Regression in TQPresenter

The MR !625 (merged) seems to have introduced a couple issues. As @sketabch informed me a couple days ago, after updating she's seeing a couple issues in the cutflow printing:

• horizontal lines are partially missing (can confirm)
• some derived quantities (signal significance) seem wrong (yet to be confirmed)

At least the first of these aspects can be understood (in retrospect) looking at the changes here: 4320e3a5 . While adding support for "merging" entries with the same name seems like a good idea this currently prevents to use more than one horizontal separator of each line multiplicity. The reason is that these are specified as

.name = "|"

in the cutflow cut files (in the process files the ".path" tag is used instead, hence, this behaves differently). Now, as entries with the same .name are merged this implies that we only see the first use of .name="|" as well as the first use of .name="||" (double line) and so on.

Tagging @tapark @cburgard as they were involved in the original MR.