Hi,
I use the default segmentation rule for colon.
Segmentation after colon usually works fine but in some cases the text is not segmented after colon. E.g.:
Is there any solution for this problem?
Sandor
I have reproduced it
using the change of "second(2)" font of a space after colon - not the "general font"
Attached sample file
Nice spot!
Displaying the non-printable characters and zooming in to the max it shows that the spaces are not the same - one placeholder is square and one is round... plus, when you select each space, the highlight shows that they have a different width:
Still, the internal XML contains just normal spaces (\u0020), so it must be the font representation of the space in the OpenXML or something what makes it a different space than the one understood as "whitespace character" by the \s regexp metacharacter in the segmentation rule.
So, after all I was actually right that the space is not just an ordinary space and that's the reason why the segmentation rule fails.
Perhaps it's time for SDL to review and update the default segmentation rules definitions...
Nice spot!
Displaying the non-printable characters and zooming in to the max it shows that the spaces are not the same - one placeholder is square and one is round... plus, when you select each space, the highlight shows that they have a different width:
Still, the internal XML contains just normal spaces (\u0020), so it must be the font representation of the space in the OpenXML or something what makes it a different space than the one understood as "whitespace character" by the \s regexp metacharacter in the segmentation rule.
So, after all I was actually right that the space is not just an ordinary space and that's the reason why the segmentation rule fails.
Perhaps it's time for SDL to review and update the default segmentation rules definitions...
Translate
