mirror/mariadb
1
0
Fork 0
mirror of https://github.com/MariaDB/server.git synced 2025-01-19 13:32:33 +01:00
Code Issues Projects Releases Packages Wiki Activity Actions
mariadb/storage/myisam/mi_create.c

885 lines
28 KiB
C
Raw Normal View History

my_strtoll10-x86.s: Corrected spelling in copyright text Makefile.am: Don't update the files from BitKeeper Many files: Removed "MySQL Finland AB & TCX DataKonsult AB" from copyright header Adjusted year(s) in copyright header Many files: Added GPL copyright text Removed files: Docs/Support/colspec-fix.pl Docs/Support/docbook-fixup.pl Docs/Support/docbook-prefix.pl Docs/Support/docbook-split Docs/Support/make-docbook Docs/Support/make-makefile Docs/Support/test-make-manual Docs/Support/test-make-manual-de Docs/Support/xwf
2006-12-31 01:02:27 +01:00
/* Copyright (C) 2000-2006 MySQL AB
Update copyright Fixed memory leak on shutdown (Affects the embedded version & MyODBC)
2001-12-06 13:10:51 +01:00
Import changeset
2000-07-31 21:29:14 +02:00
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
Many files: Changed header to GPL version 2 only
2006-12-23 20:17:15 +01:00
the Free Software Foundation; version 2 of the License.
Update copyright Fixed memory leak on shutdown (Affects the embedded version & MyODBC)
2001-12-06 13:10:51 +01:00
Import changeset
2000-07-31 21:29:14 +02:00
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
Update copyright Fixed memory leak on shutdown (Affects the embedded version & MyODBC)
2001-12-06 13:10:51 +01:00
Import changeset
2000-07-31 21:29:14 +02:00
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */
/* Create a MyISAM table */
use FT_MAX_WORD_LEN_FOR_SORT instead of HA_FT_MAXBYTELEN when calculating preferred key block length for ft index
2004-11-23 15:03:16 +01:00
#include "ftdefs.h"
This ChangeSet adds RTREE support into myisam library. RTREEs will be used for GIS extension in MySQL
2002-02-20 11:11:21 +01:00
#include "sp_defs.h"
Import changeset
2000-07-31 21:29:14 +02:00
#if defined(MSDOS) || defined(__WIN__)
#ifdef __WIN__
#include <fcntl.h>
#else
#include <process.h> /* Prototype for getpid */
#endif
#endif
#include <m_ctype.h>
Added support for key_block_size for key and table level (WL#602) Added support for key_block_size to MyISAM. Simplify interface to 'new Key' to make it easier to add new key options. mysqld option --new is used to define where key options are printed. (In 5.3 we should move all key options to after key part definition to avoid problem with reserved names) Fixed some compiler warnings and a memory leak in ssl
2006-05-03 14:59:17 +02:00
/*
Old options is used when recreating database, from myisamchk
*/
Import changeset
2000-07-31 21:29:14 +02:00
int mi_create(const char *name,uint keys,MI_KEYDEF *keydefs,
uint columns, MI_COLUMNDEF *recinfo,
uint uniques, MI_UNIQUEDEF *uniquedefs,
MI_CREATE_INFO *ci,uint flags)
{
register uint i,j;
File dfile,file;
Fixes for bugs reported by Stefano Di Paola (stefano.dipaola@wisec.it)
2005-03-03 19:51:29 +01:00
int errpos,save_errno, create_mode= O_RDWR | O_TRUNC;
Symlinking of tables should not work. Fixed error number handling bug in mysqltest. Fixed that error number from insert delayed is reported correctly. merged new vio code with old violite code.
2001-06-05 02:38:10 +02:00
myf create_flag;
use FT_MAX_WORD_LEN_FOR_SORT instead of HA_FT_MAXBYTELEN when calculating preferred key block length for ft index
2004-11-23 15:03:16 +01:00
uint fields,length,max_key_length,packed,pointer,real_length_diff,
just tried to find all 'skipp' and replace it with 'skip'.
