Sandor Juhasz 

This one is possible:

But you have to trick it!  You are forced to add a thousand separator into the settings as well as a decimal, so copy paste a zero-width space.  You can take it from here:

https://unicode-table.com/en/200B/

Just click the copy button here:

And paste straight into here:

You obviously won't see anything, but you will now be able to save your new rule and use it.

Sandor Juhasz 

This one is possible:

But you have to trick it!  You are forced to add a thousand separator into the settings as well as a decimal, so copy paste a zero-width space.  You can take it from here:

https://unicode-table.com/en/200B/

Just click the copy button here:

And paste straight into here:

You obviously won't see anything, but you will now be able to save your new rule and use it.

Maybe I was not clear.

By default, if a number contains no thousand separators in the source segment, the localised number in the target segment will not contain thousand separators either. Just the same way as we use a zero width space.

So, my question: is it possible to localise a source number without thousand separators like this:

as a target number with thousand separators like this:

Sandor Juhasz 

ok - understood.  The answer is no and the reason is because Trados Studio doesn't allow for no separator at all.  So you get this:

If I trick it I thought that might work but actually it doesn't because you would actually need to have zero width spaces as a thousand separator, like this:

But this obviously won't help you.

I think you have two workarounds.  The first is simple if you are working interactively, and that would be to use the Regex AutoSuggest Provider.  That is a simple fix, but obviously doesn't help with pre-translation or auto-propagation.

The second would be to use something like the SDLXLIFF Toolkit and search replace all your numbers in the source and add a thousand separator.  Then you would have what you really want in terms of handling the numbers in the target correctly.

But the real solution is obviously to raise an enhancement request to support no thousand separators at all.  Do that here:  Trados Studio Ideas 

Thank you Paul. I will raise an enhancement request.

Anyway, in your last example:

it seems it works. I don't really understand what happens here.

Sandor Juhasz 

It works because I added a rule that included zero-width spaces as the thousands separator AND that number actually had zero-width spaces as thousands separators.  Hence me saying this obviously won't help you.

I see now. Thanks, Paul.

By the way, Paul,

Could you please propose a regex rule for this problem to be used in Regex Match AutoSuggest Proivder?

Sandor Juhasz 

I don't know how to do this with one rule.  I'm not sure it can be... so you just simply do this:

Add as many as you think you'll need.  Pain in the neck but once you've done it you have it forever and can reuse as much as you need.

The toolkit is easier... search for this:

\d(?=(?:\d{3})+(?!\d))

Replace with this:

$0,

Then you have all your thousand separators in place.

Sandor Juhasz 

If it helps... it's quite quick to just edit this file:

c:\Users\%username%\AppData\Roaming\Capybara.EditorPlugin.RegexMASProvider\Capybara.EditorPlugin.RegexMASProvider.settings.xml

Here's the entries I set up in a couple of mins:

      <Entry>
        <IsEnabled>true</IsEnabled>
        <Description>1</Description>
        <RegexPattern>\b(\d{1}),(\d+)\b</RegexPattern>
        <ReplacePattern>$1.$2</ReplacePattern>
      </Entry>
      <Entry>
        <IsEnabled>true</IsEnabled>
        <Description>2</Description>
        <RegexPattern>\b(\d{2}),(\d+)\b</RegexPattern>
        <ReplacePattern>$1.$2</ReplacePattern>
      </Entry>
      <Entry>
        <IsEnabled>true</IsEnabled>
        <Description>3</Description>
        <RegexPattern>\b(\d{3}),(\d+)\b</RegexPattern>
        <ReplacePattern>$1.$2</ReplacePattern>
      </Entry>
      <Entry>
        <IsEnabled>true</IsEnabled>
        <Description>4</Description>
        <RegexPattern>\b(\d{1})(\d{3}),(\d+)\b</RegexPattern>
        <ReplacePattern>$1,$2.$3</ReplacePattern>
      </Entry>
      <Entry>
        <IsEnabled>true</IsEnabled>
        <Description>5</Description>
        <RegexPattern>\b(\d{2})(\d{3}),(\d+)\b</RegexPattern>
        <ReplacePattern>$1,$2.$3</ReplacePattern>
      </Entry>
      <Entry>
        <IsEnabled>true</IsEnabled>
        <Description>6</Description>
        <RegexPattern>\b(\d{3})(\d{3}),(\d+)\b</RegexPattern>
        <ReplacePattern>$1,$2.$3</ReplacePattern>
      </Entry>
      <Entry>
        <IsEnabled>true</IsEnabled>
        <Description>7</Description>
        <RegexPattern>\b(\d{1})(\d{3})(\d{3}),(\d+)\b</RegexPattern>
        <ReplacePattern>$1,$2,$3.$4</ReplacePattern>
      </Entry>
            <Entry>
        <IsEnabled>true</IsEnabled>
        <Description>8</Description>
        <RegexPattern>\b(\d{2})(\d{3})(\d{3}),(\d+)\b</RegexPattern>
        <ReplacePattern>$1,$2,$3.$4</ReplacePattern>
      </Entry>
            <Entry>
        <IsEnabled>true</IsEnabled>
        <Description>9</Description>
        <RegexPattern>\b(\d{3})(\d{3})(\d{3}),(\d+)\b</RegexPattern>
        <ReplacePattern>$1,$2,$3.$4</ReplacePattern>
      </Entry>
            <Entry>
        <IsEnabled>true</IsEnabled>
        <Description>10</Description>
        <RegexPattern>\b(\d{1})(\d{3})(\d{3})(\d{3}),(\d+)\b</RegexPattern>
        <ReplacePattern>$1,$2,$3,$4.$5</ReplacePattern>
      </Entry>
            <Entry>
        <IsEnabled>true</IsEnabled>
        <Description>11</Description>
        <RegexPattern>\b(\d{2})(\d{3})(\d{3})(\d{3}),(\d+)\b</RegexPattern>
        <ReplacePattern>$1,$2,$3,$4.$5</ReplacePattern>
      </Entry>
            <Entry>
        <IsEnabled>true</IsEnabled>
        <Description>12</Description>
        <RegexPattern>\b(\d{1})(\d{3})(\d{3})(\d{3})(\d{3}),(\d+)\b</RegexPattern>
        <ReplacePattern>$1,$2,$3,$4,$5.$6</ReplacePattern>
      </Entry>
            <Entry>
        <IsEnabled>true</IsEnabled>
        <Description>13</Description>
        <RegexPattern>\b(\d{2})(\d{3})(\d{3})(\d{3})(\d{3}),(\d+)\b</RegexPattern>
        <ReplacePattern>$1,$2,$3,$4,$5.$6</ReplacePattern>
      </Entry>
            <Entry>
        <IsEnabled>true</IsEnabled>
        <Description>14</Description>
        <RegexPattern>\b(\d{3})(\d{3})(\d{3})(\d{3})(\d{3}),(\d+)\b</RegexPattern>
        <ReplacePattern>$1,$2,$3,$4,$5.$6</ReplacePattern>
      </Entry>
            <Entry>
        <IsEnabled>true</IsEnabled>
        <Description>15</Description>
        <RegexPattern>\b(\d{1})(\d{3})(\d{3})(\d{3})(\d{3})(\d{3}),(\d+)\b</RegexPattern>
        <ReplacePattern>$1,$2,$3,$4,$5,$6.$7</ReplacePattern>
      </Entry>

Works nicely :-)

Brilliant! Thank you very much.