this is not a julia problem, but an emacs (utf8) problem using julia, but I was hoping someone had an answer for this. I use macos, iterm2 and terminal.
Copy the output of
d= DataFrame( x=1:3 )
10×1 DataFrames.DataFrame │ Row │ x │ ├─────┼────┤ │ 1 │ 1 │ │ 2 │ 2 │ │ 3 │ 3 │
Now paste it into a terminal shall via
cat > /dev/null and it works fine (suggesting that the terminal emulator and the bash configuration are all ok). Paste it into a Quartz emacs.app and it works fine, too (suggesting emacs is just fine, too).
but try pasting it into a terminal emacs, and hell breaks loose
Wrong type argument: char-or-string-p, 134217795. Unlike quartz emacs, the terminal emacs also displays UTF-8 characters badly, for example, with \u2502 instead of the vertical line. Pretty sure this is related.
there are apparently many moving parts here–utf8, terminal emulation, bash, and emacs. I have tried a whole host of related configurations, but have not stumbled onto the right ones.
what config does it take to convince terminal emacs to be julia-utf8 friendly? help, as always, appreciated.