2004-02-02 17:25:39 +01:00
key_length,info_length,key_segs,options,min_key_length_skip,
Add 0x before pointers (to help with debugging) Add support for VARCHAR with 1 or 2 length bytes Enable VARCHAR packing in MyISAM files (previous patch didn't pack data properly) Give error if we got problems in temporary tables during a SELECT Don't use new table generated by ALTER TABLE if index generation fails Fixed wrong call by range_end() (Could cause an ASSERT in debug mode)
2004-12-18 04:19:21 +01:00
base_pos,long_varchar_count,varchar_length,
Two-level index structure for FULLTEXT indexes
2003-01-21 19:24:34 +01:00
max_key_block_length,unique_key_parts,fulltext_keys,offset;
Added support for key_block_size for key and table level (WL#602) Added support for key_block_size to MyISAM. Simplify interface to 'new Key' to make it easier to add new key options. mysqld option --new is used to define where key options are printed. (In 5.3 we should move all key options to after key part definition to avoid problem with reserved names) Fixed some compiler warnings and a memory leak in ssl
2006-05-03 14:59:17 +02:00
uint aligned_key_start, block_length;
Import changeset
2000-07-31 21:29:14 +02:00
ulong reclength, real_reclength,min_pack_length;
Symlinking of tables should not work. Fixed error number handling bug in mysqltest. Fixed that error number from insert delayed is reported correctly. merged new vio code with old violite code.
2001-06-05 02:38:10 +02:00
char filename[FN_REFLEN],linkname[FN_REFLEN], *linkname_ptr;
Import changeset
2000-07-31 21:29:14 +02:00
ulong pack_reclength;
Fix for BUG#4785: * myisampack leaves key_file_length value from original table * myisamchk uses this value when calculating key file pointer length
2004-09-22 14:29:15 +02:00
ulonglong tot_length,max_rows, tmp;
Import changeset
2000-07-31 21:29:14 +02:00
enum en_fieldtype type;
MYISAM_SHARE share;
MI_KEYDEF *keydef,tmp_keydef;
MI_UNIQUEDEF *uniquedef;
MI_KEYSEG -> HA_KEYSEG _mi_key_cmp -> ha_key_cmp
2002-04-25 12:10:29 +02:00
HA_KEYSEG *keyseg,tmp_keyseg;
Import changeset
2000-07-31 21:29:14 +02:00
MI_COLUMNDEF *rec;
New thr_alarm struct for better integration with OS2 Run bootstrap in separate thread Fix bug in FLUSH TABLES table_name
2000-11-28 03:47:47 +01:00
ulong *rec_per_key_part;
Import changeset
2000-07-31 21:29:14 +02:00
my_off_t key_root[MI_MAX_POSSIBLE_KEY],key_del[MI_MAX_KEY_BLOCK_SIZE];
MI_CREATE_INFO tmp_create_info;
DBUG_ENTER("mi_create");
Bug#11824 - internal /tmp/*.{MYD,MYI} files remain, causing subsequent queries to fail Very complex select statements can create temporary tables that are too big to be represented as a MyISAM table. This was not checked at table creation time, but only at open time. The result was an attempt to delete the "impossible" table. But if the server is built --with-raid, MyISAM tries to open the table before deleting the files. It needs to find out if the table uses the raid support and how many raid chunks there are. This is done with an open "for repair", which will almost always succeed. But in this case we have an "impossible" table. The open failed. Hence the files were not deleted. Also the error message was a bit unspecific. I turned an open error in this situation into the assumption of having no raid support on the table. Thus the normal data file is tried to be deleted. This may however leave existing raid chunks behind. I also added a check in mi_create() to prevent the creation of an "impossible" table. A more decriptive error message is given in this case. No test case. The required select statement is way too large for the test suite. I added a test script to the bug report.
