On Feb 18 at 09:51 -0500, Gwern Branwen wrote:
> On Thu, Feb 18, 2010 at 3:33 AM, Peter Bienstman <[email protected]> 
> wrote:
> >> I've been using &mdash; up to this point for -- dashs, but it occurred
> >> to me that those are ugly in the source and why not use normal Unicode
> >> entities like — ?
> >>
> >> So I replaced one using the edit box, and it displayed right, so I
> >> went and did a mass-replace on my .mem and found that Mnemosyne would
> >> neither read nor write it. After some searching, I found this is
> >> apparently a known limitation of Python's pickling - it won't
> >> serialize Unicode characters above a certain codepoint, and the mdash
> >> character is apparently such a point. (My working hypothesis had been
> >> that my regexp replace had gone awry.)
> >>
> >> Is there any work around for this, or is it a legitimate Mnemosyne bug?
> >
> > If it saves OK when you manually replace one with the edit box, then I guess
> > your mem-editing went awry.
> 
> No, I checked this very carefully: I could save the em-dash when
> adding it in an edit box, but Mnemosyne would not close: on the
> console it would emit an error about writing to thus-and-thus a point
> and then failing, and this would happen even with a .mem otherwise
> unmodified. So I'm sure that it was not my editing but the em-dash
> character conversion itself.

FWIW, I've been using em-dashes (U2014) in Mnemosyne for months
without any problems.

Tim.

Attachment: signature.asc
Description: Digital signature

Reply via email to