Oct. 15, 2014
5:37 a.m.
On Oct 15, 2014, at 1:02 AM, Evan Donahue <emdonahu@gmail.com> wrote:
I am processing natural language statistics over the entire wiki corpus, not querying for specific entries. The information and entries aren't important so much as the raw quantity of words. A simple stream through a file on disk is all I need.
Maybe you can workaround the problem by writing a shell script that writes the file over a pipe or socket. Then on the that Pharo side, connect to the pipe/socket, and read the data stream.