tmx imports takes too much time
Thread poster: Worder Company

Worder Company
Russian Federation
Local time: 19:21
Russian to English
+ ...
Apr 17, 2014

I need to import a large number of tmx files into the TM in trados 2014. I cannot import all tmx files at once, as I need to type in a particular value for each tmx file within the import preparation procedure.

Sometimes, it takes up to one minute for trados to read 50 segments.

The tmx files have been prepared in memoq, since it looks like it takes less time to prepare a tmx file in memo versus the trados procedure.

How can make it quicker for trados to import tmx files into the TM.

I have done "recompute fuzzy index statistics" and "re-index translation memory", but it did not help.

I have the runner at the bottom to the left-most position for accurcy-speed. Should I take it more to the right?

I have asked the trados reps in Russia, how the changes of this runner are going to impact the qulity of the TM. The answer was that it is diffuclt to identify the outcome, so I'd better not change the positon.


 

SDL Community  Identity Verified
United Kingdom
Local time: 18:21
English
Depends on the content of the TMX Apr 17, 2014

If the TMX is full of tags, contains duplicates, filled with placeables etc. then it could indeed take a long time as Studio will do what it can to remove content it does not need.

When you say "prepare"... what do you mean by this?

Regards

Paul


 

Worder Company
Russian Federation
Local time: 19:21
Russian to English
+ ...
TOPIC STARTER
Preparation Procedure Apr 18, 2014

SDL Support wrote:

If the TMX is full of tags, contains duplicates, filled with placeables etc. then it could indeed take a long time as Studio will do what it can to remove content it does not need.

When you say "prepare"... what do you mean by this?

Regards

Paul


I have a number of files, 20 files, each of which contains 50 pages. These documents are PDFs with a lot of graphics, tables and schematics. I would like to prepare the "pseudo-memory", i.e. I would like to do the following:

1. create a list of folders,
2. name them, 1, 2, 3 ...
3, place an original file and a translation of a file in folder 1
4. open file 1 in trados (click "copy source to target if no match is found" before uploading the file to trados).
5. open file 1 in editor
6. select all segments
7. mark them as "approved"
8. update main translation memory, typing "1" in the field setting (you need to create a TM with the following field setting: column 1: something like "Register Number", colum 2: Number).

Result of the procedure:

1. open a file to be translated in editor. If you have a match in your file versus file 1, you TM results window will display this match and to the right you will see "1" in your TM results window. When you see a match, open folder 1 and copy/paste the needed sentences in your translation.

This approach should be used for a particular customer in case, when you expect a lot of matches between the files translated and files translated earlier, provided that 1) you have pdfs that are very difficult to align (otherwise I would just align them in memo), 2) you expect a lot of matches or need to develop from scract an extensive background termbase for a customer.

---

Preparation in Memo:

I would just upload all files at once in a new project. I would create a TM named "1", then open file 1 in the editor, copy source to target across the file, approve them, export tm in tmx . Then I would open trados and import tmx 1 into the TM, typing '1" in the last screen of the tmx import procedure.

P.S.

I have successfully imported large numbers of files earlier as per procedure "memo prep + tmx import with field setting used into trados TM", but this time for some reason it takes a very long time.

It's actually faster to do the update memory task as per above versus memo-created import into trados TM. But, the problem is that I would need to do a lot of actions, which is tedious, but when I upload a tmx into trados TM, I would just open two windows of trados 2014. I would place one of them so that I can see the runner at the bottom of my screen and when it's gone I would do the next tmx import.

So, my question is:

How does "accuracy-speed" runner work? Can I just go to the middle and expect satisfactory results. I mean, I have 1K pages and do not want to manually check the results of the runner changes, as, presumably, the outcome for one of tmxs may significantly differ from the outcome for the tmx runner changes...


 

CCW  Identity Verified
English to Spanish
+ ...
A very bad solution from SDL Nov 8, 2015

I have also experienced this same problem. I really do not understand why you had to design Studio like this. Why not just add an option to allow a "straight import", whithout letting Studio try to fix or remove anything, just import all of the TM without questioning anything. I have now spent more than 1.5 hours trying to import a memory of about 180,000 TUs, and I still have a lot left. This is simply not practical. Many clients keep huge TMs whit lots of repetitions and tags which you must use in each translation you do for them, no matter how small. Spending several hours in each TM import is not cost-effective. It would really be great if SDL could work out a solution to this problem. I upgraded to Studio 2014 some time ago in the hope that import of huge TMX files would be faster, but there is no difference at all compared to Studio 2009. I actually find no real improvements in the 2014 version - just a few cosmetic changes. I have worked with SDL and used their products for more than 15 years now, but I feel disappointed of the Trados Studio line of products - they have too many flaws. SDLX was faster and more reliable, but I cannot use it because almost no clients use this software.

Just my humble opinion.

Regards
ccw

SDL Community wrote:

If the TMX is full of tags, contains duplicates, filled with placeables etc. then it could indeed take a long time as Studio will do what it can to remove content it does not need.

When you say "prepare"... what do you mean by this?

Regards

Paul


Luca Tutino
 


To report site rules violations or get help, contact a site moderator:


You can also contact site staff by submitting a support request »

tmx imports takes too much time

Advanced search







PerfectIt consistency checker
Faster Checking, Greater Accuracy

PerfectIt helps deliver error-free documents. It improves consistency, ensures quality and helps to enforce style guides. It’s a powerful tool for pro users, and comes with the assurance of a 30-day money back guarantee.

More info »
SDL MultiTerm 2019
Guarantee a unified, consistent and high-quality translation with terminology software by the industry leaders.

SDL MultiTerm 2019 allows translators to create one central location to store and manage multilingual terminology, and with SDL MultiTerm Extract 2019 you can automatically create term lists from your existing documentation to save time.

More info »



Forums
  • All of ProZ.com
  • Term search
  • Jobs
  • Forums
  • Multiple search