Peter Eisentraut [Fri, 4 Nov 2005 10:50:52 +0000 (10:50 +0000)]
Remove wildly outdated Debian and Red Hat HOWTOs.
Dave Page [Thu, 3 Nov 2005 22:31:16 +0000 (22:31 +0000)]
Fix declare/fetch memory leak. Now we get a nice flat line :-)
Dave Page [Thu, 3 Nov 2005 22:07:17 +0000 (22:07 +0000)]
Add a VC++ project file.
Dave Page [Thu, 3 Nov 2005 13:25:05 +0000 (13:25 +0000)]
Remove (libpq) suffix to the driver name on dialogues etc. Use nicer font on English dialogues per grizzle from Magnus :-).
Dave Page [Thu, 3 Nov 2005 13:11:52 +0000 (13:11 +0000)]
Remove (libpq) suffix to the driver name on dialogues etc.
Dave Page [Thu, 3 Nov 2005 12:53:09 +0000 (12:53 +0000)]
Switch off echo
Dave Page [Thu, 3 Nov 2005 12:05:13 +0000 (12:05 +0000)]
Fix a major memory leak in declare/fetch mode. Ensure that forward only cursors can only be used in a forward direction.
Dave Page [Wed, 2 Nov 2005 16:29:29 +0000 (16:29 +0000)]
Fix Declare/fetch when used with bound columns.
Dave Page [Wed, 2 Nov 2005 14:49:33 +0000 (14:49 +0000)]
Fix typo, per Johann Zuschlag
Dave Page [Wed, 2 Nov 2005 10:45:43 +0000 (10:45 +0000)]
Prevent setting a connstrout value when the max length is set to zero.
Dave Page [Mon, 31 Oct 2005 19:57:06 +0000 (19:57 +0000)]
Initial fix for Declare/Fetch option [Anoop, Sivakumar and Vaidhy @ Pervasive]
Dave Page [Tue, 25 Oct 2005 07:38:54 +0000 (07:38 +0000)]
When EN_Destructor is called it searches through the connection array
and destroys any connections which belong that environment. Unfortunately
it doesn't NULL out the connection array entry afterwards so a future
call to EN_Destructor can wind up trying to destroy that connection
again if another environment has been allocated at the same address.
Also, shouldn't EN_Destructor hold the CONNS critical section while it
is playing with the connection array?
The attached patch fixes both of these issues and seems to have fixed
the crashes I was observing.
[Tom Hughes]
Dave Page [Tue, 25 Oct 2005 07:35:19 +0000 (07:35 +0000)]
This patch fixes a potential segfault that can happen if a dbc handle is created and then destroyed without attempting a connection.
[Scot Loach]
Peter Eisentraut [Thu, 20 Oct 2005 13:06:10 +0000 (13:06 +0000)]
Massive compiler warnings cleanup.
Dave Page [Wed, 12 Oct 2005 13:22:13 +0000 (13:22 +0000)]
Use program files environment var in installer per Johann Zuschlag
Dave Page [Wed, 12 Oct 2005 09:21:08 +0000 (09:21 +0000)]
Bump version number
Dave Page [Wed, 12 Oct 2005 08:15:18 +0000 (08:15 +0000)]
Use program files environment var in Windows makefile per Johann Zuschlag
Peter Eisentraut [Wed, 12 Oct 2005 08:12:30 +0000 (08:12 +0000)]
Use pg_config to find libpq.
Peter Eisentraut [Tue, 11 Oct 2005 07:47:11 +0000 (07:47 +0000)]
Fix spelling of PostgreSQL.
Dave Page [Fri, 7 Oct 2005 09:47:36 +0000 (09:47 +0000)]
Add Kerberos support.
Update paths to use libpq from PostgreSQL 8.1 by default.