2006-06-27 11:26:41 +02:00
DBUG_PRINT("enter", ("keys: %u columns: %u uniques: %u flags: %u",
keys, columns, uniques, flags));
Import changeset
2000-07-31 21:29:14 +02:00
if (!ci)
{
bzero((char*) &tmp_create_info,sizeof(tmp_create_info));
ci=&tmp_create_info;
}
Fix to not running mysqld as root by default
2001-09-02 19:42:07 +02:00
if (keys + uniques > MI_MAX_KEY || columns == 0)
Import changeset
2000-07-31 21:29:14 +02:00
{
DBUG_RETURN(my_errno=HA_WRONG_CREATE_OPTION);
}
LINT_INIT(dfile);
LINT_INIT(file);
errpos=0;
options=0;
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
bzero((uchar*) &share,sizeof(share));
Import changeset
2000-07-31 21:29:14 +02:00
if (flags & HA_DONT_TOUCH_DATA)
{
if (!(ci->old_options & HA_OPTION_TEMP_COMPRESS_RECORD))
options=ci->old_options &
(HA_OPTION_COMPRESS_RECORD | HA_OPTION_PACK_RECORD |
HA_OPTION_READ_ONLY_DATA | HA_OPTION_CHECKSUM |
HA_OPTION_TMP_TABLE | HA_OPTION_DELAY_KEY_WRITE);
else
options=ci->old_options &
(HA_OPTION_CHECKSUM | HA_OPTION_TMP_TABLE | HA_OPTION_DELAY_KEY_WRITE);
}
if (ci->reloc_rows > ci->max_rows)
New thr_alarm struct for better integration with OS2 Run bootstrap in separate thread Fix bug in FLUSH TABLES table_name
2000-11-28 03:47:47 +01:00
ci->reloc_rows=ci->max_rows; /* Check if wrong parameter */
if (!(rec_per_key_part=
(ulong*) my_malloc((keys + uniques)*MI_MAX_KEY_SEG*sizeof(long),
MYF(MY_WME | MY_ZEROFILL))))
DBUG_RETURN(my_errno);
Import changeset
2000-07-31 21:29:14 +02:00
/* Start by checking fields and field-types used */
Add 0x before pointers (to help with debugging) Add support for VARCHAR with 1 or 2 length bytes Enable VARCHAR packing in MyISAM files (previous patch didn't pack data properly) Give error if we got problems in temporary tables during a SELECT Don't use new table generated by ALTER TABLE if index generation fails Fixed wrong call by range_end() (Could cause an ASSERT in debug mode)
2004-12-18 04:19:21 +01:00
reclength=varchar_length=long_varchar_count=packed=
Import changeset
2000-07-31 21:29:14 +02:00
min_pack_length=pack_reclength=0;
for (rec=recinfo, fields=0 ;
fields != columns ;
rec++,fields++)
{
reclength+=rec->length;
if ((type=(enum en_fieldtype) rec->type) != FIELD_NORMAL &&
type != FIELD_CHECK)
{
packed++;
if (type == FIELD_BLOB)
{
share.base.blobs++;
if (pack_reclength != INT_MAX32)
{
if (rec->length == 4+mi_portable_sizeof_char_ptr)
pack_reclength= INT_MAX32;
else
pack_reclength+=(1 << ((rec->length-mi_portable_sizeof_char_ptr)*8)); /* Max blob length */
}
}
Portability fix Cleanup typos (like SKIPP -> SKIP)
2001-11-06 23:13:29 +01:00
else if (type == FIELD_SKIP_PRESPACE ||
type == FIELD_SKIP_ENDSPACE)
Import changeset
2000-07-31 21:29:14 +02:00
{
if (pack_reclength != INT_MAX32)
pack_reclength+= rec->length > 255 ? 2 : 1;
min_pack_length++;
}
else if (type == FIELD_VARCHAR)
{
Add 0x before pointers (to help with debugging) Add support for VARCHAR with 1 or 2 length bytes Enable VARCHAR packing in MyISAM files (previous patch didn't pack data properly) Give error if we got problems in temporary tables during a SELECT Don't use new table generated by ALTER TABLE if index generation fails Fixed wrong call by range_end() (Could cause an ASSERT in debug mode)
2004-12-18 04:19:21 +01:00
varchar_length+= rec->length-1; /* Used for min_pack_length */
Import changeset
2000-07-31 21:29:14 +02:00
packed--;
Add 0x before pointers (to help with debugging) Add support for VARCHAR with 1 or 2 length bytes Enable VARCHAR packing in MyISAM files (previous patch didn't pack data properly) Give error if we got problems in temporary tables during a SELECT Don't use new table generated by ALTER TABLE if index generation fails Fixed wrong call by range_end() (Could cause an ASSERT in debug mode)
2004-12-18 04:19:21 +01:00
pack_reclength++;
min_pack_length++;
/* We must test for 257 as length includes pack-length */
locking issues, test for LONGLONG_MIN in decimal.c
2005-12-23 20:50:28 +01:00
if (test(rec->length >= 257))
