Tim,
As it turns out, I talked to Monty this morning, and he said to use 'longlong' for the datatype (will require my_global.h) as it works with all UNIXs and windows that MySQL runs on. As well, he said to avoid strcasencmp. I will take a look through the perl API to see if there is a similar function.
thanks again!
Patrick
On Apr 6, 2005, at 2:07 PM, Tim Bunce wrote:
On Wed, Apr 06, 2005 at 09:33:36AM -0700, Patrick Galbraith wrote:Steve,
Thanks so much for your patch! I do need to check out windows issues. I
didn't realise that windows doesn't have LONG LONG. If possible, I'd
like to pick your brain about getting a setup to compile on windows.
I'll talk to the dev team about how to deal with unix calls that aren't
implemented for windows. The one particular call, strcasencmp, is
case-insensitive, so it could be LIMIT or limit, which would fail.
Perl provides lots of similar functions for it's own portability. There's possibly an equiv to strcasencmp.
Tim.
Also, you might've found a bug with how I check to see if the my_ulonglong returned from mysql_affected_rows is -1 (error).
Again,
thank you,
Patrick
On Apr 6, 2005, at 2:41 AM, Steve Hay wrote:
Patrick Galbraith wrote:
On Apr 5, 2005, at 3:01 PM, Tim Bunce wrote:
On Tue, Apr 05, 2005 at 11:46:21AM -0700, Patrick Galbraith wrote:
Steve,
There is a version of DBD::mysql (2.9015) that does support placeholders in the server, via CVS (or I can package it and send it to you). I would be glad to give you a version of it to try out, and I think it may deal with your problem better than the current driver (which emulates it) the new driver uses the database to handle placeholders. I have tested this code on numerous SQL statements with placeholders, and it works great.
That's great. But the behavior when placeholders are emulated[1] still could be easily improved with a one-line code change.
Tim,
I'll fix this ;)
Tim.
[1] I'm presuming that the latest version still lets you use emulated placeholders either optionally or when talking to old servers. Is that right?
Yes, if the version of mysql is less than 4.1.3, it doesn't even
compile in server-side-prepare. With a server > 4.1.3, one sets
'mysql_server_prepare=1' in the DSN to enable it, and without it, it
defaults to old behaviour.
Is the looks_like_number() issue solved already in the dev version?
I just checked out the Dev-2_9 branch. It took a bit of hacking around
to get it to build on Win32, and even then a load of tests failed, but
when I re-tried the sample program that I posted previously it worked
fine even without "mysql_server_prepare=1" in the DSN. trace() shows
that the SQL statement now contains 1 rather than 1.#INF or '1.#INF':
UPDATE foo SET str = 'one', num = 1 WHERE id = 1
Is this what you expected?
Trying again with "mysql_server_prepare=1" in the DSN, the program now
says "UPDATE affected 25388176 rows" (!), but actually didn't change
any
rows. This is presumably not what you expected, but may very well be
due to the crude hacking that I did to make it build.
Do you have a Win32 environment that you could try this out on?
If not, here are the issues that I encountered:
- long long and uint are unknown; I used LONGLONG and UINT instead; - strncasecmp is unknown; I used strncmp instead; - dbdimp.h line 257 has a syntax error; - dbdimp.c contains declarations after code at line 2823.
Attached is a patch of the changes that I made.
The test failures were as follows:
C:\Temp\modules\DBD-mysql>nmake test
Microsoft (R) Program Maintenance Utility Version 6.00.8168.0 Copyright (C) Microsoft Corp 1988-1998. All rights reserved.
C:\perl5\bin\perl.exe "-MExtUtils::Command::MM" "-e"
"test_harness(0, 'b
lib\lib', 'blib\arch')" t/*.t
t/00base...........ok
t/10dsnlist........ok
t/20createdrop.....ok
t/30insertfetch....ok
t/35limit..........DBD::mysql::st execute failed: You have an error in
your SQL
syntax; check the manual that corresponds to your MySQL server version
for the r
ight syntax to use near ''20', '50'' at line 1 at t/35limit.t line 106.
