escaping a carriage return

I am having a problem with escaping the carriage return from a memo (aka 
longText) field.

I am using Bowen's escapeURL routine (see below).  For some reason I get 
%D%A when unescaped (which of course is the translation for carriage return 
line feed I am sure).

How can I properly escape and unescape this?

Function escapeURL(cData)
 //   Bowen Moursund (dBASE, Inc.)
 //   March 2000
   local cRetStr, nLen, cChar, i
   cRetStr = ""
   nLen = len(cData)
   for i = 1 to nLen
      cChar = substr(cData, i, 1)
      if cChar = " "
         cRetStr += "+"
      elseif not cChar $ 
"&=0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ"

         cRetStr += "%"+itoh(asc(cChar))
      else
         cRetStr += cChar
      endif
   next i
   return cRetStr

Claus Mygind wrote:

> cRetStr += "%"+itoh(asc(cChar))

Does it work if you add the second parameter to the itoh() function?

Interesting, but it should capture all non-characters not just the carriage 
return.

It is possible that the problem occurs in the unEscape phase.  I did observe 
that the content which was one line then two carriage returns and then 
another line.  First word on the last line was email.  the "e" was
converted 
into a spcial charcter prefaced by the % and of couse there was no line 
returns inserted in the converted text.

Well you got me thinking now and that's good.

Thanks for the help.

"Michael Nuwer" <michael@forget.it> wrote in message 
news:793NVPYyHHA.696@news-server...
> Claus Mygind wrote:
>
>> cRetStr += "%"+itoh(asc(cChar))
>
> Does it work if you add the second parameter to the itoh() function?
>
>
> cRetStr += "%"+itoh(asc(cChar),2) 

Claus Mygind wrote:
> Interesting, but it should capture all non-characters not just the 
> carriage return.

I'm not sure I understand what your getting at here. The problem I saw 
is that the first 16 characters char(1)-char(16) have a leading "0" in

hex, but the function as used by Bowen drops the leading "0". So 
char(13) becomes "%D" whereas it should be "%0D"

> 
> It is possible that the problem occurs in the unEscape phase.  

I tested with webclass.cc's OEMFormat function to unEscape the encoding. 
"%D" fails, but "%0D" succeeds. Here is an example

set procedure to :webWizards:webclass.cc addi
o = new CGISession()
clear
for i=9 to 30
  cChar = chr(i)
  cRetStr = "%"+itoh(asc(cChar))
  ? ""+i
  ?? cRetStr at 10
  ?? asc(o.OEMFormat(cRetStr)) at 15
next


> I did 
> observe that the content which was one line then two carriage returns 
> and then another line.  First word on the last line was email.  the
"e" 
> was converted into a spcial charcter prefaced by the % and of couse 
> there was no line returns inserted in the converted text.
> 
> Well you got me thinking now and that's good.
> 
> Thanks for the help.
> 
> "Michael Nuwer" <michael@forget.it> wrote in message 
> news:793NVPYyHHA.696@news-server...
>> Claus Mygind wrote:
>>
>>> cRetStr += "%"+itoh(asc(cChar))
>>
>> Does it work if you add the second parameter to the itoh() function?
>>
>>
>> cRetStr += "%"+itoh(asc(cChar),2) 

Thanks that would have taken eons for me to figure out.  I not sure I 
understand  your function below.  But since you have clarified the problem, 
I am sure I can work through it.

"Michael Nuwer" <michael@forget.it> wrote in message 
news:dclv82eyHHA.1932@news-server...
> Claus Mygind wrote:
>> Interesting, but it should capture all non-characters not just the 
>> carriage return.
>
> I'm not sure I understand what your getting at here. The problem I saw is 
> that the first 16 characters char(1)-char(16) have a leading "0"
in hex, 
> but the function as used by Bowen drops the leading "0". So
char(13) 
> becomes "%D" whereas it should be "%0D"
>
>>
>> It is possible that the problem occurs in the unEscape phase.
>
> I tested with webclass.cc's OEMFormat function to unEscape the encoding. 
> "%D" fails, but "%0D" succeeds. Here is an example
>
> set procedure to :webWizards:webclass.cc addi
> o = new CGISession()
> clear
> for i=9 to 30
>  cChar = chr(i)
>  cRetStr = "%"+itoh(asc(cChar))
>  ? ""+i
>  ?? cRetStr at 10
>  ?? asc(o.OEMFormat(cRetStr)) at 15
> next
>
>
>> I did observe that the content which was one line then two carriage 
>> returns and then another line.  First word on the last line was email.

>> the "e" was converted into a spcial charcter prefaced by the
% and of 
>> couse there was no line returns inserted in the converted text.
>>
>> Well you got me thinking now and that's good.
>>
>> Thanks for the help.
>>
>> "Michael Nuwer" <michael@forget.it> wrote in message 
>> news:793NVPYyHHA.696@news-server...
>>> Claus Mygind wrote:
>>>
>>>> cRetStr += "%"+itoh(asc(cChar))
>>>
>>> Does it work if you add the second parameter to the itoh()
function?
>>>
>>>
>>> cRetStr += "%"+itoh(asc(cChar),2)
>>