Add 0x before pointers (to help with debugging) Add support for VARCHAR with 1 or 2 length bytes Enable VARCHAR packing in MyISAM files (previous patch didn't pack data properly) Give error if we got problems in temporary tables during a SELECT Don't use new table generated by ALTER TABLE if index generation fails Fixed wrong call by range_end() (Could cause an ASSERT in debug mode)
2004-12-18 04:19:21 +01:00
{
Import changeset
2000-07-31 21:29:14 +02:00
long_varchar_count++;
Add 0x before pointers (to help with debugging) Add support for VARCHAR with 1 or 2 length bytes Enable VARCHAR packing in MyISAM files (previous patch didn't pack data properly) Give error if we got problems in temporary tables during a SELECT Don't use new table generated by ALTER TABLE if index generation fails Fixed wrong call by range_end() (Could cause an ASSERT in debug mode)
2004-12-18 04:19:21 +01:00
pack_reclength+= 2; /* May be packed on 3 bytes */
Import changeset
2000-07-31 21:29:14 +02:00
}
}
Portability fix Cleanup typos (like SKIPP -> SKIP)
2001-11-06 23:13:29 +01:00
else if (type != FIELD_SKIP_ZERO)
Import changeset
2000-07-31 21:29:14 +02:00
{
min_pack_length+=rec->length;
packed--; /* Not a pack record type */
}
}
else /* FIELD_NORMAL */
min_pack_length+=rec->length;
}
if ((packed & 7) == 1)
{ /* Bad packing, try to remove a zero-field */
while (rec != recinfo)
{
rec--;
Portability fix Cleanup typos (like SKIPP -> SKIP)
2001-11-06 23:13:29 +01:00
if (rec->type == (int) FIELD_SKIP_ZERO && rec->length == 1)
Import changeset
2000-07-31 21:29:14 +02:00
{
BUG#26881 - Large MERGE tables report incorrect specification when no differences in tables Certain merge tables were wrongly reported as having incorrect definition: - Some fields that are 1 byte long (e.g. TINYINT, CHAR(1)), might be internally casted (in certain cases) to a different type on a storage engine layer. (affects 4.1 and up) - If tables in a merge (and a MERGE table itself) had short VARCHAR column (less than 4 bytes) and at least one (but not all) tables were ALTER'ed (even to an identical table: ALTER TABLE xxx ENGINE=yyy), table definitions went ouf of sync. (affects 4.1 only) This is fixed by relaxing a check for underlying conformance and setting field type to FIELD_TYPE_STRING in case varchar is shorter than 4 when a table is created.
2007-03-13 15:02:06 +01:00
/*
NOTE1: here we change a field type FIELD_SKIP_ZERO ->
FIELD_NORMAL
*/
Import changeset
2000-07-31 21:29:14 +02:00
rec->type=(int) FIELD_NORMAL;
packed--;
min_pack_length++;
break;
}
}
}
if (packed || (flags & HA_PACK_RECORD))
options|=HA_OPTION_PACK_RECORD; /* Must use packed records */
Fixed length rows should not use row checksum in MyISAM tables Fixed bug in handling of packed MyISAM keys
2001-06-26 09:23:50 +02:00
/* We can't use checksum with static length rows */
if (!(options & HA_OPTION_PACK_RECORD))
options&= ~HA_OPTION_CHECKSUM;
Add 0x before pointers (to help with debugging) Add support for VARCHAR with 1 or 2 length bytes Enable VARCHAR packing in MyISAM files (previous patch didn't pack data properly) Give error if we got problems in temporary tables during a SELECT Don't use new table generated by ALTER TABLE if index generation fails Fixed wrong call by range_end() (Could cause an ASSERT in debug mode)
2004-12-18 04:19:21 +01:00
if (!(options & (HA_OPTION_PACK_RECORD | HA_OPTION_COMPRESS_RECORD)))
min_pack_length+= varchar_length;
Import changeset
2000-07-31 21:29:14 +02:00
if (flags & HA_CREATE_TMP_TABLE)
Fixes for bugs reported by Stefano Di Paola (stefano.dipaola@wisec.it)
2005-03-03 19:51:29 +01:00
{
Import changeset
2000-07-31 21:29:14 +02:00
options|= HA_OPTION_TMP_TABLE;
Fixes for bugs reported by Stefano Di Paola (stefano.dipaola@wisec.it)
2005-03-03 19:51:29 +01:00
create_mode|= O_EXCL | O_NOFOLLOW;
}
Import changeset
2000-07-31 21:29:14 +02:00
if (flags & HA_CREATE_CHECKSUM || (options & HA_OPTION_CHECKSUM))
{
options|= HA_OPTION_CHECKSUM;
min_pack_length++;
}
if (flags & HA_CREATE_DELAY_KEY_WRITE)
options|= HA_OPTION_DELAY_KEY_WRITE;
WL#2575 - Fulltext: Parser plugin for FTS Manual merge.
