SonarQube does not seem to consider that the unicode character U+2028 as a valid line separator for lines of code. As a consequence, code analysis features may break for some languages (e.g. YAML) that accept this character as a valid line separator.
Example: a text file as follows should be considered as having 2 lines of code
"line 1" + U+2028 + "line 2"
Same request for U+2029 and U+0085.