When exporting Japanese DBCS (2 byte code) result set using IBM Data Virtualization Manager (DVM) for z/OS Studio to CSV format, the output appears unreadable.

Resolving The Problem

• In DVM studio, go to Window->Preferences->Data Virtualization Manager.
• Change Hex Encoding as windows-31j.
• Change File Encoding as x-UTF-16LE-BOM.
• Apply and close as shown below:

File Encoding: x‑UTF‑16LE‑BOM and Hex Encoding: windows‑31j defines how DVM writes characters into the output file (e.g., CSV, TXT). This encoding is often required by Excel or other Windows tools to correctly interpret Unicode data. When exporting to CSV with x‑UTF‑16LE‑BOM, Excel can properly read multilingual data (Japanese, Chinese, Hindi, etc.) without garbling. In contrast, windows‑31j is Microsoft’s extension of Shift_JIS (a Japanese character encoding). Windows applications (including Excel in Japanese locales) expect this format when reading Shift_JIS data.

Data containing Japanese text from DB2 will be correctly interpreted using windows‑31j. The final CSV is saved in UTF‑16LE‑BOM, ensuring that Excel can open it without corrupted or garbled characters.

Note :  It is required to switch the Windows system to Japanese encoding so that it supports Japanese characters and correctly displays DBCS characters.