Re: [Kernow] Andrew Welch's CSV2XML stylesheet and Kernow

[Dan V. <dv...@dv...>]

At 06:25 PM 7/15/2013, you wrote:
>The transformation could not get beyond the initial look at the 
>source file; error message something to the effect, something not 
>allowed in the prolog.

Are you are sure the error was about the file being processed and not 
the stylesheet? Sometimes the files will get extra characters before 
the xml prolog statement. You might want to check that.

As far as valid CSV that just depends upon who wrote the spec for the 
stylesheet as to what they expect to find. The prolog error message 
you mention sounds like an issue with the XSLT not the source file 
you are processing. If you want to test the source file, I would just 
take a couple of rows of the file, and clean them up. So I would look 
at the use of quotes (single and double), MS products like to 
substitute curly quotes and long dashes which can be hard to spot. 
I'm not familiar with this stylesheet, but I'm assuming it is looking 
for tabs as the separation character, so you might want to look for 
missing tabs such that not all the columns are there. Basically I 
would simplify the source as much as possible to create a test source 
file to verify you have everything working.

If you still can't get it to work, Ken Holman has a  general purpose 
CSV processing stylesheet that you might want to try out as well. 
<http://www.CraneSoftwrights.com/resources/#csv>http://www.CraneSoftwrights.com/resources/#csv

..dan
---------------------------------------------------------------------------
Danny Vint

Panoramic Photography
http://www.dvint.com

voice: 619-647-5780