Hi  

Just select "Add new translation units" then you update your TM:

Daniel

so, is this what you wanted ?

I am just a beginner so, I do not pretending know all.

Above screenshot is made while testing some Check/UnCheck at the TM level.
Like this (as I wrote above, it is my testing, you do yours)

And, never mind SDL Trados Studio Guru(s).

Since this is going into a more and more absurd direction, I would just want to state that Evzen's question is very valid - why would you want to store a number in a TM? The way I pointed out is the way to do this, but the question is why would you want to increase your TM size with irrelevant entries? The number is going to be substituted in later translations anyways - unless you follow BTD's route (below) which will seriously cripple your TM's performance and usability. (I am copying the screenshot into this post because some people edit and delete their posts in a rather confusing manner.)

Daniel

seems you enjoy making needless noise

Don't you know THE worst case of needless increase the size of TM ?
SDL's improvement. It wastes 3 times of all and every TU you use.

This questioner's result just increase several TUs, at most several hundred. That is tiny, tiny and tiny.

In terms of both performance and size, it effects NOTHING.

Basically, TM treats size of GB stuffs and Millions TUs. Do not forget this simple fact.

Hi  

You've had a few answers to your question now, so I think you need to make a decision on why you want to do this.  To add to the point from Evzen, the out of the box solution in Trados is that you would translate a number once and then never need to do this again as Studio holds the information to handle future numbers already.  So if you translate this:

$1,255

Then these would be 100% matches and they would not be stored in the TM:

$899

$2,476

If you translate this:

It cost me $25 to buy dinner.

Then these would also be 100% matches and only the first one would be stored in the TM:

It cost me $271 to buy dinner.

It cost me $3 to buy dinner.

If you turn off the number recognition as Kelly suggested then you would never get 100% matches for these repetetive segments and each unique number would have to be stored in the TM.  You can do this, but I am wondering why you would want to do this?