From: anne001 <anne@...>
Date: 2005-11-24T03:17:26+09:00
Subject: Re: Parsing excel CVS data on a mac OSX to extract blocks of cells

I have moved onto Tiger. Still much installing to do. Sorry in the
delay in processing your responses.

The line ending is extremely simple,
http://www.delorie.com/gnu/docs/emacs/emacs_220.html

....-unix     newline to separate lines.
....-dos      carriage-return +  linefeed to separate lines
....-mac     carriage-return to separate lines

But the editors show what they want. Text wrangler lets you replace \r
with \n
but when you look, it still looks as a \r. Did it make the
substitution?
Emacs has a command to change the file to unix. But the ^M are not
changed. Did something change? hard to say. So I gave up on that.

I like the command, thank you
ruby -p -i.bak -e 'sub("\r", "\n")' your_csv_file_here.csv

It seems to work, it creates your_cvs_file_here.csv.bak and csv read no
longer gives an error, but... It seems to parse each cell, with realms
of NIL in between bits of text.
Conceptually I think it is easier to process line by line. But I will
look into CVS to see what it can do.

In any case, I can combine the ruby line command with the readlines
ruby function. So I should be OK.