Dave Page [Fri, 7 Oct 2005 09:03:34 +0000 (09:03 +0000)]
Add OpenSSL and MIT Kerberos acknowledgements
Dave Page [Thu, 6 Oct 2005 10:01:43 +0000 (10:01 +0000)]
Create logfiles in mode 0600 on *nix
Dave Page [Wed, 5 Oct 2005 20:31:40 +0000 (20:31 +0000)]
Don't reuse the UNICODE macro
Dave Page [Tue, 4 Oct 2005 19:03:19 +0000 (19:03 +0000)]
I've attached a patch that uses the SQLSTATE from the backend any time
the error condition after executing a query is
STMT_ERROR_TAKEN_FROM_BACKEND. Besides solving the problem I reported,
this seems to make much more sense than setting the SQLSTATE arbitrarily
to HY000 "General Error" in this situation. I look forward to your
feedback.
[Chris Ingram]
Dave Page [Wed, 28 Sep 2005 08:33:24 +0000 (08:33 +0000)]
Map 'name' to SQL_VARCHAR rather than SQL_CHAR - this fixes an error with SQL Server linked servers which gets confused by the mismatch between the column size and the data size.
Dave Page [Fri, 23 Sep 2005 09:35:58 +0000 (09:35 +0000)]
Further multibyte tweaks from Hiroshi.
Dave Page [Wed, 21 Sep 2005 20:26:17 +0000 (20:26 +0000)]
Update multibyte encodings and mappings [Hiroshi Saito]
Anoop Kumar [Mon, 19 Sep 2005 13:44:09 +0000 (13:44 +0000)]
These files are no longer used by the Libpq version of psqlodbc.
Anoop Kumar [Mon, 19 Sep 2005 13:31:59 +0000 (13:31 +0000)]
Socket code is removed from the source. Now, psqlodbc driver relies completely on Libpq for the communication.
Dave Page [Sat, 17 Sep 2005 22:58:51 +0000 (22:58 +0000)]
Don't lose notices that don't come from a query.
Dave Page [Sat, 17 Sep 2005 22:27:24 +0000 (22:27 +0000)]
This patch fill fix the notice handler so that the odbc driver doesn't output notices from the backend to stderr.
[Scot Loach]
Dave Page [Sat, 17 Sep 2005 22:16:08 +0000 (22:16 +0000)]
This patch corrects the previous fix for returning 08S01 error code for the server disconnecting.
The previous patch sent this error code for any error on the statement (such as a primary key collision).
[Scot Loach]
Dave Page [Fri, 16 Sep 2005 23:12:17 +0000 (23:12 +0000)]
Remove old doc file.
Dave Page [Fri, 16 Sep 2005 23:10:28 +0000 (23:10 +0000)]
Bump version number
Dave Page [Fri, 16 Sep 2005 23:05:00 +0000 (23:05 +0000)]
Don't include win_unicode.c twice
Dave Page [Fri, 16 Sep 2005 16:01:43 +0000 (16:01 +0000)]
Backout previous rowcount patch which seems to be causing some odd errors with MS ADO.
Dave Page [Thu, 15 Sep 2005 08:16:55 +0000 (08:16 +0000)]
Fix paths for new build targets
Dave Page [Thu, 15 Sep 2005 08:09:43 +0000 (08:09 +0000)]
Allow SQLRowCount to work following SELECT queries [Scot Loach]
Dave Page [Thu, 15 Sep 2005 07:45:44 +0000 (07:45 +0000)]
Fix identifier case sensitivity [Chris Ingram]
Dave Page [Tue, 6 Sep 2005 15:19:13 +0000 (15:19 +0000)]
Explain which driver users should choose.
Dave Page [Tue, 6 Sep 2005 07:47:26 +0000 (07:47 +0000)]
Re-enable ANSI/Unicode compilation.
Dave Page [Fri, 2 Sep 2005 14:31:42 +0000 (14:31 +0000)]
Remove unnecessary query length limitation, and clean up so dodgy code.
Dave Page [Fri, 2 Sep 2005 08:40:59 +0000 (08:40 +0000)]
Add basic support for savepoints [Hiroshi Saito/Dave Page]
Dave Page [Thu, 1 Sep 2005 14:45:16 +0000 (14:45 +0000)]
Only mark the connection dead when it actually fails, not for any error.
