Changing the colon segmentation rule

Question

I am new to Regex and am trying to change the existing colon rule in Trados/Groupshare 2017 so that it only breaks when a colon is followed by a hyphen/dash. 
 Here is an example of the kind of source text (German) I mean: 
 Technische Daten: 
 -Eingangsspannung: 100-240V 50/60 Hz 
 -Konstanter Strom: 700mA DC 
 So I've tried changing the existing colon rule to the following: 
 Before break = .[:]+ 
 After break = [- &ndash; &mdash;] 
 But something seems to be wrong - presumably my Regex - as it hasn't done the trick. 
 I hope one of you will be able to point me in the right direction :-) 
 Thanks in advance! 
 Rachel

Nora Díaz · Accepted Answer

Sorry, I had written another post about modifying the colon segmentation rule but in fact, removing the default colon segmentation rule is all that's needed in this case. 
 The key to understanding why is that two different segmentation rules come into play when you have text like this. 
 Technische Daten: (1) 
 -Eingangsspannung: (2) 100-240V 50/60 Hz 
 -Konstanter Strom: (2) 700mA DC 
 
 Text is being segmented at (1) not because of the colon, but because of the new line (likely a hard return). 
 By removing the colon segmentation rule, text won't be segmented at (2) anymore.

Paul · Answer

Hi 
 
Does that mean the source text is actually like this? 
 
Technische Daten: -Eingangsspannung: 100-240V 50/60 Hz -Konstanter Strom: 700mA DC 
 
Or are you trying to achieve this: 
 
Technische Daten: 
- 
Eingangsspannung: 
100-240V 50/60 Hz 
- 
Konstanter Strom: 
700mA DC 
 
Sorry for the question... I'm a bit confused by what you have now and what you are trying to achieve?

Paul · Answer

Hi , In this case you would have less problem if you were given the html as it would simple to achieve what you need just by removing the colon rule as previously suggested. The

tag would ensure that "Technische Daten:" is indeed in a segment on its own.

Paul · Answer

Thanks 
 
I think Evzen is probably right. I can only parse this XLIFF anyway by removing the <HtmlText> element prior to the CDATA section as this is incorrect. I also don't think the segmentation rule will have an effect if you are handling the file as XLIFF. You might have to handle it as XML instead and deal with the target element only... then I think you'll have more success. 
 
If you're dealing with the text extraction already discussed then the segmentation rule changes should work fine.

Trados Studio > 5. Regex and XPath

Changing the colon segmentation rule