Translating Assets and Tasks Offline

Idiom Frequently asked questions

Translating Assets and Tasks Offline?

You can work on WorldServer translations directly in WorldServer DVX. A single mouse click in WorldServer opens a project in WorldServer DVX, and a single menu selection in WorldServer DVX uploads a project to WorldServer. There is no need to export and unzip projects, and to zip and import projects.
You can also choose to work on translations offline in Trados TagEditor or SDLX.
This chapter provides an overview of how to do the following:

  • Download, install, and use the functionality available in WorldServer DVX to translate assets or tasks
  • Export assets or tasks from WorldServer to use in Trados TagEditor or SDLX
  • Use Trados TagEditor to translate assets or tasks
  • Use SDLX to translate assets or tasks
  • Import translated assets or tasks back into WorldServer


Using WorldServer DVX
A single mouse click in WorldServer opens a project in WorldServer DVX, and a single menu selection in WorldServer DVX uploads a project to WorldServer. There is no need to export and unzip projects, and to zip and import projects.
Online freeform TM search provides a freeform SQL query facility. This allows WorldServer DVX users who have connectivity to their WorldServer server to query a WorldServer translation memory.
WorldServer DVX allows you to take advantage of WorldServer translation memories and term databases while translating assets or tasks offline. To use the WorldServer DVX, you download it from the WorldServer Home tab and launch WorldServer DVX from Explorer or the Task list.

Installing WorldServer DVX
The WorldServer Home tab provides a link to download WorldServer DVX installer. This download link is located in the Useful Links section.
To download WorldServer DVX, click the link and follow the prompts.
To install WorldServer DVX, double-click the install executable and follow the prompts. When installing, you have three install choices:

  • Typical installs all spell check tools.
  • Custom allows you to select which spell check tools to install.
  • Compact installs without spell check tools.


Creating a WorldServer DVX Project
Before you can begin translating your WorldServer assets or tasks in WorldServer DVX, you need to create a project that contains the following:

  • The assets
  • Translation memory
  • Term database


NOTE: Each time you create a project and translate assets, you will generate several different types of files. You may want to devise a file structure to organize your work.

To create a new project, choose File>New>Project. The project creation wizard steps you through the process, where you specify which content will be in the project, and create or select a new translation memory and a term database.
There are two major sections of the WorldServer DVX main window: the project navigator and the translation window. When you create or open a project, the project navigator appears. The translation window does not appear until you open an asset to translate. The next two sections explain these windows.

About the Project Navigator
The project navigator is the panel at the right side of WorldServer DVX that displays the names of the assets in the current project. Double-click an asset to display its contents in the translation window.

NOTE: If you have closed the project navigator, you can re-open it by choosing View > Project Navigator.

Working with the Translation Window
The translation window displays the source and target segments of the selected asset. The translation window is where you translate the asset.
When you first open an asset in the translation window (by double-clicking it in the project navigator), any 100% matches from WorldServer translation memory automatically populate the appropriate target segments. To make use of fuzzy matches or term database entries, you can use any of the following options:

  • The Options dialog box (Tools>Options) sets some default general behavior for automatic actions and spell checking behavior.
  • Pretranslate populates the target segments with matches it finds in the translation memory and through assembling portions.
  • Translation Lookup searches the translation memory for matches to the current segment.
  • Term Lookup searches the term database for matches to the selected word or phrase.
  • Populate copies the source text to the target segment for any segments that are empty and do not have embedded codes (placeholders).
  • Propagate finds segments that have source text that matches the selected segment and populates the matching target with the target text of the selected segment.

The target segments use color coding to indicate their translation status.

About the Options Dialog Box
The Options dialog box provides general options, including settings for AutoAssemble, AutoSearch, AutoPropagate, and AutoSend as well as spelling options. Choose Tools > Options to open the Options dialog box.

Pretranslating Assets
You can use the pretranslate feature to get the most out of your translation memory matches, term database matches, and lexicon. Depending on what options you choose in the pretranslate dialog box, and how many entries you have in your translation memory and term database, pretranslating can make translating your task or asset easier. Pretranslate can involve assembling a translation from portions.Choose Translation > Pretranslate to access the pretranslate options.

