Importing a Transit XV TMX into SDL Trados - not all TU are imported

Hello,

I have finally been able to export a Transit XV TM via Transit NXT into TMX so I can import it in SDL Trados Studio 2014. All seems to work well, except less than half of the TU in the TMX are being imported in SDL Trados. I don't know what is causing the problem. I've noted that Trados says "10,000 TU read and 4,000 TU imported". What happened to the other 6,000 translation units? Has anyone ever had the same problem and is there a solution? I really would like to get all TU from Transit into Trados.

 

Many thanks in advance for your help.

Best regards,

Denise Jacobs

  • My next step would be to check what I exported. Are there duplicates? If yes, how many. Studio deletes duplicates when converting TMX files to TMs. I would not be surprised if this might be one possible explanation. And this would, in my opinion not be a problem of Studio.

  • Hi Denise,

    Maybe I can help, I have been migrating tons of tmx from Transit XV via Transit NXT into SDL TMs ;)

    First of all, SDL TMs avoid to upload duplicates. So if your TU occurs more than once in your NXT exported tmx (which is very normal in Transit since the system manages a lot of reference files which can have coincident sentences), SDL will import only one instance. Feel free to contact me if you want more info about this and how to convert and manage Transit reference materials.

    Secondly, not every Transit exported TU is valid for an SDL TM - usually it is due to an unrecognized tag sequence. So those invalid TUs will be rejected and lost unless you save the file containing them. There is an option when importing, see screenshot:

     

     

    By clicking this option, later you will be able to edit the saved tmx and remove all those tags and error messages. Here is an example for a problematic Transit NXT TU:

       <!--Error: TagAnchorNotOpen-->

       <tu changedate="20071226T133123Z" changeid="xxxx">

         <prop type="x-OriginalFormat">Unknown</prop>

         <tuv xml:lang="en-GB">

           <seg><bpt i="1" type="1" x="1" /><bpt i="2" type="2" x="2" />- Um welche Formate handelt es sich bei den Musikdateien?<ept i="1" /><ept i="2" /></seg>

         </tuv>

         <tuv xml:lang="es-ES">

           <seg><bpt i="1" type="1" x="1" /><bpt i="2" type="2" x="2" />- ¿De qué formato de archivos de música se trata?<ept i="1" /><ept i="2" /></seg>

         </tuv>

       </tu>

    You can edit those tmx with Notepad++ and run a macro which simply clean all those tags. Then save the cleaned tmx and re-import it in your SDL TM. You will lose formatting for those TU but contents are preserved so you will get fuzzies for those TUs - I think it is worth.

     

     

    Hope it helps,

     

    Almudena

     

     

     

     

    emoji