If you use a full-featured CAT Tool to translate your documents, you probably already know that you can search for all instances of a term or phrase such as patria potestad via the concordance function. That way, you do not always have to rely on your memory to know how you have consistently translated a term.
This is not that easy if you have to wade through dozens of Word and PDF documents. If you use Translation Office 3000, and you assign each job a fitting description, with the aid of your memory, you could probably locate most documents and perform a search within each of those documents. But how long would that take? How motivated would you be to do all of that?
There are several desktop search engines such as Copernic Desktop Search and Google Desktop. They include sound, video and image files in addition to text. I have found them to be a bit bloated at times, especially because as a translator, I am mainly interested in searching through text. Then there is dtSearch, which is good if you create your own specialized indexes. I know that several translators use dtSearch. I for one have been testing Archivarius and like it so far. It does not take up an inordinate amount of RAM, and you can define your indexes. Many times I have come across documents that I wasn't looking for, but that are nonetheless useful!