2005-12-28 13:05:30 +01:00
if (flags & HA_CREATE_RELIES_ON_SQL_LAYER)
options|= HA_OPTION_RELIES_ON_SQL_LAYER;
Import changeset
2000-07-31 21:29:14 +02:00
packed=(packed+7)/8;
if (pack_reclength != INT_MAX32)
pack_reclength+= reclength+packed +
test(test_all_bits(options, HA_OPTION_CHECKSUM | HA_PACK_RECORD));
min_pack_length+=packed;
concurrent-insert can now be set to 2 for concurrent inserts when there is holes in the data file myisam_max_extra_sort_file_size is depricated Ensure that myisam_data_pointer_size is honoured when creating new MyISAM files Changed default value of myisam_data_pointer_size from 4 to 6 to get rid of 'table-is-full' errors
2005-05-13 11:08:08 +02:00
if (!ci->data_file_length && ci->max_rows)
Import changeset
2000-07-31 21:29:14 +02:00
{
concurrent-insert can now be set to 2 for concurrent inserts when there is holes in the data file myisam_max_extra_sort_file_size is depricated Ensure that myisam_data_pointer_size is honoured when creating new MyISAM files Changed default value of myisam_data_pointer_size from 4 to 6 to get rid of 'table-is-full' errors
2005-05-13 11:08:08 +02:00
if (pack_reclength == INT_MAX32 ||
(~(ulonglong) 0)/ci->max_rows < (ulonglong) pack_reclength)
Import changeset
2000-07-31 21:29:14 +02:00
ci->data_file_length= ~(ulonglong) 0;
else
ci->data_file_length=(ulonglong) ci->max_rows*pack_reclength;
}
else if (!ci->max_rows)
ci->max_rows=(ha_rows) (ci->data_file_length/(min_pack_length +
((options & HA_OPTION_PACK_RECORD) ?
3 : 0)));
if (options & (HA_OPTION_COMPRESS_RECORD | HA_OPTION_PACK_RECORD))
myisam_data_pointer_size
2004-05-01 15:41:59 +02:00
pointer=mi_get_pointer_length(ci->data_file_length,myisam_data_pointer_size);
Import changeset
2000-07-31 21:29:14 +02:00
else
myisam_data_pointer_size
2004-05-01 15:41:59 +02:00
pointer=mi_get_pointer_length(ci->max_rows,myisam_data_pointer_size);
Import changeset
2000-07-31 21:29:14 +02:00
if (!(max_rows=(ulonglong) ci->max_rows))
max_rows= ((((ulonglong) 1 << (pointer*8)) -1) / min_pack_length);
real_reclength=reclength;
if (!(options & (HA_OPTION_COMPRESS_RECORD | HA_OPTION_PACK_RECORD)))
{
if (reclength <= pointer)
reclength=pointer+1; /* reserve place for delete link */
}
else
Add 0x before pointers (to help with debugging) Add support for VARCHAR with 1 or 2 length bytes Enable VARCHAR packing in MyISAM files (previous patch didn't pack data properly) Give error if we got problems in temporary tables during a SELECT Don't use new table generated by ALTER TABLE if index generation fails Fixed wrong call by range_end() (Could cause an ASSERT in debug mode)
2004-12-18 04:19:21 +01:00
reclength+= long_varchar_count; /* We need space for varchar! */
Import changeset
2000-07-31 21:29:14 +02:00
max_key_length=0; tot_length=0 ; key_segs=0;
Two-level index structure for FULLTEXT indexes
2003-01-21 19:24:34 +01:00
fulltext_keys=0;
Import changeset
2000-07-31 21:29:14 +02:00
max_key_block_length=0;
share.state.rec_per_key_part=rec_per_key_part;
share.state.key_root=key_root;
share.state.key_del=key_del;
if (uniques)
{
Cleaned up SSL documentation Fixes for embedded server Made key_cache more configurable Fixed that one can change key blocksize in MyISAM A lot of optimizations to make MyISAM slightly faster
2001-10-02 04:53:00 +02:00
max_key_block_length= myisam_block_size;
Merge with 3.23
2003-01-16 19:47:46 +01:00
max_key_length= MI_UNIQUE_HASH_LENGTH + pointer;
Import changeset