Using Translation Lookup
Translation lookup allows you to search the translation memory for matches. It displays all matches found for the currently selected segment as well as the score for the match.
Click the segment for which you want to find matches and choose Translation > Translation Lookup or press Ctrl+S to search the translation memory for matches to the current segment. If there are any matches, the translation memory match results window appears. Translation lookup is done on a segment-by-segment basis. If you want to search the translation memory for all segments in the file, or in the project, use pretranslate instead.

Using Term Lookup
Term lookup allows you to search the term database for matches to the selected word or phrase. Select the word or phrase for which you want to find term database entries and choose Translation > Term Lookup or press Ctrl+Shift+S. If any entries match the selected word or phrase, the term database lookup results window appears, delineating the matches. Select the entry to see the score (fuzziness) and the date that the term was entered into the term database. You can insert the selected entry or cancel.
Term lookup is done on a segment-by-segment basis. If you want to search the term database for all segments in the file, or in the project, use pretranslate instead.

Using the Populate Feature
Populate copies the source text to the target segment for any segments that are empty and do not have embedded codes (placeholders). Choose Insert > Populate > Current Sentence to populate the currently selected segment. Choose Insert > Populate > Current Language to populate all empty segments in the project.

Using Propagation
Propagate finds segments that have source text that matches the selected segment and populates the matching target with the target text of the selected segment. Choose Translation > Propagate in Current File or Translation > Propagate to All Files to search for and populate matching segments in the current asset or in all assets in the WorldServer DVX project.
If you want to always propagate matching segments, turn on AutoPropagation. When enabled, AutoPropagate fills in the target of any segments that have an identical source to the current source when you press Ctrl+DownArrow (to move to the next segment) or Alt+DownArrow (to move to the next untranslated segment).
Enable AutoPropagate in the Options dialog box (Tools > Options) or by clicking the APr button in the lower right of the Translation Window.

Assembling Translations
Assemble looks for translation memory matches, term database entries, and lexicon matches at the word level. WorldServer DVX then assembles the segment from the translation memory, term database, and lexicon. Target segments that are populated by an assemble process have a blue label.
You can pretranslate with assemble enabled or choose Translation>Assemble.
When AutoAssemble is enabled, WorldServer DVX automatically assembles the next untranslated row when you press Alt+DownArrow.
Enable AutoAssemble in the Options dialog box (Tools > Options) or by clicking AAs in the bottom right corner of the translation window.

Finding and Replacing
You can find words or phrases in the source or target of the asset loaded into the translation window. In addition, you can find and replace words or phrases in the target. Choose Edit > Find or click Ctrl+F.

Changing Case
You can change the capitalization of a segment. The change case feature changes all words in the segment to all caps the first time, then sentence capitalization, then all lowercase. Choose Edit > Change Case or click Shift+F3.

Joining and Splitting Sentences
Joining and splitting sentences in WorldServer DVX is similar to splitting and merging sentences in the Translation Workbench in the browser.
You may want to split a segment to increase the chances of finding translation memory matches or term database entries or to make a large segment easier to translate. Likewise, you can merge segments that are better translated together.
Clicking "Join" (or choosing Edit > Join Sentences) merges the currently selected segment with the segment directly under it. Clicking "Split" (or choosing Edit > Split Sentences) breaks the selected segment where the cursor is currently positioned.

Using the Translation Memory
When you open an asset or task in WorldServer DVX, it contains any translation memory (TM) matches that were found. When you create a WorldServer DVX project and open the asset in the translation window, these WorldServer translation memory matches are inserted into the WorldServer DVX translation memory. 100% WorldServer matches automatically appear in the appropriate target segments. Throughout the translation process, you can search the WorldServer DVX TM for fuzzy matches, and can browse and edit the TM. However, when you import translated assets back into WorldServer, changes or entries that you have made to the offline TM do not get inserted into the WorldServer TM. Only matched segments that are in the actual assets are inserted into the WorldServer TM (only if you select that option upon import).
Maintaining the WorldServer DVX TM will make it easier for you to translate using features such as assemble and AutoSearch. There are several ways to search the TM, send translations to the TM, and to maintain the TM.

Searching the Translation Memory

  • To search for TM matches for a selected segment, choose Translation > Translation Lookup.
  • To search for TM matches for all segments, use Pretranslate (Translate > Pretranslate), and choose the desired options.
  • To search the TM for a particular word or phrase, open the translation memory file (*.dvmdb) and type the word or phrase into the "Locate Sentence" field. You can also type a word or phrase into the "Filter" field and click the Filter button to view only matches that contain that word or field.

