Error Code page conversion table for ISO8859-1 to SHIFT-JIS was not found in convmap.cp. (6063)

qadvietnam

Member
Hi all,

I am facing problem issue when I try to convert DB from ISO8859-1 to SHIFT-JIS
All the steps I did:

0. Proshut db
1. proutil c:\mfgsvr\db\mfgempty -C convchar convert undefined
2. proutil c:\mfgsvr\db\mfgempty -C convchar convert SHIFT-JIS
3. proutil c:\mfgsvr\db\mfgempty -C idxbuild all
4. proutil c:\mfgsvr\db\mfgempty -cpinternal SHIFT-JIS -C idxbuild all
5. Truncate DB
6. Startup db

and then I try to run character client, I got error msg

Code page conversion table for ISO8859-1 to SHIFT-JIS was not found in convmap.cp. (6063)

How can I solve it?
Thanks
 

TomBascom

Curmudgeon
Which part of this answer, courtesy of RealHeavyDude, from your other thread is unclear?

The two code pages you mention are not compatible. There is no painless,
lossless conversion between them two possible. Full stop.
...
Forgot to mention: Although most people I know who work in IT are
completely oblivious to what code pages and character sets are, they are
fundamental in any string processing and therefore I consider it basic IT
knowledge. If you would like to know more then I recommend you reading this
for a starter: http://en.wikipedia.org/wiki/Code_pages
 

qadvietnam

Member
Hi,

- sorry i do not think it related with the old topic.
- others, make sure before I can do that with the same issue in progress 10.0B05 on RH4 Ent. But now I can not do that on Windows OS with OE10.2B?
- The quote as you post I am sure I do not understand, can you explain for me clear?

Thnks
 

RealHeavyDude

Well-Known Member
Just to be clear: When you move a database to another platform then you need to create a new database on the target platform. Exactly at that point you decide which code page the database will have. You make that decision by chosing an empty database from which you take a copy. You will find a prolang subdirectory which contains subdirectories for the languages that are available which has been decided when Progress was installed. All of these language subdirectory will contain copies of the different empty database ( for different blocksizes respectively ).

Nevertheless, you can change the code page of the database later on but there restrictions: As soon as there is data in the database which contains characters that are not compatible with the target code pages ( which means the character set does not contain them ) they will either be lost or screwed up. There is a utility that lets you identify the records which contain such characters ( have a look a the proutil <dbname> -C convchar command ).

If I undstand you correct then somebody has already created a database on the target system for you. That somebody obviously did not know that she/he should have created that database with the correct code page or did not think about code pages at all. That might be the reason why the code page of the target database is iso8859-1 which is the default American / Western European code page.

Therefore I would recommend you to urge whoever is in charge of that database that she/he corrects her/his mistake or converts the database to the code page you need or UTF-8 - which can serve data from most code pages as it contains all characters from most code pages. But, since that is a least common denominator there might be other implications that won't suit with your application.

Not only is the code page of the database relevant, but also the code page which the client uses. They must be compatible too.

Plus, you might also want to have a look at the collation which is relevant for the sort order. Different code pages usually have different sort orders - for example in which order A and the German Umlaut Ä are sorted. In some code pages their weight is equal in others not. Therefore, most likely, not only the code page iso8859-1 of the target database does not fit your needs, the sort order will also be different from what you have now.

Another option - but I have never done that plus I don't know anybody who ever did that - would be that you create your own conversion table between the code pages so that Progress knows what it should do. You could roll your own and extend the convmap.cp with it.

Heavy Regards, RealHeavyDude.
 

qadvietnam

Member
hi,
I have a look on KB here https://knowledgebase.qad.com/kmp/article/AA-27849

How do you think about this issue?
Thanks

Code page conversion table for ISO8859-1 to SHIFT-JIS was not found in convmap.cp. (6063)
Author: Edward Grize Article ID: AA-27849 Views: 153 Created: 11-15-2004 05:30 PM Last Updated: 01-31-2008 03:09 PM0 Rating/ Voters
ico-rating-g.gif
ico-rating-g.gif
ico-rating-g.gif
ico-rating-g.gif
ico-rating-g.gif

fact:

Double Byte Enabled



fact: Progress 9

fact: MFG/PRO eB

fact: MFG/PRO eB2

fact: code page

fact: convmap.cp

symptom: Code page conversion table for ISO8859-1 to SHIFT-JIS was not found in convmap.cp. (6063)

symptom: Progress error (6063)

symptom: Convert English to Shift-JIS

symptom: How to convert to DBE database

cause: You can NOT place double-byte (such as Japanese) data into a single-byte (English language) database.However, you CAN place single-byte data, such as English characters, into a double-byte database.

fix:


For Example : Converting a US codepage to Japanese.


Using Progress Proutil


1) Convert code-page to double byte code-page (For example: from ISO8859-1 to SHIFT-JIS)

$DLC/bin/proutil <db> -C convchar convert undefined

$DLC/bin/proutil <db> -C convchar convert SHIFT-JIS


Once you have converted from undefined to SHIFT-JIS codepage, make sure for all steps that you are using -cpinternal SHIFT-JIS.If -cpinternal is any value other than SHIFT-JIS, then the MFG/PRO clients are not using the proper double-byte aware algorithms.


2) Load collation table.This sets the sorting sequence for the proper code-page.

Load $DLC/prolang/jpn/_trans.df into database.

3) Rebuild all database indexes

$DLC/bin/proutil <db> -C idxbuild all

4) Also need to specify code-page parameters (For Example: -cpstream SHIFT-JIS -cpinternal SHIFT-JIS -cprcodein SHIFT-JIS -cpcoll basic) in the MFG/PRO server and client startup scripts.

The database language should be jpn (Japanese) now.

Article Properties
Article Type:Functional
Review Status:Verified
Database: Progress
Original Solution ID:qad32287
 
Top