Difference: UnicodeCNV_BINStmt (2 vs. 3)

Line: 1 to 1

META TOPICPARENT	name="UnicodeTestPlan"

CNV BIN

Line: 35 to 35

convert valid literal unicode value to alpha/national and alpha/unicode
try to CNV BIN from literal into alpha/national and alpha/unicode that are not 1 character (Em build error)
try to CNV BIN from literal into alpha/national and alpha/unicode from invalid numerics (negative, >=4294967296)

Added:

>
>

Changed:

<
<

Painted 1 character alpha/national field on image, displayed as 4 characters. alpha/unicode was ok

>
>

Painted 1 character national field on image, displayed as 4 characters. unicode was ok

CNV BIN from negative literal does not generate compile time error, generates 'Invalid Sign' runtime error. Might not be actual bug, just different implementation?

Changed:

<
<

>
>

Not a bug, just different implementation from above: CNV BIN from national/unicode field >1 character is rejected by ILF editor. If forced via Opt 88, correctly produces compile error
National fields are not transcoded to their proper character set. I can CNV BIN a character from the Unicode character set (but that does not exist in the National character set as specified on the SYSPARM file) into a National field, and it will display the Unicode character. Instead, it should transcode the Unicode character to the National character, and display that.
Overflow errors are not occurring at the expected limit. The max value for a 4 byte unsigned number is 4294967295, if I use this value I get overflow errors. If I decrease that by 1 to 4294967294, I do not get an overflow.

-- JeanNeron - 2011-03-11