Adding, Editing, and Deleting Translation Memory Matches
To add, edit, or delete TM matches, open the translation memory file *.dvmdb), then do one of the following:

  • To add a match, click "Add" and type the word or phrase into the line that appears at the bottom of the source list. Then click "Add" on the target side and type the translation.
  • Edit either the source or the target of the TM match directly in the list.
  • To merge entries, select the terms to merge and click "Merge". Merging of entries takes the translations of the various entries and puts them all together. So if you have an English-German pair and an English-Japanese pair, merging them will give you an English-German+Japanese entry. You should only merge translations with the same source.
  • To remove a match, click on the match and click "Remove".

Sending Translations to the Translation Memory

  • To manually send a selected segment to the TM, choose Translation > Add Pair to Translation Memory (Ctrl+F12).
  • To send segments to the TM as you translate, enable AutoSend in the Options Dialog Box (Tools > Options) or click the ASd button at the bottom of the translation window. After each segment that you complete, press Ctrl+DownArrow to send that segment to the TM.
  • To send all translations to the TM, choose Project > Add to Translation Memory. This sends all source and target pairs in the project to the WorldServer DVX TM.



Using the Terminology Database
When you import WorldServer assets or tasks into WorldServer DVX, applicable terms from the WorldServer term database are inserted into the WorldServer DVX term database. These terms will be useful when using the WorldServer DVX assemble feature as well as the AutoSearch feature. You can also perform term database lookups as you translate your files.
You can add to and edit the term database to make it more useful while working in WorldServer DVX, however, the term database is not imported back into WorldServer with the assets or tasks.

  • To search for term database entries for a selected word or phrase, highlight the word or phrase for which you want to search for a match and choose Translation > Terminology Lookup (Ctrl+Shift+S).
  • To search for term database entries for all segments, use Pretranslate (Translate > Pretranslate), and choose the desired options.
  • To search the term database for a particular word or phrase, open the term database file (*.dvtdb) and type the word or phrase into the "Locate Term" field.
  • To check the translation for correct term database usage, choose Translation > Check Terminology (Ctrl+Shift+F7) to check for term database entries that are not being used in the target.
  • To add, edit, or delete term database entries, open the term database file (*.dvmdb).
  • To add a match, click "Add", type the word or phrase into the "Lemma" field, click "Translation", add the translation for the word, and click "OK".
  • To edit the entry, select it in the list and click "Edit". You edit the source and the target terms separately.
  • To remove an entry, click the entry and click "Remove".
  • To manually send a segment to the term database, click in the segment. Choose Translation > Add Pair to Term Database (Ctrl+F11).
  • To merge term database entries, select the terms to merge and click "Merge". Merging of entries takes the translations of the various entries and puts them all together. So if you have an English-German pair and an English-Japanese pair, merging them will give you an English-German+Japanese entry. You should only merge terms with the same source.
  • To create relations (translation, synonym, antonym) between target words, select the target term, click "Add". In the Add Relation dialog box, select the relation type. Type the term to locate or add. If it is a new term, click "Add" to add the term to the source; if it is an existing term, make sure it is selected in the list. Click "Add". The related term will appear in the source side, depending upon the selection in the "Related Terms" dropdown menu.


Using the Lexicon
The lexicon is a list of all the source language words or phrases in the project. AutoAssemble and AutoSearch use the lexicon to help you translate untranslated segments.
You can build a lexicon to your specifications by using Project > Lexicon > Build. You can also resolve the lexicon to your term database and translation memory. This means that the lexicon will be searched for any TM matches or term database entries. Any matches will be inserted into the lexicon. Choose Project > Lexicon > Resolve with Translation Memory or Project > Lexicon > Resolve with Term Database.
Once the lexicon is built and translated, it will become very useful in AutoAssemble and AutoSearch operations.

Checking Terminology
You can check the translation for term database usage. Choose Translation > Check Terminology (Ctrl+Shift+F7) to check for term database entries that are not being used in the target.

Example: If there is a term database entry that specifies that "pen" should be translated as "stylo" in French, if the word "pen" appears in the source side of a segment, and the word "stylo" does not appear in the target side of that segment, when you check terminology, the word "pen" will turn red, indicating that the term database entry is not being used properly.


