Encoding problem with Oracle 12c after update to 2.2 EF core

[SergeyPechenik]

We updated from EF Core 2.1 with Devart.Data.Oracle.EFCore 9.6.621
to EF Core 2.2 with Devart.Data.Oracle.EFCore 9.6.696

But we get strange strings now. It's looks like that:
"\0L\0o\0r\0e\0m\0 \0i\0p\0s\0u\0m\0 \0d\0o\0l\0o\0r\0 \0s\0i\0t\0 \0a\0m\0e\0t\0"
and like that:
"\u0004#\0 \u0004?\u0004>\u0004?\u00040\0 \u00041\u0004K\u0004;\u00040\0 \u0004A\u0004>\u00041\u00040\u0004:\u00040\0 \u0004>\u0004=\0 \u00045\u0004Q\0"

Previous version was worked fine. 'Unicode=true' and 'Unicode=false' connection string parameter, does not make any effect.

Oracle server: Oracle Database 12c Enterprise Edition Release 12.2.0.1.0 - 64bit Production
Collation:
NLS_SORT BINARY
NLS_LANGUAGE AMERICAN
NLS_TERRITORY AMERICA
NLS_CHARACTERSET AL32UTF8

How to solve this problem?

P.S. Columns type NVARCHAR2

[Shalex]

We cannot reproduce the issue in our environment.

Please specify:

1) your connection string (mark confidential information with asterisks)

2) does the issue persist with reading the existing records or with inserting the new records?

3) give us an example of the original string that got corrupted after the upgrade

[SergeyPechenik]

Sorry, for long response.

1) Data Source=(DESCRIPTION=(ADDRESS_LIST=(ADDRESS=(PROTOCOL=TCP)(HOST=***)(PORT=***)))(CONNECT_DATA=(SERVER=DEDICATED)(SERVICE_NAME=***)));User Id=***;Password=***;Direct=true;license key=***

2) the issue persist with reading the existing records and with inserting the new records;

when inserting string with dotConnect: Encoding problem with Oracle 12c after update to 2.2 EF core
in dbForge i see: 䕮捯摩湧⁰牯扬敭⁷楴栠佲慣汥‱㉣⁡晴敲⁵灤慴攠瑯′⸲⁅䘠捯牥
But i can read this string in dot connect fine!

if i insert string in dbforge, cant read it in dotConnect.

3) Original string: Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.

readed string: "\0L\0o\0r\0e\0m\0 \0i\0p\0s\0u\0m\0 \0d\0o\0l\0o\0r\0 \0s\0i\0t\0 \0a\0m\0e\0t\0,\0 \0c\0o\0n\0s\0e\0c\0t\0e\0t\0u\0r\0 \0a\0d\0i\0p\0i\0s\0c\0i\0n\0g\0 \0e\0l\0i\0t\0,\0 \0s\0e\0d\0 \0d\0o\0 \0e\0i\0u\0s\0m\0o\0d\0 \0t\0e\0m\0p\0o\0r\0 \0i\0n\0c\0i\0d\0i\0d\0u\0n\0t\0 \0u\0t\0 \0l\0a\0b\0o\0r\0e\0 \0e\0t\0 \0d\0o\0l\0o\0r\0e\0 \0m\0a\0g\0n\0a\0 \0a\0l\0i\0q\0u\0a\0.\0 \0U\0t\0 \0e\0n\0i\0m\0 \0a\0d\0 \0m\0i\0n\0i\0m\0 \0v\0e\0n\0i\0a\0m\0,\0 \0q\0u\0i\0s\0 \0n\0o\0s\0t\0r\0u\0d\0 \0e\0x\0e\0r\0c\0i\0t\0a\0t\0i\0o\0n\0 \0u\0l\0l\0a\0m\0c\0o\0 \0l\0a\0b\0o\0r\0i\0s\0 \0n\0i\0s\0i\0 \0u\0t\0 \0a\0l\0i\0q\0u\0i\0p\0 \0e\0x\0 \0e\0a\0 \0c\0o\0m\0m\0o\0d\0o\0 \0c\0o\0n\0s\0e\0q\0u\0a\0t\0.\0 \0D\0u\0i\0s\0 \0a\0u\0t\0e\0 \0i\0r\0u\0r\0e\0 \0d\0o\0l\0o\0r\0 \0i\0n\0 \0r\0e\0p\0r\0e\0h\0e\0n\0d\0e\0r\0i\0t\0 \0i\0n\0 \0v\0o\0l\0u\0p\0t\0a\0t\0e\0 \0v\0e\0l\0i\0t\0 \0e\0s\0s\0e\0 \0c\0i\0l\0l\0u\0m\0 \0d\0o\0l\0o\0r\0e\0 \0e\0u\0 \0f\0u\0g\0i\0a\0t\0 \0n\0u\0l\0l\0a\0 \0p\0a\0r\0i\0a\0t\0u\0r\0.\0 \0E\0x\0c\0e\0p\0t\0e\0u\0r\0 \0s\0i\0n\0t\0 \0o\0c\0c\0a\0e\0c\0a\0t\0 \0c\0u\0p\0i\0d\0a\0t\0a\0t\0 \0n\0o\0n\0 \0p\0r\0o\0i\0d\0e\0n\0t\0,\0 \0s\0u\0n\0t\0 \0i\0n\0 \0c\0u\0l\0p\0a\0 \0q\0u\0i\0 \0o\0f\0f\0i\0c\0i\0a\0 \0d\0e\0s\0e\0r\0u\0n\0t\0 \0m\0o\0l\0l\0i\0t\0 \0a\0n\0i\0m\0 \0i\0d\0 \0e\0s\0t\0 \0l\0a\0b\0o\0r\0u\0m\0."