DBD::mysql::st fetchall_arrayref failed: fetch() without execute() at
t/35limit.
t line 110.
t/35limit..........FAILED tests 107-109
Failed 3/113 tests, 97.35% okay
t/35prepare........ok
t/40bindparam......ok
t/40blobs..........ok
t/40listfields.....Use of uninitialized value in numeric eq (==) at
t/40listfiel
ds.t line 132.
t/40listfields.....ok
t/40nulls..........ok
t/40numrows........ok
t/50chopblanks.....ok
t/50commit.........ok
t/60leaks..........skipped
all skipped: $ENV{SLOW_TESTS} is not set or Proc::ProcessTable
not insta
lled
t/ak-dbd...........ok 12/90DBD::mysql::st execute failed: You have an
error in y
our SQL syntax; check the manual that corresponds to your MySQL server
version f
or the right syntax to use near 'LISTFIELDS testae' at line 1 at
t/ak-dbd.t line
143.
t/ak-dbd...........FAILED test 13
Failed 1/90 tests, 98.89% okay
t/akmisc...........ok 56/351DBD::mysql::st execute failed: You have an
error in
your SQL syntax; check the manual that corresponds to your MySQL server
version
for the right syntax to use near 'LISTFIELDS testae' at line 1 at
C:\Temp\module
s\DBD-mysql\blib\lib/Mysql.pm line 175.
Can't call method "name" on an undefined value at t/akmisc.t line 660.
t/akmisc...........dubious
Test returned status 2 (wstat 512, 0x200)
DIED. FAILED tests 73-351
Failed 279/351 tests, 20.51% okay
t/dbdadmin.........ok
t/insertid.........ok
t/mysql............DBD::mysql::st execute failed: Table 'table01'
already exists
at C:\Temp\modules\DBD-mysql\blib\lib/Mysql.pm line 175.
DBD::mysql::st execute failed: You have an error in your SQL syntax;
check the m
anual that corresponds to your MySQL server version for the right
syntax
to use
near 'LISTFIELDS TABLE02' at line 1 at
C:\Temp\modules\DBD-mysql\blib\lib/Mysql.
pm line 175.
Can't call method "name" on an undefined value at t/mysql.t line 502.
t/mysql............dubious
Test returned status 2 (wstat 512, 0x200)
DIED. FAILED tests 43-68
Failed 26/68 tests, 61.76% okay
t/mysql2...........ok
Failed Test Stat Wstat Total Fail Failed List of Failed
--------------------------------------------------------------------- --
--------
t/35limit.t 113 3 2.65% 107-109
t/ak-dbd.t 90 1 1.11% 13
t/akmisc.t 2 512 351 557 158.69% 73-351
t/mysql.t 2 512 68 52 76.47% 43-68
1 test skipped.
Failed 4/20 test scripts, 80.00% okay. 309/909 subtests failed, 66.01%
okay.
NMAKE : fatal error U1077: 'C:\perl5\bin\perl.exe' : return code '0xff'
Stop.
If you're able to address any of these issues and would like me to re-test, I would be very happy to do so.
- Steve
------------------------------------------------ Radan Computational Ltd.
The information contained in this message and any files transmitted with it are confidential and intended for the addressee(s) only. If you have received this message in error or there are any problems, please notify the sender immediately. The unauthorized use, disclosure, copying or alteration of this message is strictly forbidden. Note that any views or opinions presented in this email are solely those of the author and do not necessarily represent those of Radan Computational Ltd. The recipient(s) of this message should check it and any attached files for viruses: Radan Computational will accept no liability for any damage caused by any virus transmitted by this email. --- dbdimp.h.orig 2005-04-05 00:53:46.000000000 +0100 +++ dbdimp.h 2005-04-06 10:16:27.435430200 +0100 @@ -173,7 +173,7 @@ char * data; double ddata; long ldata; - long long lldata; + LONGLONG lldata;
} imp_sth_fbh_t;
