Kbase P110878: Invalid UTF-8 characters in output from a 4GL program.
| Autor |
Progress Software Corporation - Progress |
| Acesso |
Público |
| Publicação |
16/10/2008 |
|
Status: Unverified
SYMPTOM(s):
Invalid UTF-8 characters in file.
FACT(s) (Environment):
File is created via 4GL/ABL.
Using OUTPUT TO ... CONVERT TARGET "UTF-8".
Session internal codepage ( -cpinternal ) is a single-byte codepage, such as iso8859-1.
Only occurs under ChUI.
protermcap was customized via the IN and OUT options, to allow correct input and visualization of extended characters, such as Ä and é.
UNIX
Progress/OpenEdge Versions
CAUSE:
The IN and OUT options are impacting how the UTF-8 file is produced.
After the conversion from the -cpinternal codepage to UTF-8, a further conversion is done relying on the mappings from the protermcap file. This can lead to incorrect or malformed UTF-8 characters.
FIX:
Add the NO-MAP option to the OUTPUT statement:
OUTPUT TO "myFile" NO-MAP CONVERT TARGET "UTF-8".