The problem is that hypview uses Enca and its automatic encoding detection
internally on the hyp @node contents. Enca does is not able, in case the
@node contents is short enough or doesn't contain enough diacritics, to
detect the correct source detection.
One possible solution may be to let enca sift through first few nodes (or all as it
is quite fast) of the .hyp document. This source encoding property could be then
cached in some way so that the detection only happens once.
The problem is that hypview uses Enca and its automatic encoding detection
internally on the hyp @node contents. Enca does is not able, in case the
@node contents is short enough or doesn't contain enough diacritics, to
detect the correct source detection.
One possible solution may be to let enca sift through first few nodes (or all as it
is quite fast) of the .hyp document. This source encoding property could be then
cached in some way so that the detection only happens once.