cuneiform team mailing list archive

["Jussi Pakkanen" <jpakkane@xxxxxxxxx>]

On Tue, Oct 21, 2008 at 4:11 PM, Ladwig, Mike <mike.ladwig@xxxxxxx> wrote:

> First, does this project remain active?

Yes. I have just been busy with other stuff recently.

> I have been able to get the "winport" branch to build correctly under
> Windows with Magick++ using the MSYS environment.  However, OCR fails in
> recognition because "load_stat_dict" reading language data apparently has
> data format problems.
>
> I'm guessing this code has been reworked as part of the Linux port.  Any
> pointers on what to look for?

I changed the "end-of-line" character from '\n' to '\r' because some
of the data files have DOS-type newlines (CR+LF). I should have done
it cleanly the first time, but I was lazy.

Anyhow, the code should look roughly like this (and at some source
files it already is):

#if WIN32

#define EOL '\n'

#else

#define EOL '\r'

#endif

....

// read_stuff_until_newline
if(EOL == '\r')
  read_one_char();