Help neded with data driven category

[Carlo Didier]

I'm trying to make a data-driven category as follows: Show a sub-category for each folder under a specific path.
Why? I have copies of images which I have published online in subfolders like this:
P:\published
   +-- Facebook
      +-- Album xyz
   +-- Smugmug
   +-- ...
and I would like to have the images automatically assigned to categories depending on which folder they are in.
I can of course do this with an event script, but I thought it should be possible with a data driven category. I tried what you see in the screenshots, but it didn't give me the result I was looking for. I tried various other things, but I just can't get what I want. I'm nearly there, but I can't get the next level of folders as subcategories (I got the second level only because the name of the first is a fixed length).



[attachment deleted by admin]

[JohnZeman]

Hi Carlo,

I have read your post several times but I'm just not grasping the concept of what it is you're actually trying to do.

Can you rephrase your question differently?

[Carlo Didier]

Of course, John.
When I create jpeg version of images to post online, I save those copies to specific subfolders under a folder P:\published.
I would like to have a data-driven category that gives me a main category "published" with subcategories for each subfolder tree.

Example Facebook:
P:\published\facebook
P:\published\facebook\Album Bruges
P:\published\facebook\Profile pictures
P:\published\facebook\Cover photos
etc

From that, I want a top-level category "published" and under that one sub-category containing everything under "facebook" and all sub-folders.

I know, I could simply create those categories as static, but I thought it should be possible to do it with a data-driven category (which would automatically add/remove subfolders or folder renames).

[monochrome]

I'm going to go out on a limb here and declare what you are trying to do to be impossible.

The reasons for this being impossible are:

• There is no way to create a data-driven category from a subset of images - you can't have the result of a category formula feeding the DD category, for example - so there's no way to just filter out the "published" images before creating the DD category hierarchy.
• The filter expression is applied after the hierarchical name has been split into path elements, and applied on each path element individually - so there's no way to just filter out the "published" images after creating the DD category hierarchy.

I think what you have now, a flat list of all directories under "published" is the best you can get. Use the folder tag. Turn off "Detect hierarchy". Add a filter that filters on "Published".

The real reason you're running into this is that metadata is supposed to be encoded in the image file itself (or sidecar), and not in its place in the directory hierarchy. If you want to track where images were published, perhaps the IPTC field "Job Identifier" is suitable? "This ID is typically added by the creator or image provider for transmission and routing purposes and related to the job for which the image is supplied" -- http://www.iptc.org/std/photometadata/documentation/GenericGuidelines/Documents/iptccorestatussection.htm

You could then create a DD category based on this tag.

[Mario]

Using Metadata as the 'state' indicator would also be my preference.

XMP labels have been created to carry the 'state' of an image, and you even have automatic visual feedback for which images new, edited, for review, published or whatever. Labels can hold only one 'state', which is usually sufficient for the professional workflows.

If your files can be in more than one state at a time (e.g., published to several sites), use a metadata tag of your choice to track this. E.g. separating multiple sites by ; in the Job Id or another tag.

When you have the information in the metadata and not buried somehow in folder names, you can easily access it in data-driven categories. And you can use the Renamer to automatically copy/move your files into folders corresponding to to their state, by using the Job Id to create folders as needed and then copy/move the file.

[Carlo Didier]

Thanks for your replies, everyone.
What you propose is probably the logical procedure in a professional workflow, but I'm not a professional and saving the published versions (actually before they are published) to specific folders is for me much, much simpler.
I'll just manually create the corresponding categories by directly referencing those folders.
You may think it a disadvantage to have multiple copies of those images in these folders, but those are just lower resolution copies of my originals (so they are automatically identified as versions) and I may use different pixel sizes, renderings (added border or not, adding copyright watermark or not) depending on the site where it is published. Only these versions shall be assigned the corresponding published categories, not the originals.

[monochrome]

What you propose is probably the logical procedure in a professional workflow, but I'm not a professional and saving the published versions (actually before they are published) to specific folders is for me much, much simpler.

Ok.

Just a final note: If you use the IMatch batch processor to create the published versions you can have it set the metadata as part of the batch process.

[ubacher]

I don't quite get why you like to have categories for these published images. They are easy to find when you need them
because of your folder structure. And easy to use in filters. Just put a bookmark to the top folder (published)
into your M&F panel for easy access.

Do you have a scenario where you can't work with these images (filter/find/select) unless they have categories
assigned to them?

[Carlo Didier]

I don't quite get why you like to have categories for these published images. They are easy to find when you need them
because of your folder structure. And easy to use in filters. Just put a bookmark to the top folder (published)
into your M&F panel for easy access.

Do you have a scenario where you can't work with these images (filter/find/select) unless they have categories
assigned to them?

The simple answer is that I want to do everything with categories because then its most simple to use searches, filters, the builder, etc if I only have to cope with categories. It's just simpler.
There are just more possibilities with categories (like color coding for example).
Why use multiple techniques to organize your images if you can do it all with just one. Keep it simple.

[sinus]

Carlo,

if you would do it, like for example Mario and monochrome suggested, you would have exactly, what you want.

The only thing is, would would have an additional step, to bring the category into one metadata-field.

What would be  for you an additional "security", because you would have the category in the image itself. Not necessary of course, but does also not harm.

I understand, that you like to work in the cat-view (me too), but I think, you can get, what you want, but maybe not on the direct way to Rome, maybe you have to take a second way over Milano. 

[Carlo Didier]

... because you would have the category in the image itself

which is exactly what I want to avoid ...
apart from the first ingestion and basic post-processing, the DNGs aren't touched anymore. Though I never had any problems, the less often files are written, the less chances of something going wrong. And if a file isn't modified, it does not need to be re-backuped.

[sinus]

... because you would have the category in the image itself

which is exactly what I want to avoid ...

I see. So it will not easy, then ... Formula-categories maybe?

[Carlo Didier]

... Formula-categories maybe?

That's what I've done now (using @RFolder)



[attachment deleted by admin]