the common formats that are already registered. For example, the .doc extension is associated with 7 different formats in dm_format. I have no way of telling which is the right one for the file the user has selected. I was hoping the content server, with all its built-in content management expertise, would have the ability to recognize the format of a given file. Sadly, it doesn't, which means the format recognition will have to be done on my side.