2000-07-31 21:29:14 +02:00
}
for (i=0, keydef=keydefs ; i < keys ; i++ , keydef++)
{
This ChangeSet adds RTREE support into myisam library. RTREEs will be used for GIS extension in MySQL
2002-02-20 11:11:21 +01:00
share.state.key_root[i]= HA_OFFSET_ERROR;
use FT_MAX_WORD_LEN_FOR_SORT instead of HA_FT_MAXBYTELEN when calculating preferred key block length for ft index
2004-11-23 15:03:16 +01:00
min_key_length_skip=length=real_length_diff=0;
Import changeset
2000-07-31 21:29:14 +02:00
key_length=pointer;
This ChangeSet adds RTREE support into myisam library. RTREEs will be used for GIS extension in MySQL
2002-02-20 11:11:21 +01:00
if (keydef->flag & HA_SPATIAL)
{
WL#1163 To make spatial code optional myisam spatial code isolated
2004-03-12 15:51:03 +01:00
#ifdef HAVE_SPATIAL
This ChangeSet adds RTREE support into myisam library. RTREEs will be used for GIS extension in MySQL
2002-02-20 11:11:21 +01:00
/* BAR TODO to support 3D and more dimensions in the future */
Two-level index structure for FULLTEXT indexes
2003-01-21 19:24:34 +01:00
uint sp_segs=SPDIMS*2;
This ChangeSet adds RTREE support into myisam library. RTREEs will be used for GIS extension in MySQL
2002-02-20 11:11:21 +01:00
keydef->flag=HA_SPATIAL;
Import changeset
2000-07-31 21:29:14 +02:00
This ChangeSet adds RTREE support into myisam library. RTREEs will be used for GIS extension in MySQL
2002-02-20 11:11:21 +01:00
if (flags & HA_DONT_TOUCH_DATA)
{
Two-level index structure for FULLTEXT indexes
2003-01-21 19:24:34 +01:00
/*
This ChangeSet adds RTREE support into myisam library. RTREEs will be used for GIS extension in MySQL
2002-02-20 11:11:21 +01:00
called by myisamchk - i.e. table structure was taken from
Two-level index structure for FULLTEXT indexes
2003-01-21 19:24:34 +01:00
MYI file and SPATIAL key *does have* additional sp_segs keysegs.
Fix for bug #10893 (myisamchk fails on packed spatial) need to be fixed in 4.1
2005-07-23 17:04:15 +02:00
keydef->seg here points right at the GEOMETRY segment,
so we only need to decrease keydef->keysegs.
(see recreate_table() in mi_check.c)
This ChangeSet adds RTREE support into myisam library. RTREEs will be used for GIS extension in MySQL
2002-02-20 11:11:21 +01:00
*/
Fix for bug #10893 (myisamchk fails on packed spatial) need to be fixed in 4.1
2005-07-23 17:04:15 +02:00
keydef->keysegs-=sp_segs-1;
This ChangeSet adds RTREE support into myisam library. RTREEs will be used for GIS extension in MySQL
2002-02-20 11:11:21 +01:00
}
for (j=0, keyseg=keydef->seg ; (int) j < keydef->keysegs ;
j++, keyseg++)
{
if (keyseg->type != HA_KEYTYPE_BINARY &&
Add 0x before pointers (to help with debugging) Add support for VARCHAR with 1 or 2 length bytes Enable VARCHAR packing in MyISAM files (previous patch didn't pack data properly) Give error if we got problems in temporary tables during a SELECT Don't use new table generated by ALTER TABLE if index generation fails Fixed wrong call by range_end() (Could cause an ASSERT in debug mode)
2004-12-18 04:19:21 +01:00
keyseg->type != HA_KEYTYPE_VARBINARY1 &&
keyseg->type != HA_KEYTYPE_VARBINARY2)
This ChangeSet adds RTREE support into myisam library. RTREEs will be used for GIS extension in MySQL
2002-02-20 11:11:21 +01:00
{
my_errno=HA_WRONG_CREATE_OPTION;
goto err;
}
}
keydef->keysegs+=sp_segs;
key_length+=SPLEN*sp_segs;
length++; /* At least one length byte */
just tried to find all 'skipp' and replace it with 'skip'.
