technology under the hood;
SDL TB : Microsoft Access
SDL TM : SQLite

0.7 Mil TU is not so big size

You can delete duplicates with easy ignoring Numbers, Tags. SDL Trados Studio do has this function, but not useful in this case.

Detouring tmx is not a bad idea. But, do NOT use 3rd party tools. Some could be dangerous. Because, SDL Trados Studio has its own bizarre items within it. And, I guess, it will take million years to finish it.

Whatever you do, keep your original copy in the safe place.

Good Luck To You

I have tested this one.

my TM
- size 800 MB
- TUs 0.7 Mil

Elapsed time to finish "Remove Duplicates"
- 23 Min

Performance
- disappointing

I have tested this one.

my TM
- size 800 MB
- TUs 0.7 Mil

Elapsed time to finish "Remove Duplicates"
- 23 Min

Performance
- disappointing

Unknown said:

I have tested this one.

Good for you!

I'd be quite interested in your feedback on it as I know you have done a fair bit working directly on the SQL as it's more performant than using the APIs.  We actually took a more direct approach for many things with the SDL Data Protection Suite for the same reason.

The TM Management Utility Tool is quite popular although it hasn't changed since being introduced a very long time ago.... what was your verdict on it?  What improvements would you prefer to see?  I know you can't/won't do them even though it's opensourced but we might if they seem worthwhile.

github.com/.../TranslationMemoryManagementUtility

you are saying "version 2019 only" again ?

I am a version 2017 customer.

Do not ignore your valuable customer.

Nagoya Kita said:

you are saying "version 2019 only" again ?

I am a version 2017 customer.

Do not ignore your valuable customer.

Nope.  In this case the app is available for all versions since 2009.  But in github we only support 2015, 2017 and 2019 so we have two branches when we update the app (2015/2017 is one branch).