Hi,
We have a Dedalo installation in which we are using different sections to store images.

We want to unify all these sections into one and start using the generic Dedalo image section (rsc170).

Furthermore, this information comes from a Dedalo 5 that we are taking advantage of to migrate to Dedalo 6. This way in Dedalo 6 all the images will remain within the same section.

Later we will do the same with the rest of the media (av, pdfs,...) that we also have divided into different sections.

To do this, we have copied the images from the /media/images directory that we had classified in different subfolders (catalogo, diapositiva, digital,...) within the 1.5MB, original, modified and thumb folders, depending on their quality, and we have renamed them to rsc29_rsc170_x.

Once this is done and the image field cache (rsc29) has been updated to synchronize the disk data with the database data, the images are displayed correctly in the section, both in the list and within the tab, so far so good.

The problem comes with some images that do not have a 1.5MB or thumb version, when trying to generate it with the Generate Version option of the media versions tool it gives us the error Error: Unable locate source_file. File does not exist when the file does exist and has the correct permissions.

Looking a little at the Dedalo code, it seems that to generate the different versions, some original_normalized_name and modified_normalized_name variables are used that are filled in and saved when the media is uploaded to Dedalo and I understand that, having directly copied the files to the server, they do not exist.

If this is so, would there be a way to solve it so that I can use the generate version tool with media copied directly to the server.

I can think of some options, I don't know if they would be feasible or if they are unfeasible due to having side effects in other parts of the code that I don't know about:

When updating the cache of the rsc29 field, if these values (original_normalized_name and modified_normalized_name) are empty, fill them if the original and modified qualities exist.
When using the version generation tool, if original_normalized_name and modified_normalized_name are empty, try to obtain the name from the file_name attribute of files_info with original or modified quality.

Maybe you can come up with a more suitable solution, or this is really not the reason and you can shed some light on the problem.

We have not tested whether the same thing happens for the rest of the media, but it would be good if it could be solved for all cases.