FreeTDS Development Group

[galt AT folkplanet.com]

It would be nice to be able to say how the NULL value
should be expressed. Currently it looks like 
an unquoted empty string, i.e. if your separator is 
<my#$#@separator><my#$#@separator>.  

MySQL 5 load data local infile will give a bunch of 
warnings with scary messages like out of range value
and data truncated.  For CSV, it would like to see
unquoted NULL (i.e. the only thing that ISN'T quoted), 
and for special separator strings (non-csv), it wants it to 
be escaped as \N.  

If you gave freebcp a way to express nulls, e.g. -Null '' (equivalent to 
what happens now) -Null '"NULL"' (actually this is probably not a useful 
example) -Null 'NULL' (great for csv) -Null '\N' (great for mysql with 
custom field/line separator strings) etc, then the user could get what he 
needs.

Right now I am having to write a custom program to do 
this because sed is bombing out on big input files
(several tables around 6GB).

Plus, even then, it will end up turning all "" into NULL.

-Galt



On Mar 31 2006, Bort, Paul wrote:

> > 
> > Ahem.  I understand they're technically different, of course, 
> > but how are
> > they semantically different?  Which user can tell the 
> > difference between a
> > null field and a zero-length string?  And what is the 
> > difference between
> > an empty field and a field that's empty?  
> > 
>
> Just a quick example from my current project of how '' and NULL are
> different and that difference can be important:
>
> We have a table that holds user settings, and a null ( or a missing row
> ) means that the default (from a separate table) should be used. OTOH,
> if the column is an empty string, that means that the setting should be
> an empty string, even if the default says otherwise. If a setting is a
> file prefix, an empty string is a reasonable value.
>
> _______________________________________________
> FreeTDS mailing list
> FreeTDS AT lists.ibiblio.org
> http://lists.ibiblio.org/mailman/listinfo/freetds