PROBLEM

At times it happens, that all content of a content-exchange in the WYSIWYG goes into the first detected part of the target-segment although there are multiple parts. That e.g. leds to all words of a segment being bold although just the e.g. first word should be bold.

Another deficiancy is, that often segments ending with a superscript/subscript character like "©" will not be correctly distributed, as a product-name like "XYZ©" will be handled as a single word.

Example tasks: see nextcloud/Developer/test-data/visual/problem-textdistribution-WYSYWIG

SOLUTION

a) Find the reason for the distribution algorithm not to work & fix it

b) Improve the distribution for edge cases: The last part of the target-segment contains just 1 character AND the last word of the segment-content ends with a character qualified to represent superscript/subscript - or is just a single character. In this case put the last char to the last part and continue the distribution of the rest as usual.