There are a number of special things about the way TTX (Trados TagEditor) files are handled inside Déjà Vu.

1. Preparing the TTX files

The TTX files should be pre-segmented by Trados Workbench by the tool used to generate them.

To pre-segment the file, you need to do the following:

1. Open SDL Trados Translator's Workbench.
2. Create an empty translation memory by accessing the menu option File>New.
3. Click Create, and give a name to the Translation Memory.
4. Access the menu option Tools>Translate.
5. In the Translate Files dialog box, enable the Segment unknown sentences check box.
6. Add the files you want to segment by clicking Add...
7. Click Translate.

2. Importing TTX files into Déjà Vu

Add the files to the project in the usual way (i.e. with the Project Wizard or with the Project Explorer). Just before they are imported, you may be presented with a window asking you to confirm which languages are present in the file(s). It will show the language abbreviation (as present in the TTX file) in the left column and give you an option of changing the language that Déjà Vu treats it as.

If, after importing the file(s), you see that they are empty and don't contain any rows, then you should go back to step 1. and make sure that you (or your client) selected "segment unknown sentences". If the file imported successfully, then you should see the source sentences along with some of their translations.

If there are no useful translations in the original TTX file (because e.g. it was segmented with an empty Trados memory), then you can delete all existing translations in the file by right-clicking in the project area and selecting "Clear all Translations".

You may also be presented with orange colored rows, which indicate XTranslated segments.

3. Tags inside TTX projects

When importing a TTX file, Déjà Vu keeps all existing formatting information from the original.

Since Trados and other tools do not have the same concept of tags as Déjà Vu does, this may result in difficulties.

If the formatting of a target sentence is different from the source formatting, then the tags in that sentence pair will not match.

In contrast to what happens with all other file formats imported into Déjà Vu, you don't need to correct these code mismatches if they are present right after you imported the TTX file.

You can safely ignore these particular code mismatches. However, when you are manually translating a given sentence, you should still try and place the source codes at the appropriate places in the target sentence. Please note that if you are translating between languages with different scripts, you should refer to step 4.

4. Handling of different character sets

This applies to people who translate between languages that do not use the same script.

This happens for example if you are translating from English (Latin script) to Russian (Cyrillic script) or Far East languages (CJK). TTX files contain information about which font should be used to display a given sentence or part of a sentence. Many times, people use a different font for different scripts, so this font information will not be the same in the source and the target sentences that are imported. This will result in many segments with mismatched tags, but there is a way to eliminate this effect. Déjà Vu can ignore the font information on import and then re-assign a font with the correct script when it exports the file. Note that this will still import font attributes such as bold, italic and size as tags. If you would like to use this feature, click on the version of Déjà Vu you are using on the table below to see the steps you need to follow:

Déjà Vu X	Déjà Vu X2	Déjà Vu X3
Download this file. Double-click on it and confirm when Windows asks you whether you would like to merge the contents with your registry. If you would like to turn this feature off again, please follow these steps: Download this file. Double-click on it and confirm when Windows asks you whether you would like to merge the contents with your registry. Download this file. Double-click on it and confirm when Windows asks you whether you would like to merge the contents with your registry. If you would like to turn this feature off again, please follow these steps: Download this file. Double-click on it and confirm when Windows asks you whether you would like to merge the contents with your registry. Download this file. Double-click on it and confirm when Windows asks you whether you would like to merge the contents with your registry. If you would like to turn this feature off again, please follow these steps: Download this file. Double-click on it and confirm when Windows asks you whether you would like to merge the contents with your registry.

Please note that if you are sharing TTX projects between different users and translate across different scripts, both the person who imports the file and the person who exports the file should have the same settings.

Example:

When translating a TTX file from English to Russian, the following sentence appears:

The project was set up with an empty Trados memory, so when you import the TTX file, you will get the following:

Source (English)	Target (Russian)
{1}This is a great restaurant{2}	{3}This is a great restaurant{2}

The source begins with the tag {1} while the target begins with the tag {3}. In this case, this is entirely due to the font information stored in the TTX file. If you highlight the code {1} in the source segment and press Ctrl+F6 to view its contents, you should see something like:

<df Font="Arial">

Now the code {3} in the target is similar, but different:

<df Font="Arial Cyrillic">

If the segment is translated manually, the Cyrillic font should be retained in the target sentence. This is done by leaving the tag {3} at the beginning of the target sentence. A translation could look like:

Source (English)	Target (Russian)
{1}This is a great restaurant{2}	{3}Это будет большим трактиром{2}

If you choose to ignore the font information with the registry file above, there will be no tags for font definitions to worry about. The same segment will be imported in the following way:

Source (English)	Target (Russian)
This is a great restaurant	This is a great restaurant

Déjà Vu will automatically reinsert the font "Arial Cyrillic" around the target segment, so that non-Unicode programs can properly handle the resulting file.