RE: column_query buffer in PGAPI ColumnPrivileges

From: "Wal, Jan Tjalling van der" <jan_tjalling(dot)vanderwal(at)wur(dot)nl>
To: foxi_yiyi12081003 <foxi_yiyi12081003(at)outlook(dot)com>, "pgsql-odbc(at)postgresql(dot)org" <pgsql-odbc(at)postgresql(dot)org>
Subject: RE: column_query buffer in PGAPI ColumnPrivileges
Date: 2022-11-09 21:45:16
Message-ID: AM0PR01MB563486B9018D8DD263AF1044DD3E9@AM0PR01MB5634.eurprd01.prod.exchangelabs.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-odbc

Dear foxi_yiyi12081003,

I’m not an expert on the inner workings of this specific driver, but in my opinion this is probably correct.
The query that is defined inside appendPQExpBufferStr, asks for data from column table_schema to be returned using a different name: TABLE_SCHEM.
When the results of that query are used further down, the correct name to use them will be that new name.

When the query is run against an running instance of a postgres-database it gives results (over 9000), here limited to just 5.
select '' as TABLE_CAT, table_schema as TABLE_SCHEM,
table_name, column_name, grantor, grantee,
privilege_type as PRIVILEGE, is_grantable from
information_schema.column_privileges where true limit 5;
"table_cat"
"table_schem"
"table_name"
"column_name"
"grantor"
"grantee"
"privilege"
"is_grantable"

"information_schema"
"routines"
"scope_schema"
"xxxxxxxxxxxxx"
"xxxxxxxxxxxxx"
"UPDATE"
"YES"

"information_schema"
"routines"
"dtd_identifier"
"xxxxxxxxxxxxx"
"xxxxxxxxxxxxx"
"INSERT"
"YES"

"pg_catalog"
"pg_stat_progress_vacuum"
"datid"
"xxxxxxxxxxxxx"
"xxxxxxxxxxxxx"
"SELECT"
"YES"

"information_schema"
"role_udt_grants"
"grantor"
"xxxxxxxxxxxxx"
"xxxxxxxxxxxxx"
"SELECT"
"NO"

"pg_catalog"
"pg_namespace"
"nspname"
"xxxxxxxxxxxxx"
"xxxxxxxxxxxxx"
"SELECT"
"YES"

Kind regards JT

From: foxi_yiyi12081003 <foxi_yiyi12081003(at)outlook(dot)com>
Sent: 09 November 2022 04:18
To: pgsql-odbc(at)postgresql(dot)org
Subject: column_query buffer in PGAPI ColumnPrivileges

Hi,
Is that a bug in psqlodbc-13.02.0000 release version ?
file: info.c : 3734 ?
the SQL Command in the second appendPQExpBuffers
【 and table_schem %s'%s' 】table_schem or table_schema ??

code:
appendPQExpBufferStr(&column_query, "select '' as TABLE_CAT, table_schema as TABLE_SCHEM,"
" table_name, column_name, grantor, grantee,"
" privilege_type as PRIVILEGE, is_grantable from"
" information_schema.column_privileges where true");
op_string = gen_opestr(like_or_eq, conn);
eq_string = gen_opestr(eqop, conn);
if (escSchemaName)
appendPQExpBuffer(&column_query, " and table_schem %s'%s'", eq_string, escSchemaName);
if (escTableName)
appendPQExpBuffer(&column_query, " and table_name %s'%s'", eq_string, escTableName);
if (escColumnName)
appendPQExpBuffer(&column_query, " and column_name %s'%s'", op_string, escColumnName);
if (PQExpBufferDataBroken(column_query))

and I also found the same condition in master branch。
[cid:image001(dot)png(at)01D8F48C(dot)1DC554E0]

In response to

Responses

Browse pgsql-odbc by date

  From Date Subject
Next Message Chiang Chan-i 2022-11-10 01:36:44 RE: column_query buffer in PGAPI ColumnPrivileges
Previous Message foxi_yiyi12081003 2022-11-09 03:18:10 column_query buffer in PGAPI ColumnPrivileges