Another one:
Was: У попа была собака он её любил, она съела кусок мяса он её убил, закопал её в могилу на могиле написал, у попа была собака он её любил...
Became: "\u0004#\0 \u0004?\u0004>\u0004?\u00040\0 \u00041\u0004K\u0004;\u00040\0 \u0004A\u0004>\u00041\u00040\u0004:\u00040\0 \u0004>\u0004=\0 \u00045\u0004Q\0 \u0004;\u0004N\u00041\u00048\u0004;\0,\0 \u0004>\u0004=\u00040\0 \u0004A\u0004J\u00045\u0004;\u00040\0 \u0004:\u0004C\u0004A\u0004>\u0004:\0 \u0004<\u0004O\u0004A\u00040\0 \u0004>\u0004=\0 \u00045\u0004Q\0 \u0004C\u00041\u00048\u0004;\0,\0 \u00047\u00040\u0004:\u0004>\u0004?\u00040\u0004;\0 \u00045\u0004Q\0 \u00042\0 \u0004<\u0004>\u00043\u00048\u0004;\u0004C\0 \u0004=\u00040\0 \u0004<\u0004>\u00043\u00048\u0004;\u00045\0 \u0004=\u00040\u0004?\u00048\u0004A\u00040\u0004;\0,\0 \u0004C\0 \u0004?\u0004>\u0004?\u00040\0 \u00041\u0004K\u0004;\u00040\0 \u0004A\u0004>\u00041\u00040\u0004:\u00040\0 \u0004>\u0004=\0 \u00045\u0004Q\0 \u0004;\u0004N\u00041\u00048\u0004;\0.\0.\0."

[SergeyPechenik]

One more thing!

I just updated DbForge and now all strings looks the same like in new dotConnect, but in SQL Developer strings looks like it was in older version of dbForge.

[Shalex]

I just updated DbForge and now all strings looks the same like in new dotConnect, but in SQL Developer strings looks like it was in older version of dbForge.

1. You mean that now the same problem persists with both dotConnect for Oracle and dbForge Studio for Oracle products, don't you?

2. Could you try connecting with dotConnect for Oracle v9.6.725 in the OCI mode (via Oracle Client)? Does the problem occur in this case as well?

3. Specify your value of NLS_NCHAR_CHARACTERSET.

[SergeyPechenik]

1) Yes, Exactly.

2) In OCI Mode everything fine

3) NLS_NCHAR_CHARACTERSET = UTF8

[Shalex]

We are investigating the issue.

[TheGramix]

Hello,
any news about this issue? I have the same problem on my oracle 12c :-(

[Shalex]

We have reproduced the bug with Oracle 12c (NLS_NCHAR_CHARACTERSET=UTF8). It occurs when the value is passed as a plain text to NVARCHAR2 column, in case of using a parameter it works. We will notify you when the issue is fixed.

any news about this issue? I have the same problem on my oracle 12c :-(

Please specify the full version, edition of your Oracle 12c and its value of NLS_NCHAR_CHARACTERSET.

[TheGramix]

Hello,
my Oracle is 12.1.0.2.0 and NLS_NCHAR_CHARACTERSET is UTF8.
I tried to use your Entity Developer 6.3.631 to create my model, I don't use any parameter at the moment, just a "Retrieve data" from a table with NVARCHAR2 columns.

Many thanks!

[Shalex]

The investigation is in progress. Once the issue is fixed, we will notify you.

[seb1487]

Hi,
i have the same issue. The Bug exists since version 9.6.646 and only with NLS_NCHAR_CHARACTERSET=UTF8.

Greetings,
seb1487

[Shalex]

* The bug with inserting a long string into the VARCHAR2 column on Oracle Server with NLS_CHARACTERSET=UTF8 in the Direct mode is fixed
* The bug with reading data from the NVARCHAR2 column on Oracle Server with NLS_NCHAR_CHARACTERSET=UTF8 in the Direct mode is fixed

The internal build with the fix is available at https://www.devart.com/pub/nuget_oracle_9_7_756.zip.

We will notify you when the new public build is available for download.

[Shalex]

New build of dotConnect for Oracle 9.7.770 is available for download now: viewtopic.php?f=1&t=38884.