fix(plugin-oracle): read NCHAR columns under a UTF8 national charset

[datlechin]

Why

On an Oracle database whose national character set is UTF8, any query returning an NCHAR, NVARCHAR2 or NCLOB value failed with "The Oracle driver reported an error (nationalCharsetNotSupported)", and the next statement on that connection failed as "connection closed". oracle-nio only handled AL16UTF16. UTF8 values arrive as CESU-8 (measured on Oracle 23ai).

Change

• Re-pins oracle-nio to fix: decode values of a UTF8 national character set oracle-nio#14, which converts UTF8 national character set values to UTF-16 while reading rows.
• A national character set rejection now resets the connection, so the next statement redials instead of hitting a decoder that can no longer read.
• scripts/probes/check-oracle-national-charset.sh checks NCHAR reads and writes against a live database.

Verified

Fork unit tests cover CESU-8 and row framing. The probe passes on a UTF8 and an AL16UTF16 database: Lao text, U+1D11E, a 20,008-character NCLOB, literal and bind writes.

Not in this PR

• Needs fix: decode values of a UTF8 national character set oracle-nio#14 merged with a merge commit, then a plugin-oracle-v* release.
• On a database whose own character set is not Unicode, grid saves and filters send NCHAR-column text through that character set, so Lao is stored as ¿¿¿ with no error (measured on WE8MSWIN1252). Separate fix: UNISTR literals or NCHAR binds.

[mintlify]

Preview deployment for your docs. Learn more about Mintlify Previews.

Project	Status	Preview	Updated
TablePro	🟢 Ready	View Preview	Oct 8, 2026, 3:21 AM

💡 Tip: Enable Automations to automatically generate PRs for you.