As strange as it might seem - there is no "standard" for CSV other than common convention and the name.
CSV's or Comma Seperated Value files are just that - a text file containing tabluar data where each line is a "row" seperated into "columns" by commas.
It is usually used as a common file format for sharing data between spreadsheets (i.e. Excel, Lotus, Quatro, whatever).
I thinks the rules are something like this
* file is plaintext, human readable and viewable in any text editor
* each line is terminated with a newline
* each line has the same number of elements
* the element seperator is a comma (it puts the C in CSV)
* commas may have zero or more spaces around them, but they are ignored upon parsing
* if an element contains a comma, that element needs to be placed in "double quotes"
* optional, the first row contains column names
* fields with newlines need to be wrapped in "double quotes"
* fields with leading or trailing spaces need to be wrapped in " double quotes "
* for excel, if you want to force your entry to be a string - put a single ' quote at the beginning.

Why bring this up? Well, a customer's lead supplier got the bright idea to suffix the 8bit ascii char 0xA0 to every single field!
Apparently they were haveing problems with excel importing east coast zip codes with leading zeros as zip codes, so rather than prefix those fields with a (') they decided to add (what is on the US code page) is a non printable space.
From what I've seen, this same character is an (a') on some euro code pages.