Dave Page [Wed, 31 Aug 2005 07:31:44 +0000 (07:31 +0000)]
Please find the patch for the other bug pointed out by Scot. (Thanks to
Sivakumar)
"Another difference is the "state" that is returned from SQLGetDiagRec
when a connection is lost during a query. This used to return 08S01.
It now returns HY000."
Dave Page [Tue, 30 Aug 2005 14:57:50 +0000 (14:57 +0000)]
Bloody line ends. Grrr.
Dave Page [Tue, 30 Aug 2005 11:27:07 +0000 (11:27 +0000)]
Fix line ends.
Dave Page [Tue, 30 Aug 2005 10:28:32 +0000 (10:28 +0000)]
Test commit - ignore
Anoop Kumar [Tue, 30 Aug 2005 09:19:50 +0000 (09:19 +0000)]
Fix for making the SQLGetConnectAttr API work with SQL_ATTR_CONNECTION_DEAD attribute.
Dave Page [Tue, 30 Aug 2005 08:37:06 +0000 (08:37 +0000)]
Allow SSL modes to be translated [Hiroshi Saito]
Dave Page [Sun, 28 Aug 2005 19:43:47 +0000 (19:43 +0000)]
Don't install docs by default, as most likely only developers/sysadmins will need them.
Dave Page [Sat, 27 Aug 2005 23:19:32 +0000 (23:19 +0000)]
Update installer:
- Install in configurable directories.
- Add documentation, with a shortcut from the start menu.
- Update the banner bitmap.
Dave Page [Sat, 27 Aug 2005 22:16:46 +0000 (22:16 +0000)]
Wix got finickity about short filenames.
Dave Page [Sat, 27 Aug 2005 22:00:54 +0000 (22:00 +0000)]
Add FAQ.
Dave Page [Sat, 27 Aug 2005 20:14:55 +0000 (20:14 +0000)]
Further documentation updates
Dave Page [Sat, 27 Aug 2005 19:44:47 +0000 (19:44 +0000)]
Move docs into the source tree.
Anoop Kumar [Wed, 24 Aug 2005 06:09:52 +0000 (06:09 +0000)]
Fixed the SQLExecDirec bug which returned SQL_SUCCESS instead of SQL_NO_DATA when no row got affected in UPDATE/DELETE. (Thanks to Sivakumar)
Dave Page [Fri, 12 Aug 2005 09:47:55 +0000 (09:47 +0000)]
Fix SQLRowCount [Anoop Kumar]
Dave Page [Fri, 12 Aug 2005 09:19:42 +0000 (09:19 +0000)]
Oops, missed this commit
Dave Page [Thu, 11 Aug 2005 15:35:16 +0000 (15:35 +0000)]
Further column length tweaks
Dave Page [Thu, 11 Aug 2005 14:08:28 +0000 (14:08 +0000)]
Improve field length setup. Use the correct values for varchar and char, and SQL_NO_TOTAL for bytea and text.
Dave Page [Fri, 5 Aug 2005 15:19:45 +0000 (15:19 +0000)]
Contrary to what SQLGetInfo thinks, PostgreSQL doesn't support catalog names.
Dave Page [Fri, 5 Aug 2005 12:37:20 +0000 (12:37 +0000)]
Post release version number bump
Dave Page [Fri, 5 Aug 2005 12:19:48 +0000 (12:19 +0000)]
Fix default numeric precision.
