Welcome to a new tooltip about Transit NXT. Today we’ll deal with pretty basic stuff, in fact one of the most basic operations one must know to use this tool (or any CAT tool) productively and one that makes computer-assisted translation tools what they are. Their basic principle is don’t reinvent the wheel, recycle instead. You do this thanks to the translation memory functionality.
In Transit NXT, a project’s translation memory (or TM) is commonly referred to as the reference material that you assign to it. What is this reference material? Let’s remember that Transit NXT internal working format is the language pair, that is, a bi-text, or two parallel documents (one in the source language, and the other in the target language). This means that when you translate you’re always working with two text files. When the translation is finished, these two text files can be added as a source of fuzzy matches to future projects, and this is what we call the reference material. Therefore, the reference segments that come up as fuzzy matches are stored in text files too.
In fact, there’s no difference between the language pairs in the project that you have to translate and the language pairs that you include in your project’s translation memory, except the status of their segments (the latter is already translated and confirmed, the former not yet, or not completely). This is not a trivial issue — under a normal configuration your reference segments will only come up as fuzzy matches if their status is Translated or higher.
Moreover, a project’s active translation memory is composed of the segments stored in the reference material plus the segments which have already been confirmed as translated in the working files — namely, when you translate and confirm a segment, it becomes a reference segment that will pop up as an exact or fuzzy match to help you translate identical or similar future segments.
Let’s see how to add reference material to your project. Open your project Settings.
Then go to the Reference material tab.
You will see three buttons: Add projects, Add folder and Add files. Let us focus now on the last two. Adding files will let you choose the specific language pairs that you want to use, whereas adding folders would allow you to select a folder that contains them. In this case, let’s choose a language pair (second option). You will only see files in the project’s source language, which you should have previously defined. That is fine, select the ones you want added:
After doing that, the selected file or files will be added to your project and all the segments they contain will be available as exact or fuzzy matches when you come across something identical or similar enough. That’s it!
Adding a folder instead of a file could be handier in certain cases, as you could add and remove files from that folder and your reference files would change accordingly without needing to modify your project settings. Just a tip ;)
As we hinted earlier, when you are done with your translation and all the segments in the target language file are confirmed with status Translated, you have a new language pair that you can use as reference material adding it to future projects. Remember that all working segments, as long and as soon as they are confirmed as translated, become also part of the active translation memory that the project is resorting to (meaning that the active translation memory of the project is composed of all the segments with status Translated, whatever their source file — working file or reference file).
This is all. This can be a bit complex for people who are not used to Transit NXT’s genuine architecture, I hope I have put it plainly enough. Thank you for reading. Please do not hesitate to send your comments and remember that you can send questions or ask for specific tooltips.