A simple way to work with complex text formats
Using our computer-assisted translation tool (or, in short, CAT tool) is such a normal part of our everyday work that we almost take this software for granted. The translation memory saves every sentence we translate and then offers them as suggestions when it recognises similar text in subsequent projects. Furthermore, the tool has a terminology database, in which we store and consult specialist terms and their translations in different languages. This system saves both us and our clients valuable time and resources, and also ensures lasting consistency. But what happens when the format of the text that needs to be translated is so technically complex that even a professional CAT tool is at a loss to know what to do with it? Then we relish the challenge! Just like we did with our client Viacar AG, who we “brainwashed” our software for.
Viacar AG is based in Aarau and supports Swiss driver and vehicle licensing offices with its modern integrated IT solution that encompasses all the important processes, such as vehicle registrations, changes of address or ordering driving licences. The software is an efficient tool that the public can use to take advantage of modern and convenient e-government services.
So far, we have translated approximately 1,000 so-called test cases for Viacar’s software. These test cases contain instructions, error messages, expected values and information in the broadest sense in order to check Viacar’s software. We receive the texts as Excel documents. The text boxes in these documents are embedded in long chains of HTML programming code that is not only not to be translated, but should also not be changed under any circumstances. It looks something like this:
Simple, right? The text that is to be translated appears in each Excel cell twice: the first version should remain in the source language, while the second should be overwritten with the translation in the end.
To ensure our translators can work effectively (and fast!) and our client’s translation memory stays up to date, we work on these complex texts with the help of our CAT tool, as with all our other translations. But what happens to all that programming code in the text? The tool cannot make head nor tail of it: the entire text gets all mashed up in the program and turns into an illegible chaos. Where is the translator meant to start? Or the accounting department? How are you expected to weed through this alphabet jungle and accurately calculate how much of the text is actually part of the translation order? Of course, we do not want to include the code in the calculation, as we will not be working on that.
We called on our CAT tool specialist and his programming expertise. Our expert has many years’ experience with CAT tools and knows every single secret trick to our program. For this project type, he made the tool undergo a sort of brainwashing, with lots of testing, programming, extracting and fine-tuning. And our expert was successful: the system has now been programmed to “understand” which text passages are to be translated for Viacar’s projects. The CAT tool automatically recognises the programming code that appears between the text and blanks it out. The translators simply have to apply a few special settings to the document in the CAT tool and they can get to work as usual. Finally, when the translator exports the finished text into an Excel document, the program once again includes the programming code, as if by magic.
The bottom line: Viacar AG offers its clients a personalised tool for e-government services, while we, for our part, support Viacar in creating this effective tool with the help of our own customised translation program.