Created attachment 38834 [details] [review]
proposed patch (only for text-delete events for single-line texts)

Remus, in your patch above, in srl_event_compute_line(), you check 'if
(event->detail1)', but if not, you continue to use it (in the line two lines
later).  Wouldn't it be better to simply return if detail1 or detail2 are null,
and otherwise then continue with the rest of your logic?

Also, later in the patch in srl_event_listener() why not use the
just-computed/cached GPOINTER_TO_INT (data) in ev->type?

Otherwise I see nothing to object to in this patch.  We understand that this is
addressing a specific case where we have a known problem, but that this more
general solution maybe should be used (in the future) elsewhere.  And at that
point we'll perhaps need another patch...

event->detail1 and event->detail2 represent the start position and the length of
the deleted text. The new text is obtained by concatenating text before
"detail1" and the text after "detail1 + detail2" (= removing the text between
detail1 and detail1 + detail2). 

>Also, later in the patch in srl_event_listener() why not use the
>just-computed/cached GPOINTER_TO_INT (data) in ev->type?
No reason, probably a copy-paste issue. I agree that event->type should be used.

Created attachment 46527 [details] [review]
reworked patch

Great.  Please apply.

There have not been any code commits to Gnopernicus for more than two years, and Gnopernicus has been superseded by Orca in GNOME.
Gnopernicus has been moved to the SVN archives today, and the developers did not respond to any emails.
Hence I'm closing all remaining 25 open bug reports as WONTFIX.