CTRL-M Problem

Hi,

I'm moving some tables from Sybase to Oracle (9i). I can dump the data
out from Sybase into files using ~ as the field seperator and <CR> for
the record end. The problem I have is that some of the fields have ^M
in them. I have used sed to replace the ^M's with <br> html tag (The
data ends up in Oracle and is accessed using PHP via a browser). The
problem is that there is still a <CR> after the <br> tag meaning that
sqlldr has trouble loading these records. I have tried sed
s/^M$/\<br\>/g without luck. An example record which causes problems
looks like:

3~5HIAA 24 Hr Urine                                                   
           ~U5HIAAT                                           ~~St
Vincent's~Biochemistry                                      ~Note UVOL
and  pH on request form. Aliquot 100mls send refrigerated.  Acid
bottles (30ml 60% HCl) required for collection.~Urine container is
available from Pathology Ext 3223. Please provide details of current
medications.^M
Dietary instructions:  for at least 12 hours prior to and during
collection of 24 hour urine, avoid consuming the following - avocados,
bananas, pineapples, plums, tomatoes, nuts, chocolate, eggplant,
kiwifruit, figs, grapes.  These foods contain chemicals which may
interfere with laboratory testing.~10 days~~WD   ~E   ~~7~~~~~~

Using the above sed gives me ^M replaced by <br> but it still has the
<CR> in it meaning that sqlldr sees it as two lines.

Anyone sorted this one out before?

Thanks,
Wayne
--
http://www.freelists.org/webpage/oracle-l

Other related posts: