Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - daodug

Pages: [1]
1
Hi Justin,

I created the database following https://www.ibm.com/support/pages/creating-remote-microsoft-sql-server-database-ibm%C2%AE-content-manager-ondemand.

There's no mention of the collation, so as the install worked for double byte with FileNet, I left it alone.  The original collation was Latin1_General_CI_AS.  That didn't work.
I created a new DB with SQL_Latin1_General_CP1_CI_AS.  I found a reference (see PDF attachment https://www.ibm.com/support/pages/node/616131) that showed that as a selection on install.  It doesn't work for me.

The documentation https://www.ibm.com/support/knowledgecenter/SSEPCD_10.5.0/com.ibm.ondemand.installmp.doc/dodnl005.htm doesn't provide absolute facts.  It describes vague information and then more vague examples.  It describes that if you had this value then some other thing would be required, but doesn't provide either guidance on where the value is stored, how it is changed or what the other thing that might be required actually is, or what the possible values are.  Very obtuse, poorly constructed, incomplete and confusing documentation.

With SQL Server there is no 'code page' thing.  There's a collation.  There's no documentation I can find that specifies what the collation should be for any given OnDemand configuration.  My SQL Server install doesn't support UTF-8, without me upgrading, which I really don't want to do, yet, but I don't know if it's using UTF-8 either. Latin characters are fine, double byte ones are not.  How do you tell  it to do something different?  The documentation references ARS_ORIGINAL_CODEPAGE but wanders around from one release to another with possible indeterminate values.  It looks like for 10.5 a value of 0 is required.  There's no documentation that describes what value 0 means, or what other values are possible or what they do.

I've been down several rabbit holes already, rebuilding the DB, hacking columns to NVARCHAR, and without some absolute definition I'm just wasting my time.

There's no definition of what OnDemand is doing when it inserts data from the admin tool into the DB. How do you configure OnDemand to use a specific codepage?  Do I need to? How does that map to any concrete value in SQLServer?  It should be a trivial thing and it should be the same regardless of it being a remote SQL, so I'm hoping that someone that uses DBCS has some guidance on what the configuration values that work are.

Thanks,
David


2
Hi,

I'm trying to set up some translations of folder names, field names and field values on a 10.5 system with Remote SQL Server and I'm having issues with ?? values being stored in the database and displayed in the admin UI after entering valid  names in non-latin languages (eg Arabic, Japanese, Chinese, Korean).

The documentation is verbose yet doesn't explain what the system is doing.  As it's a remote SQL server, the documentation is sparse in describing the installation and any dependencies.  The database server is an existing one (SQL Server 2016) that works fine with other products (eg IBM FileNet) in correctly storing and retrieving double byte characters.  The database system cannot be changed easily, so before I do anything dramatic I'd like to understand what the system is doing, what else I may have missed, and if anyone else is already using these translation features in the administration tool.

I've tried different database collations.  One document I found showed SQL_Latin1_General_CP1_CI_AS being used.  I can't find any documentation that describes what the collation for SQL Server should be.  The other databases on the same server that work with DBCS are using Latin1_General_CI_AS.  That same collation also doesn't work for OnDemand.  The same behaviour occurs, the data can be entered and displays correctly, when it's saved to the DB it's shown as ??, when it's retrieved in the UI, it's a series of ??.  My system is in Sydney, so the base locale info is en-au.  All of the latin languages (eg spanish, french, german) work.

I've discussed with a colleague the same actions on an OnDemand with DB2.  They work fine.  It seems to be some additional or different settings needed for SQL Server.

With all that background:
1. Has anyone got an English based OnDemand system using SQL Server that has any double byte character translations for Cabinet, Folder, Hold, Field or Field Value entries that persists the data correctly?  I'm using 10.5 but any indication on any version will give me more of an idea what the issue might be.
2. There are lots of pieces of documentation around code pages but no sensible definition of where and how they're used, exactly.  There's also lots of references to configuration files that don't exist anymore in the documentation.  Without some solid documentation I'm not ready to hack too much more.  Is this where I'm going wrong?

I think it's code page related, but the doco is truly awful.  There's no complete discussion that covers my corner case and I'm not ready to waste more time guessing.

Any guidance is greatly appreciated.

David

3
Report Indexing / Re: Report not being broken into multiple documents
« on: July 15, 2014, 04:49:30 PM »
Thanks, but I don't understand why I need to use the asciinp.  The index values are being generated correctly and the system can read and interpret the input stream.  Each of my indexes is being located and correctly put into the database.  The problem is that each index gets the entire document added rather than just the pages of the document that relates to that index.  From my reading, the load/index process should identify the triggers/indexes and break up the document based on those settings.  Why would I need to change the input data for it to do that, when it's already identfying the data correctly?

4
Report Indexing / Report not being broken into multiple documents
« on: July 14, 2014, 02:21:00 PM »
I've got a set of customer statements in a single file which is being indexed correctly (?) but isn't being broken into different documents.  It's a small 36 pages which should break into around 25 individual statements.  I've got 25 individual indexes but each index retrieves the entire 36 pages.  I can't seem to change any of the break/group settings in the index definition and everything is set to break (see below).  I've only been working with the product (9.0.3 with content navigator) for a few weeks so any help greatly appreciated.

CC=NO
CCTYPE=Z
CONVERT=NO
CPGID=850
MCF2REF=CPCS
NEWPAGE=12
TRC=NO
FILEFORMAT=STREAM,(NEWLINE=X'0D0A')
TRIGGER1=*,96,X'312020',(TYPE=GROUP)                                     /* 1               */
FIELD1=-9,51,61,(TRIGGER=1,BASE=0)
FIELD2=-6,45,40,(TRIGGER=1,BASE=0)
FIELD3=-6,86,18,(TRIGGER=1,BASE=0)
FIELD4=-3,86,11,(TRIGGER=1,BASE=0)
FIELD5=-3,106,11,(TRIGGER=1,BASE=0)
FIELD6=0,87,4,(TRIGGER=1,BASE=0)
INDEX1=X'4163636F756E7454797065',FIELD1,(TYPE=GROUP,BREAK=YES)           /* AccountType     */
INDEX2=X'4163636F756E744E756D626572',FIELD3,(TYPE=GROUP,BREAK=YES)       /* AccountNumber   */
INDEX3=X'437573746F6D65724E616D65',FIELD2,(TYPE=GROUP,BREAK=YES)         /* CustomerName    */
INDEX4=X'53746174656D656E744E756D626572',FIELD6,(TYPE=GROUP,BREAK=YES)   /* StatementNumber */
INDEX5=X'4461746546726F6D',FIELD4,(TYPE=GROUP,BREAK=NO)                  /* DateFrom        */
INDEX6=X'44617465546F',FIELD5,(TYPE=GROUP,BREAK=NO)                      /* DateTo          */
DCFPAGENAMES=NO
UNIQUEBNGS=YES
IMAGEOUT=ASIS
INDEXOBJ=GROUP
INDEXSTARTBY=1
INSERTIMM=NO
RESTYPE=NONE

Pages: [1]