I did an export from Palm Desktop. I have 3 dummy
records. One is comma seperated and the other is tab
seperated. DSV correctly guesses the comma seperated
case, but not the tab seperated case. I have attached
the tab seperated file.
Might I also suggest adding an optional original
filename argument to guessDelimiter. If it ends in tsv
then the delimiter is likely to be tab, if in csv then
comma and if something else then the existing
heuristics near the end of the function are fine
(although I would argue that comma should come
before tab on line 281).