Skip to Content
0
Former Member
Jan 24, 2007 at 09:55 AM

Interesting invalid char. issue: all error string broken at 13th Unicode ch

49 Views

We keep getting data load error for the data source of 0FI_GL_4 (one record error on our test system, and two records on our production system) at the InfoObj. 0POSTXT (with a length defination of 60 on BW, the corresponding R3 field is SGTXT with a length defination of 50). All the three record error msgs are listed below and also the match of each four HEX digit and one unicode char. are together listed correspondingly, through the match, find a very interesting phenomenon that all these three error record Unicode always broken at 13th char.

1. One record error on our Test system and the error msg and the match table:

Value 'rsz bijdragen w#40169547#06-02-06#625,24(hex.

'00720073007A002000620069006A0064007200610067006500') of

characteristic 0P OSTXT contains invalid characters

Here I separat each four HEX digits in this string and match these every four HEX digits and input char. in the table below:

r s z b i j d r a g e n

0072 0073 007A 0020 0062 0069 006A 0064 0072 0061 0067 0065 00

char. n should be 006E, but somehow the unicode was broken at this char. which is the unlucky 13th char.

2. One record error on our PROD system and the error msg and the match table:

Value '114280 WESCO DRUM LIFTER # 1000-LB. CAPA ' (hex.

'00310031003400320038003000200057004500530043004F00') of

characteristic 0P OSTXT contains invalid characters

1 1 4 2 8 0 W E S C O

0031 0031 0034 0032 0038 0030 0020 0057 0045 0053 0043 004F 00

space should be 0020, but somehow the unicode was broken at this char. which is the unlucky 13th char.

3. Another record error on our PROD system and the error msg and the match table:

Value 'AMK MANUFACTURING ROADMAG 72IN. LENGTH' (hex. '0041004D004B0020004D0041004E0055004600410043005400') of characteristic 0POSTXT contains invalid characters

A M K M A N U F A C T U

0041 004D 004B 0020 004D 0041 004E 0055 0046 0041 0043 0054 00

char. U should be 0055, but somehow the unicode was broken at this char. which is the unlucky 13th char.

As long as we manually edit this Item Text field (no matter what char. we input with our standard English keyboard including # and fill the field to a full lengh

) on R3 or PSA on BW and save it, and then load to BW, there is no any error at all. Our RSKC value is ALL_CAPITAL as recommended.

Any idea why and the solution?

Thanks a lot!