1 – Creation of a MultiTerm Termbase Version 2011 and upwards

  1. Open your MultiTerm Desktop application;
  2. In version 2011, from the tab Termbase, select Create Termbase;
    In version 2014 and upwards, from the tab File, select New > Create Termbase
  3. Enter the name for the termbase and its location, then click Save
  4. The Termbase Wizard opens; click Next
  5. In step 1 of the Wizard, select ‘Load an existing termbase definition file’
  6. Browse for the IATE_SDL*.xdt file delivered in the package and click Open
  7. Back in step 1 of the Wizard, click Next
  8. In Step 2 of the Wizard, fill in a friendly name, and description if you wish, and click Next
  9. In step 3 of the Wizard, click Next
  10. In step 4 and step 5 of the Wizard, click Next
  11. Click Finish to complete the setup of the Termbase

2 – Fill the new termbase

  1. If your new termbase is not still open, open it
  2. Open, in SDL MultiTerm 2011 Desktop, the Catalog view or, in 2014 and upwards, the Termbase management view, click on the Friendly Name defined before, and select Import > Process
  3. Browse for the .xml files in the package and select one, click Open
  4. In step 2 of the Wizard, click Next
  5. In step 3 of the wizard, you are asked for an exclusion file; don’t fill anything in but click Save As, and specify a file name, click Save As
  6. Back in step 3 of the wizard, click Next
  7. You arrive now in Step 7 of the Wizard, click Next
  8. Entries are now imported; there should be 100 000 entries imported (except for the last file in the set, or for the file in the example); import is fast, unfortunately reorganization of the termbase takes ages, even when it has been disabled….

Important: keep an eye on the size of the termbase. As Multiterm is based on the MS Jet engine, the max. file size is 2 GB.
When this size is reached, the file will be corrupted. This may be around importing the third .xml-file. So whenever you come near to that limit, import the remaining entries into a new termbase.

Should there be irregularities, especially when import finishes amazingly fast after it has done only 20% or so, or the Wizard complains about problems, inspect the log file and the exclusion file created in the process. The log file tells how many records have been imported correctly and which line has caused the Fatal error; note the line number with the Fatal error, edit the .xml file in an editor capable of showing line numbers, like NotePad++, remove all lines from line 2 until and including the line with that line number. Then save the edited file, quit the editor and reimport the .xml file. In this way, the offending record will be skipped, and no records will be entered twice. In such a case, please inform me about the error, so I will be able to correct it for future use.

3 – Creation of a Trados Studio Translation Memory

  1.  In Trados Studio 2011, select the Translation Memories View, then select File > New > Translation Memory; in 2014 and upwards, select File > New > Translation Memory;
  2. In the screen New Translation memory – General, fill in at least the Name, select the Location and the source- and target languages, then click Next;
  3. In the screen Fields and Settings, nothing needs to be filled in; click next;
  4. In the screen Language Resources, click Finish, and in the following screen, click Close.

4 – Fill the new memory

  1. If your new memory is not still open, open it;
  2. Select Import… with the right mouse button;
  3. In the screen Import Files, click Add Files and select the file  IATE_SDL_{lang1}_{lang2}_TM.tmx from the package; click Open;
  4. Back in Import Files, click Next;
  5. In the screen TMX Import Options, click Next;
  6. In the screen General Import Options, click Finish
  7. Should there be irregularities, the wizard will tell what line and what character is offending; edit the .tmx file with a suitable text editor and remove all the lines below the <body> line until the line with the offending entry: the first line to be removed is line 15, beginning with <tu>, the last line the offending line or the one below that, ending with </tu>; then reimport the rest of the data. In such a case, please inform me about the error, so I will be able to correct it for future use.