Re: Camel Case

FWiW: Unfortunately that algorithm would generally produce what is typically considered "A Title Case" value instead of "aCamelCase" value; plus has an unstated assumption that the input value must not be all blanks nor an empty string. As others had provided similarly, that algorithm also does not support making lowercase from an international character set. Consider that the upper case name [the OP wants to process names] "EMIL SOMMARSTRÖM" processed by that algorithm produces a Title Case value with a Camel Case value for the second token, i.e. the value "Emil SommarstrÖm", instead of producing the Title Case value "Emil Sommarström".

Regards, Chuck

On 11-Feb-2012 06:37 , Emil Sommarström wrote:

More samples:)

ptoNameCase...
p b
d pi 200a varying
d name 200a value varying

d uc c 'ABCDEFGHIJKLMNOPQRSTUVWXYZ'
d lc c 'abcdefghijklmnopqrstuvwxyz'
d pos s 5u 0
/free

name = %xLate(uc:lc:%trim(name));
dou pos = 0;
pos += 1;
%subst(name:pos:1) = %xLate(lc:uc:%subst(name:pos:1));
pos = %scan(' ':name:pos);
endDo;
return name;

/end-free
p e

On Thu, Feb 9, 2012 at 2:58 AM, Shannon ODonnell wrote:

Thanks!

sjl on Wednesday, February 08, 2012 3:34 PM wrote:

Here is what I was going to suggest:

I would store the conversion information in a dictionary file as
pairs of data, with the "From" word in all caps. In order to
perform conversion, take a word from the string being converted
and convert the value to all caps, then CHAIN to the dictionary
file with the value in all caps.

If found, the "to" value would be returned in CamelCase format.

Sample Dictionary file DDS:

UNIQUE
R @XLAT
WORDFROM 60A
WORDTO 60A

K WORDFROM

Example data:
WORDFROM WORDTO
------------------ ------------
AGRIFINANCE AgriFinance

Matt wrote:

If it's the first item from sjl then I would just create a
dictionary of either upper or lower cased words and their mixed
case counter parts.