🚀 go-pugleaf

In article <Bu9Lv8.6pL@ireq.hydro.qc.ca>, beaurega@ireq.hydro.qc.ca (Denis Beauregard) writes:
> There is a "natural" method of sorting according to which language
> (i.e. French, not Basic :-)  ).  DOS sort in recent versions takes care
> of that.  The method is quite simple (no idea if they do that, but I
> do that in my own qsort-based psort program) : if under 0x80, keep it
> as is (lowercase or lowercase if needed), if over, using a table lookup,
> remap the character...

I've always wonderred what the "proper" way is to sort letters with accent
marks (grave, acute, umlaut, etc). Could somebody please explain the order of,
for example, the standard IBM PC extended characters:  131, 132, 133, 134 and
160, which, presumably, come after 97 (a) but before 98 (b)?

I presume all this changes when different country pages are selected, and when
you are running MS-Windows, and when outputting to a variety of printers all
with their own characters sets (and so collating sequences).

Mark Aitchison, University of Canterbury, New Zealand.