Dave Page [Thu, 4 Aug 2005 13:39:28 +0000 (13:39 +0000)]
Fix leak in SC_Destructor [Kelly Burkhart] - archives.postgresql.org/pgsql-odbc/2005-03/msg00091.php
Dave Page [Thu, 4 Aug 2005 12:26:57 +0000 (12:26 +0000)]
If I bind a 32 bit integer using SQL_INTEGER/SQL_C_LONG and pass a value of -1, the insert fails with an 'integer out of range' error. The reason this happens is the sprintf near line 2658 of convert.c. [Kelly Burkhart]
http://archives.postgresql.org/pgsql-odbc/2005-03/msg00126.php
Dave Page [Wed, 3 Aug 2005 08:19:28 +0000 (08:19 +0000)]
Fix per bug ref 1187 [Scot Loach]
Dave Page [Wed, 3 Aug 2005 08:09:33 +0000 (08:09 +0000)]
Handle SQL_DATE from ODBC2 apps, per bug report 1292 [ alic <NOSPAM> sokrates.hr]
Dave Page [Wed, 3 Aug 2005 07:27:30 +0000 (07:27 +0000)]
Add comment about a known leak bug that needs fixing
Dave Page [Wed, 3 Aug 2005 07:26:01 +0000 (07:26 +0000)]
Fix bug that stopped Use Declare/Fetch working, and prevented ODBC bench running in multi-thread, multi-transaction mode.
Dave Page [Wed, 3 Aug 2005 07:24:49 +0000 (07:24 +0000)]
Fix some logging that was bugging me
Dave Page [Tue, 2 Aug 2005 13:31:46 +0000 (13:31 +0000)]
Remove commented out code.
Dave Page [Tue, 2 Aug 2005 12:52:03 +0000 (12:52 +0000)]
Fix unicode copy & convert bug. SQLGetData with SQL_C_WCHAR now returns SQL_SUCCESS once the whole string is sent, and doesn't lose odd characters like it did.
Dave Page [Mon, 1 Aug 2005 20:56:46 +0000 (20:56 +0000)]
Fix memory leak per gborg bug 1356 [pauldaugherty]
Dave Page [Mon, 1 Aug 2005 20:38:37 +0000 (20:38 +0000)]
Fix SQL_MAX_IDENTIFIER_LEN per gborg bug ref: 1348
Dave Page [Mon, 1 Aug 2005 20:17:43 +0000 (20:17 +0000)]
Include win_unicode.c in Unix build because it contains functions that are used under unix.
Dave Page [Wed, 27 Jul 2005 07:36:21 +0000 (07:36 +0000)]
Japanese GUI update from Hiroshi Saito
Dave Page [Tue, 26 Jul 2005 09:29:09 +0000 (09:29 +0000)]
Get proper error messages from the server, rather than just saying the database doesn't exist.
Dave Page [Mon, 25 Jul 2005 20:38:31 +0000 (20:38 +0000)]
Add SSL support.
Dave Page [Fri, 22 Jul 2005 22:11:38 +0000 (22:11 +0000)]
In SQLGetDescFieldW:128 blen is not initialised before sent to utf8_to_ucs2(). The same thing exists in SQLColAttributeW:287 and SQLGetDiagFieldW:345, per Tomas Skäre
Dave Page [Fri, 22 Jul 2005 22:05:57 +0000 (22:05 +0000)]
In SC_pos_reload_needed:2189 rows_per_fetch is uninitialised, if create_from_scratch is false, per Tomas Skäre
Dave Page [Fri, 22 Jul 2005 21:52:31 +0000 (21:52 +0000)]
Fix some errors found by static source analysis (using Klocwork). [Tomas Skäre]
Dave Page [Fri, 22 Jul 2005 21:17:35 +0000 (21:17 +0000)]
Bump version number
Dave Page [Wed, 20 Jul 2005 14:50:21 +0000 (14:50 +0000)]
self->pgconn needs to be set as NULL after PQfinish(self->pgconn) on line 3117. [Marko Ristola]
Dave Page [Wed, 20 Jul 2005 11:54:36 +0000 (11:54 +0000)]
Remove exit_nicely function, along with it's exit(1) call, per Marko Ristola.
Dave Page [Tue, 19 Jul 2005 10:17:36 +0000 (10:17 +0000)]
Fix memory leak spotted by Marko Ristola
Anoop Kumar [Tue, 19 Jul 2005 06:40:28 +0000 (06:40 +0000)]
Removes an unnecessary call to PGgetisnull, and reduces the memory leakage (Patch from Dave Page)
Dave Page [Fri, 15 Jul 2005 10:14:09 +0000 (10:14 +0000)]
Re-add previously removed variable, ifdef'd appropriately.