2004-02-02 17:25:39 +01:00
min_key_length_skip+=SPLEN*2*SPDIMS;
WL#1163 To make spatial code optional myisam spatial code isolated
2004-03-12 15:51:03 +01:00
#else
my_errno= HA_ERR_UNSUPPORTED;
goto err;
#endif /*HAVE_SPATIAL*/
This ChangeSet adds RTREE support into myisam library. RTREEs will be used for GIS extension in MySQL
2002-02-20 11:11:21 +01:00
}
Add support for up to VARCHAR (size up to 65535) Renamed HA_VAR_LENGTH to HA_VAR_LENGTH_PART Renamed in all files FIELD_TYPE_STRING and FIELD_TYPE_VAR_STRING to MYSQL_TYPE_STRING and MYSQL_TYPE_VAR_STRING to make it easy to catch all possible errors Added support for VARCHAR KEYS to heap Removed support for ISAM Now only long VARCHAR columns are changed to TEXT on demand (not CHAR) Internal temporary files can now use fixed length tables if the used VARCHAR columns are short
2004-12-06 01:00:37 +01:00
else if (keydef->flag & HA_FULLTEXT)
Import changeset
2000-07-31 21:29:14 +02:00
{
keydef->flag=HA_FULLTEXT | HA_PACK_KEY | HA_VAR_LENGTH_KEY;
options|=HA_OPTION_PACK_KEYS; /* Using packed keys */
for (j=0, keyseg=keydef->seg ; (int) j < keydef->keysegs ;
j++, keyseg++)
{
if (keyseg->type != HA_KEYTYPE_TEXT &&
Add 0x before pointers (to help with debugging) Add support for VARCHAR with 1 or 2 length bytes Enable VARCHAR packing in MyISAM files (previous patch didn't pack data properly) Give error if we got problems in temporary tables during a SELECT Don't use new table generated by ALTER TABLE if index generation fails Fixed wrong call by range_end() (Could cause an ASSERT in debug mode)
2004-12-18 04:19:21 +01:00
keyseg->type != HA_KEYTYPE_VARTEXT1 &&
keyseg->type != HA_KEYTYPE_VARTEXT2)
Import changeset
2000-07-31 21:29:14 +02:00
{
my_errno=HA_WRONG_CREATE_OPTION;
goto err;
}
Add 0x before pointers (to help with debugging) Add support for VARCHAR with 1 or 2 length bytes Enable VARCHAR packing in MyISAM files (previous patch didn't pack data properly) Give error if we got problems in temporary tables during a SELECT Don't use new table generated by ALTER TABLE if index generation fails Fixed wrong call by range_end() (Could cause an ASSERT in debug mode)
2004-12-18 04:19:21 +01:00
if (!(keyseg->flag & HA_BLOB_PART) &&
(keyseg->type == HA_KEYTYPE_VARTEXT1 ||
keyseg->type == HA_KEYTYPE_VARTEXT2))
{
/* Make a flag that this is a VARCHAR */
keyseg->flag|= HA_VAR_LENGTH_PART;
/* Store in bit_start number of bytes used to pack the length */
keyseg->bit_start= ((keyseg->type == HA_KEYTYPE_VARTEXT1)?
1 : 2);
}
Import changeset
2000-07-31 21:29:14 +02:00
}
Two-level index structure for FULLTEXT indexes
2003-01-21 19:24:34 +01:00
fulltext_keys++;
make a clear distinction between max_word_length in *characters* and in *bytes*
2003-12-04 21:58:28 +01:00
key_length+= HA_FT_MAXBYTELEN+HA_FT_WLEN;
Import changeset
2000-07-31 21:29:14 +02:00
length++; /* At least one length byte */
just tried to find all 'skipp' and replace it with 'skip'.
2004-02-02 17:25:39 +01:00
min_key_length_skip+=HA_FT_MAXBYTELEN;
use FT_MAX_WORD_LEN_FOR_SORT instead of HA_FT_MAXBYTELEN when calculating preferred key block length for ft index
2004-11-23 15:03:16 +01:00
real_length_diff=HA_FT_MAXBYTELEN-FT_MAX_WORD_LEN_FOR_SORT;
Import changeset
2000-07-31 21:29:14 +02:00
}
else
{
Fixed bug when having many binary log files Fixed bug when deleting keys with possible NULL values many changes to mysqltest
2001-01-07 23:04:30 +01:00
/* Test if prefix compression */
if (keydef->flag & HA_PACK_KEY)
Import changeset
2000-07-31 21:29:14 +02:00
{
Fixed bug when having many binary log files Fixed bug when deleting keys with possible NULL values many changes to mysqltest
2001-01-07 23:04:30 +01:00
/* Can't use space_compression on number keys */
if ((keydef->seg[0].flag & HA_SPACE_PACK) &&
keydef->seg[0].type == (int) HA_KEYTYPE_NUM)
keydef->seg[0].flag&= ~HA_SPACE_PACK;
Fixed some varnings from valgrind Set min value of max_allowed_packet to 1024 Fixed problem with UNION's without braces and SQL_CALC_FOUND_ROWS, LIMIT #,# and ORDER BY...LIMIT
2003-08-20 15:25:44 +02:00