@@ -254,7 +254,7 @@ #define do_error mysql_dr_error #define dbd_db_type_info_all mysql_db_type_info_all #define dbd_db_quote mysql_db_quote -#ifdef DBD_MYSQL_INSERT_ID_IS_GOOD /* prototype was broken in some versions of dbi */j +#ifdef DBD_MYSQL_INSERT_ID_IS_GOOD /* prototype was broken in some versions of dbi */ #define dbd_db_last_insert_id mysql_db_last_insert_id #endif
--- dbdimp.c.orig 2005-04-05 00:53:46.000000000 +0100 +++ dbdimp.c 2005-04-06 10:21:53.498475800 +0100 @@ -538,7 +538,7 @@ if (! limit_flag) { if ((*statement_ptr == 'l' || *statement_ptr == 'L') && - !strncasecmp(statement_ptr+1, "imit", 4)) + !strncmp(statement_ptr+1, "imit", 4)) limit_flag = 1; } switch (*statement_ptr) @@ -1481,7 +1481,7 @@
safe_hv_fetch
*/
-static char *safe_hv_fetch(HV *hv, const char *name, uint name_length)
+static char *safe_hv_fetch(HV *hv, const char *name, UINT name_length)
{
SV** svp;
STRLEN len;
@@ -2162,10 +2162,10 @@
for ( i = 0; i < statement_length - 1; i++) {
searchptr= &statement[i];
/* prepared statements for SHOW commands are not supported */
- if (!strncasecmp(searchptr, "SHOW ", 5))
+ if (!strncmp(searchptr, "SHOW ", 5))
imp_sth->use_server_side_prepare = 0;
/* if there is a 'limit' in the statement... */
- if (!limit_flag && !strncasecmp(searchptr, "limit ", 6))
+ if (!limit_flag && !strncmp(searchptr, "limit ", 6))
{
limit_flag = 1;
i += 6;
@@ -2211,7 +2211,7 @@
*/
/* this is a better way to do this */ - if ( !strncasecmp(statement, "listfields ", 11) && imp_sth->use_server_side_prepare ) + if ( !strncmp(statement, "listfields ", 11) && imp_sth->use_server_side_prepare ) { if (dbis->debug >= 2) PerlIO_printf(DBILOGFP, "\"listfields\" Statement: %s\n setting use_server_side_prepare to 0\n", statement); @@ -2356,7 +2356,7 @@ *result= NULL; }
- if (slen >= 11 && !strncasecmp(sbuf, "listfields ", 11)) + if (slen >= 11 && !strncmp(sbuf, "listfields ", 11)) { /* remove pre-space */ slen -= 10; @@ -2418,7 +2418,7 @@ else rows= mysql_num_rows(*result);
- if ((long long)rows == -1) + if ((LONGLONG)rows == -1) return(-1);
return(rows); @@ -2527,7 +2527,7 @@ rows= mysql_stmt_num_rows(stmt); } } - if ((long long)rows == -1) + if ((LONGLONG)rows == -1) return(-1); return(rows);
@@ -2641,7 +2641,7 @@ imp_sth->row_num); }
- if ((long long) imp_sth->row_num == -1) + if ((LONGLONG) imp_sth->row_num == -1) return -1; else return (int) imp_sth->row_num; @@ -2778,6 +2778,9 @@ AV *av; MYSQL_ROW cols; unsigned long* lengths; + int rc; + imp_sth_fbh_t * fbh; + MYSQL_BIND *bind;
#if MYSQL_VERSION_ID >=SERVER_PREPARE_VERSION if (imp_sth->use_server_side_prepare) @@ -2820,9 +2823,6 @@ }
#if MYSQL_VERSION_ID >=SERVER_PREPARE_VERSION - int rc; - imp_sth_fbh_t * fbh; - MYSQL_BIND *bind; if (imp_sth->use_server_side_prepare) { if (dbis->debug >= 2) --- mysql.xs.orig 2005-04-05 00:53:46.000000000 +0100 +++ mysql.xs 2005-04-06 10:21:20.859104200 +0100 @@ -564,7 +564,7 @@ D_imp_sth(sth); char buf[64];
- if ((long long)imp_sth->row_num == -1) + if ((LONGLONG)imp_sth->row_num == -1) sprintf(buf, "%d", -1); else sprintf(buf, "%llu", imp_sth->row_num);
Patrick Galbraith Senior Software Developer [EMAIL PROTECTED] http://www.mysql.com
"Whatever action a great man performs, common men follow. Whatever standards he sets by exemplary acts, all the world pursues" -- Bhagavad Gita
Patrick Galbraith Senior Software Developer [EMAIL PROTECTED] http://www.mysql.com
"Whatever action a great man performs, common men follow. Whatever standards he sets by exemplary acts, all the world pursues" -- Bhagavad Gita