Dave Page [Fri, 15 Jul 2005 09:53:34 +0000 (09:53 +0000)]
Remove an usued variable, and initialise another to zero to fix compiler warnings.
Anoop Kumar [Fri, 15 Jul 2005 09:08:50 +0000 (09:08 +0000)]
To fix the bug of PQgetvalue returning an empty string instead of a null value.
Dave Page [Wed, 13 Jul 2005 19:55:02 +0000 (19:55 +0000)]
Post-snapshot version number bump
Dave Page [Wed, 13 Jul 2005 19:50:05 +0000 (19:50 +0000)]
Fix format string [Marko Ristola]
Dave Page [Wed, 13 Jul 2005 15:41:31 +0000 (15:41 +0000)]
Rename driver for parallel testing with stable version. Update installer.
Dave Page [Wed, 13 Jul 2005 14:23:36 +0000 (14:23 +0000)]
New version of the driver with libpq based comms layer, courtesy of Anoop Kumar of Pervasive PostgreSQL. Version number bumped, and extra strings modified in the resource file to make it obvious when the libpq version of the driver is being run.
Dave Page [Thu, 30 Jun 2005 09:30:38 +0000 (09:30 +0000)]
According to MSDN docs for SQLSetStmtAttr, if you try to request a
cursor type which isn't supported, the driver should fall back to a
supported type, set an appropriate SQLState and return
SQL_SUCCESS_WITH_INFO.
Currently, if a SQL_ATTR_CURSOR_TYPE of SQL_CURSOR_DYNAMIC is requested,
we fall back to SQL_CURSOR_FORWARD_ONLY and no warning is given; this is
simply because we don't explicitly check for SQL_CURSOR_DYNAMIC.
A better solution is to check for SQL_CURSOR_DYNAMIC, fall back to
SQL_CURSOR_KEYSET_DRIVEN and set a warning accordingly.
Also, currently if SQL_CURSOR_KEYSET_DRIVEN is requested and only static
cursors are supported, we fall back without returning a warning.
[Mark G. Adams]
Dave Page [Wed, 27 Apr 2005 10:34:00 +0000 (10:34 +0000)]
Bug ref: 1262
When specifiying a client_encoding, the driver will always use UTF-8.
The problem is in connection.c, line number 1024, the code that reads:
if (!self->client_encoding || stricmp(self->client_encoding, "UNICODE"))
should read:
if (!self->client_encoding || !stricmp(self->client_encoding, "UNICODE"))
(the check with stricmp will always be true when client_encoding is different from UNICODE, not when it is UNICODE, wich is what's intended).
[melkor <NOSPAM> odyssey.com.uy]
Dave Page [Wed, 27 Apr 2005 10:27:29 +0000 (10:27 +0000)]
Bug ref: 1276
When I attempt to bind a null-terminated Unicode string parameter using
SQLBindParameter, the database engine receives a string representation of
today's date instead of the data I passed. The same issue does not occur if
I explicitly specify the length of the string instead of using SQL_NTS.
[jeremy <NOSPAM> accessdata.com]
Dave Page [Mon, 25 Apr 2005 08:08:22 +0000 (08:08 +0000)]
This patch returns the remarks (description) on a table, if available. If there is no description, it returns an empty string when SQLTables API is called. This patch works for versions with schema support.
[Anoop Kumar]
Dave Page [Mon, 25 Apr 2005 07:54:17 +0000 (07:54 +0000)]
Bump version to 08.01.0102
Peter Eisentraut [Fri, 18 Mar 2005 15:34:48 +0000 (15:34 +0000)]
When inserting SHRT_MIN into an int2 column, the ODBC driver comes up
with the following SQL:
insert into dbc_test ( i16 ) values ( -32768::int2 );
This fails with an out of range error. Changed to:
insert into dbc_test ( i16 ) values ( (-32768)::int2 );
patch by Kelly Burkhart