SE Tool Kit

[deleted-user-I64nRgrUqUEC]

We are running ArcSDE 9.3.1 and are currently using the se_toolkit-9.1. But recently we have been trying to upgrade to the newer version of the toolkit and I downloaded the setk101b11_LIN64_931sp0.tgz build. But when I tried to run the asc2sde utility I got the following error:

error while loading shared libraries: libsde.so: Wrong ELF class: ELFCLASS32

And the linux server version we are running is:
SUSE Linux Enterprise Server 10 (x86_64)

Any help will be appreciated.

Thanks,
Mahesh

[VinceAngelo]

You downloaded the 64-bit executable.  It will only work with a 64-bit ArcSDE install.
What does 'file $SDEHOME/lib/libsde.so' report?

There isn't any reason for you to be using ArcSDE 9.3.1 (Final), and every reason to
be using the terminal service pack (SP2).  If you don't keep current on your software,
you don't get the benefits of ongoing development.  When se_toolkit-10.2 comes
out, 9.3.1 support will be limited to SP2.

- V

[deleted-user-I64nRgrUqUEC]

Thanks Vince for the prompt response. You were right, we have the 32 bit installed on our 64 bit OS and that was creating the issues. Once I downloaded the setk101b11_LIN32_931sp0.tgz it worked. But the reason we were planning to upgrade to the new tool is because the records containing the special characters gets rejected and we get the following error:

Invalid UTF-8 formatted string
Unable to parse column 15 (INSD_NAME) -- row rejected

Some of the  special characters in the name field are like �?,ì,û,§,®�? etc. So all we are trying is to get these records loaded and we were hoping that the new version of the toolkit would load it but we get the same error. We are using the the following command in the bin directory of the toolkit to load the records:

./asc2sde -o create -f /datafiles/Policy.txt -C /control/PolicyControl.ctl -l POLICY_FILE,shape -k DEFAULTS -u username -p password

Thanks,
Mahesh

[VinceAngelo]

The Dynamic ASCII Table (DAT) module on which 'ascinfo' and 'asc2sde' are built are
"UTF-8 aware".  The STRING parser checks for illegal UTF-8 encoding.

You didn't include the content of your 'PolicyControl.ctl' or a reject file from the load,
so I can't make a specific recommendation, but you're probably going to need to use
the codepage-aware XASCII parser.  You'll need to figure out what codepage is in use
(se_tools/xascii.c currently only supports cp437, cp850, cp858, & cp1252) with a
syntax of:

#colname       XASCII(codepage)    -   width   Boolean
insd_name      XASCII(1250)        -   40      N

- V

[deleted-user-I64nRgrUqUEC]

It wouldn't let me upload the .ctl file so I put the contents in a text file uploaded it. I have also included few records that were rejected in the loading process. Also when you mention that I will have to use the codepage-aware XASCII parser, is it something I have toconfigure in the tool kit?

Thanks,
Mahesh

[VinceAngelo]

The data you provided didn't correlate to the control file.

Try changing the "STRING" in the rejected column to "XASCII(1250)", and look to see if
that results in correct data.  If it doesn't, try the other supported codepages.  If it still
doesn't work, you'll need to convert the file to UTF-8 some other way. 

Note: Fixed formatting and UTF-8 are very nearly incompatible, so your conversion
process will likely need to add delimiters to make the output parsable.

- V

[deleted-user-I64nRgrUqUEC]

Changing the "STRING" in the rejected column to "XASCII(1250)" didn't work. It gave me the following error:

Column 'INSD_NAME' refernces unknow codepage '1250'.

Also how can I find out the other supported codepages?

Thanks,
Mahesh

[VinceAngelo]

Doh! 1252, not 1250 (I gave the list, and the file it came from two posts ago).

- V