Welcome to a new tooltip about Transit NXT. In this tooltip we will see how to integrate data coming from a source other than Transit NXT into your translation memory and the role that exchange standards play in that. Let's see a simple case. Imagine a client has asked you to translate the new version (v. 2.0) of a document that was already translated by someone else into your target language. You will be able to obtain perfect (or very similar) matches or to pretranslate all old sentences (i.e. all sentences already present identically in the old version) if you have access to that old material and, for obvious reasons, it is in the interest of the client himself to provide you with it. For that, the best option is that the client sends to you a TMX file with the translation of version 1.0 that they got from their previous vendor.

The purpose of the Translation Memory eXchange (TMX) international standard format, developed by LISA’s OSCAR workgroup, is to exchange translation data with other users. This is a useful functionality that all important CAT tools include nowadays. It allows you to use translation memories regardless of the format in which they were originally created (Trados, Across, DéjàVu, Wordfast, MemoQ, etc. ), which means you may use Transit NXT and work with clients that use another assisted-translation system, or use public memories which are available in this format (e.g. DGT, UN Corpora, etc., about which we will have a word in the near future, stay tuned!). The first time you do this, you might be tempted to try to add the TMX as reference material. If you do that, there will be no perfect matches, because TMX is just an exchange format, not one that Transit NXT can use natively (as it might be the case with other tools, such as OmegaT). Instead, you have to import the TMX file so as to obtain a language pair out of it or, in other words, to convert it to a format that Transit NXT can understand. Let's see what you need to do, then, to convert a TMX file sent by a partner or client into a language pair. Just go to the Reference material menu in the resources bar, then to TMX interface, and finally to Import TMX.

Then you will need to provide the path to the input TMX file, and the path and name to the output language pair (which will be generated). You can use the actual name of the TMX file (including the .tmx extension) for the name of the output language pair files. Transit NXT will automatically assign the correct extensions according to each particular language.

You will find the generated language pairs in the output folder defined.

Once you have the language pairs, in order to use them just copy them to the reference folder you added to your project, or add the folder containing them as a reference material folder in the project settings, or you can add the language pairs directly too. And that's all. In a future tooltip we will see how to perform the inverse operation, that is, convert one or more language pairs to a TMX file, which you could send to a client or share with other people using different tools (as long as they have this standard implemented). Thanks for reading, and please do not hesitate to send your comments or questions or to ask for specific tooltips.