Re: Selecting a constant question

Поиск
Список
Период
Сортировка
От Dann Corbit
Тема Re: Selecting a constant question
Дата
Msg-id D425483C2C5C9F49B5B7A41F894415470100072E@postal.corporate.connx.com
обсуждение исходный текст
Ответ на Re: Selecting a constant question  ("Larry McGhaw" <lmcghaw@connx.com>)
Список pgsql-hackers
> -----Original Message-----
> From: Larry McGhaw
> Sent: Tuesday, June 12, 2007 1:40 PM
> To: Martijn van Oosterhout
> Cc: Andrew Dunstan; Hannu Krosing; Tom Lane; Alvaro Herrera; Dann
Corbit;
> Gregory Stark; pgsql-hackers@postgresql.org
> Subject: RE: [HACKERS] Selecting a constant question
>
> For what its worth .. Your statement about why we are the first people
to
> mention this problem really got me thinking.  Anyone who would attempt
to
> write an ODBC driver for Postgres would run into the exact same issue.
> So I installed the official Postgres ODBC driver, and ran the
identical
> query and here are my results:
>
> I probably should have looked at this first .... There is a whole
Postgres
> ODBC dialog dedicated to the very subject of this thread:
> Handling of "unknown" data sizes.   The pgodbc driver is configured to
> treat unknowns as varchar(255) by default,
> As shown by my example below.  This can be configured up or down as
> desired.
>
> SQLExecDirect:
> In: hstmt = 0x003C18E0, szSqlStr = "Select a,b,c, '123' ,
'123'::char(3),
> '123'::varchar(3) from...", cbSqlStr = -3
> Return:    SQL_SUCCESS=0
>
> Describe Column All:
> icol, szColName, *pcbColName, *pfSqlType, *pcbColDef, *pibScale,
> *pfNullable
> 1, a, 1, SQL_VARCHAR=12, 20, 0, SQL_NULLABLE=1
> 2, b, 1, SQL_CHAR=1, 10, 0, SQL_NULLABLE=1
> 3, c, 1, SQL_INTEGER=4, 10, 0, SQL_NULLABLE=1
> 4, ?column?, 8, SQL_VARCHAR=12, 255, 0, SQL_NULLABLE=1
> 5, bpchar, 6, SQL_CHAR=1, 3, 0, SQL_NULLABLE=1
> 6, varchar, 7, SQL_VARCHAR=12, 255, 0, SQL_NULLABLE=1
>
> From psqlodbc.h
>
> #define MAX_VARCHAR_SIZE        255    /* default maximum size
of
>                          * varchar fields (not
including null
> term) */
>
> So I guess the bottom line is that we are not the first to encounter
this
> problem .. Its just been covered up by assigning
> An arbitrary maximum size .. So I guess we will do the same and make
it
> configurable like the official postgres driver.

Of course, the downside here is that choosing a default will truncate
the data when the actual data is larger than the default chosen.



В списке pgsql-hackers по дате отправления:

Предыдущее
От: "Dann Corbit"
Дата:
Сообщение: Selecting a constant question: A summary
Следующее
От: "Joshua D. Drake"
Дата:
Сообщение: Re: Selecting a constant question: A summary