firebird-automations commented 13 years ago

Submitted by: Thomas Steinmaurer (tsteinmaurer)

Hello,

in a Firebird 2.5 database created with ISO8859_1 and also using ISO8859_1 at connect time, I get the following in fsql:

SQL> select 'ä' from rdb$database;

CONSTANT

ä

But the raw output of a trace session gives me:

2010-11-19T11:02:11.6010 (4588:00000000011ABAE0) START_TRANSACTION tourism.fdb (ATT_1437, TOURISM:NONE, ISO88591, TCPv4:127.0.0.1) C:\Program Files\Firebird\Firebird_250_3051\bin\fsql.exe:5280 (TRA_24528, CONCURRENCY | WAIT | READ_WRITE)

2010-11-19T11:02:11.6020 (4588:00000000011ABAE0) EXECUTE_STATEMENT_START tourism.fdb (ATT_1437, TOURISM:NONE, ISO88591, TCPv4:127.0.0.1) C:\Program Files\Firebird\Firebird_250_3051\bin\fsql.exe:5280 (TRA_24528, CONCURRENCY | WAIT | READ_WRITE)

Statement 1906: ------------------------------------------------------------------------------- select '„' from rdb$database ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ PLAN (RDB$DATABASE NATURAL)

I don't see the 'ä' correctly. So, are there any problems dealing with special (Non-Ansi?) characters in the raw trace data output?

firebird-automations commented 13 years ago

Modified by: @dyemanov

assignee: Vlad Khorsun \[ hvlad \]

firebird-automations commented 13 years ago

Commented by: @hvlad

What is real encoding used in fsql session ? Is it used OEM codepage as usual console application ?

Trace does not perform conversion of characters of users strings (queries, parameters, etc). So, in general, yes, we have a problem - trace output with data from different users connections could contain strings in different encodings.

We can convert all this pieces into UTF8, i think, and have consistent trace output, but it will cost us some performance penalty.

FirebirdSQL / firebird

Garbage with e.g. umlauts in the raw trace output [CORE3280] #3648

CONSTANT