When this happens, choose Translation > Term Lookup to see the entry, and make corrections as necessary. This command checks for the first misuse. Choose the command again to see the next instance of a missing term.

Checking Numerals in Text
You can check the numerals in your translation to make sure that the numbers that appear on the source side also appear on the target side. Choose Translation > Check Numerals to check your numerals. You can also automatically check for numerals while checking terminology; this is an option that you can select from the Options dialog box (Tools>Options).

Checking Embedded Codes (Placeholders)
You can check if your embedded codes are correct before exporting a finished translation. You cannot import assets whose placeholders are incorrect into WorldServer.
WorldServer DVX checks to make sure that the placeholders that appear on the source side also appear on the target side. If there are any missing, extra, or incorrect placeholders on the target side, WorldServer DVX marks the segment with a red "x". Make sure you resolve incorrect placeholders before importing back into WorldServer.

Using the Spell-Checking Tool
Choose Tools>Spelling (F7) to check your target spelling. The spell check dictionaries you have available depend on the install option you chose ("Typical" includes all spell check dictionaries, "Custom" includes only those dictionaries that you specify, and "Compact" does not include any dictionaries).
The WorldServer DVX Spell Check tool works the same way as most word processor spell check tools.

NOTE: You can change your spell check options in the Spelling tab of the Options dialog box (Tools > Options).

Using the Word Count Tool
The word count tool provides a detailed account of the words in the document, listing different types of matches separately. To find the word count of a file, choose Tools > Word Count. Choose the files, languages, and row types and then click "Calculate".

Exporting Finished Translations
When you are finished translating your assets or tasks in WorldServer DVX and are ready to upload back into WorldServer, choose File > Export > Upload Project to WorldServer.

NOTE: For the best results, you should have a separate folder for your finished translations than the folder into which the IDZ files were exported from WorldServer. This facilitates troubleshooting problems that may arise.

Exporting Assets and Tasks
To translate assets or tasks offline in Trados or SDLX, you export them in the appropriate format. You export assets from Explorer, and tasks from the task list.

Using Trados TagEditor
When you export tasks or assets to be translated offline in Trados, WorldServer exports the following files into a ZIP file:

  • Idiom.ini
  • IdiomAsset.dtd
  • Idiom.tmx
  • Items index.html
  • metadata
  • translate/*.sgm


The only files that you should open and/or edit are the Items index.html, which contains a list of all the assets or tasks exported, and the SGM files in the "translate" folder, which are the translatable assets or tasks.


To translate assets or tasks in Trados

  • Launch Trados Translator's Workbench.
  • Create a new translation memory by choosing File > New.
  • Import the Idiom.tmx file into the translation memory. Use File > Import and choose files of type TMX. Select the "Check Matching Sub-Languages" at the bottom of the dialog box to ensure that dialects are also included. Open the Idiom.tmx file that was exported.
  • Launch Trados TagEditor.
  • If you are exporting to Trados for the first time, in Tools > DTD settings, choose to open an existing DTD settings file, then select Idiom.ini.
  • Open the translate directory that was exported.
  • Translate each asset.
  • When translation is complete, choose File > Save Target As to save the target into the translate directory. Use the same names that were used for the export.
  • Repackage the files into a ZIP file (the translate folder with the target files and the metadata file are the only files that WorldServer needs; the others will be ignored). You can replace the untranslated files in the original expanded ZIP folder with the newly translated files and re-ZIP the folder for import into WorldServer, or you can create a new folder with the translated files and the metadata file and package this as a ZIP file to import into WorldServer.
  • Import the ZIP file into WorldServer.

The source and target segments from the imported files will populate the translation memory if you select the "Update the translation memory using imported assets" check box upon import.

Using SDLX
When you export tasks or assets to be translated offline in SDLX version 3.5 or version 4.2, WorldServer exports the files into a ZIP file.
After translating your tasks or assets in SDLX, you import the translations back into WorldServer.

Importing Assets or Tasks into WorldServer
You import tasks or assets that have been translated offline in WorldServer DVX or in Trados.

NOTE: Translation memory from offline translation tools is not imported - instead, the source and target segments are aligned and inserted into the translation memory, if that option is selected. You can load an external translation memory by importing a TMX file from Tools > Translation Memories.


Comments