Our translators (using SDL Trados) are complaining about the long file names generated for translation purposes. This is what one of them is writing:

"I mean to bring the attention on the fact that long names and long paths could regularly cause some stopping issues during the creation of translation project. As a matter of fact, we’re always forced to “quarantine” files into local very short directories in order to spare characters and make the system working and kick off projects since our CAT system paths are compounded of filenames and paths themselves hence resulting in very long intermediate structures which might exceed sometimes 260 characters.

I repeat, we have a workaround for that and we share this info with translators which are also invited to store their own working projects into local short folder names, still I wonder if any chance to get shorter source names and structures “by design” in order to avoid those issues upfrontly (in case they were not generated automatically from the CMS, of course). I’m also aware they are “talking” names hence very useful to understand context, so I reckon we would lose some important piece of information by reducing them to shorter alphanumeric strings, and this might return into an undesired drawback."

As you know, the file names are created by adding the crucial GUID, version and language information to the metadata titles we have created for the objects. So yes, unfortunately some of these file names are long. At the same time, the metadata titles contain valuable information for sorting searching and context. Reducing these names to a minimum would greatly reduce the usability of our database.

Is there a way you can “cut out” pieces of our metadata text before you add the GUID, version and language information??