/* Only use HA_PACK_KEY when first segment is a variable length key */
Fixed bug when having many binary log files Fixed bug when deleting keys with possible NULL values many changes to mysqltest
2001-01-07 23:04:30 +01:00
if (!(keydef->seg[0].flag & (HA_SPACE_PACK | HA_BLOB_PART |
Add support for up to VARCHAR (size up to 65535) Renamed HA_VAR_LENGTH to HA_VAR_LENGTH_PART Renamed in all files FIELD_TYPE_STRING and FIELD_TYPE_VAR_STRING to MYSQL_TYPE_STRING and MYSQL_TYPE_VAR_STRING to make it easy to catch all possible errors Added support for VARCHAR KEYS to heap Removed support for ISAM Now only long VARCHAR columns are changed to TEXT on demand (not CHAR) Internal temporary files can now use fixed length tables if the used VARCHAR columns are short
2004-12-06 01:00:37 +01:00
HA_VAR_LENGTH_PART)))
Fixed bug when having many binary log files Fixed bug when deleting keys with possible NULL values many changes to mysqltest
2001-01-07 23:04:30 +01:00
{
/* pack relative to previous key */
keydef->flag&= ~HA_PACK_KEY;
keydef->flag|= HA_BINARY_PACK_KEY | HA_VAR_LENGTH_KEY;
}
else
{
keydef->seg[0].flag|=HA_PACK_KEY; /* for easyer intern test */
keydef->flag|=HA_VAR_LENGTH_KEY;
options|=HA_OPTION_PACK_KEYS; /* Using packed keys */
}
Import changeset
2000-07-31 21:29:14 +02:00
}
Fixed bug when having many binary log files Fixed bug when deleting keys with possible NULL values many changes to mysqltest
2001-01-07 23:04:30 +01:00
if (keydef->flag & HA_BINARY_PACK_KEY)
Import changeset
2000-07-31 21:29:14 +02:00
options|=HA_OPTION_PACK_KEYS; /* Using packed keys */
fix auto-increment in sub-key and insert ... select
2003-03-02 01:48:57 +01:00
if (keydef->flag & HA_AUTO_KEY && ci->with_auto_increment)
Fixed bug when having many binary log files Fixed bug when deleting keys with possible NULL values many changes to mysqltest
2001-01-07 23:04:30 +01:00
share.base.auto_key=i+1;
for (j=0, keyseg=keydef->seg ; j < keydef->keysegs ; j++, keyseg++)
Import changeset
2000-07-31 21:29:14 +02:00
{
Fixed bug when having many binary log files Fixed bug when deleting keys with possible NULL values many changes to mysqltest
2001-01-07 23:04:30 +01:00
/* numbers are stored with high by first to make compression easier */
switch (keyseg->type) {
case HA_KEYTYPE_SHORT_INT:
case HA_KEYTYPE_LONG_INT:
case HA_KEYTYPE_FLOAT:
case HA_KEYTYPE_DOUBLE:
case HA_KEYTYPE_USHORT_INT:
case HA_KEYTYPE_ULONG_INT:
case HA_KEYTYPE_LONGLONG:
case HA_KEYTYPE_ULONGLONG:
case HA_KEYTYPE_INT24:
case HA_KEYTYPE_UINT24:
case HA_KEYTYPE_INT8:
keyseg->flag|= HA_SWAP_KEY;
Add support for up to VARCHAR (size up to 65535) Renamed HA_VAR_LENGTH to HA_VAR_LENGTH_PART Renamed in all files FIELD_TYPE_STRING and FIELD_TYPE_VAR_STRING to MYSQL_TYPE_STRING and MYSQL_TYPE_VAR_STRING to make it easy to catch all possible errors Added support for VARCHAR KEYS to heap Removed support for ISAM Now only long VARCHAR columns are changed to TEXT on demand (not CHAR) Internal temporary files can now use fixed length tables if the used VARCHAR columns are short
2004-12-06 01:00:37 +01:00
break;
Add 0x before pointers (to help with debugging) Add support for VARCHAR with 1 or 2 length bytes Enable VARCHAR packing in MyISAM files (previous patch didn't pack data properly) Give error if we got problems in temporary tables during a SELECT Don't use new table generated by ALTER TABLE if index generation fails Fixed wrong call by range_end() (Could cause an ASSERT in debug mode)
2004-12-18 04:19:21 +01:00
case HA_KEYTYPE_VARTEXT1:
case HA_KEYTYPE_VARTEXT2:
case HA_KEYTYPE_VARBINARY1:
case HA_KEYTYPE_VARBINARY2:
Add support for up to VARCHAR (size up to 65535) Renamed HA_VAR_LENGTH to HA_VAR_LENGTH_PART Renamed in all files FIELD_TYPE_STRING and FIELD_TYPE_VAR_STRING to MYSQL_TYPE_STRING and MYSQL_TYPE_VAR_STRING to make it easy to catch all possible errors Added support for VARCHAR KEYS to heap Removed support for ISAM Now only long VARCHAR columns are changed to TEXT on demand (not CHAR) Internal temporary files can now use fixed length tables if the used VARCHAR columns are short
2004-12-06 01:00:37 +01:00
if (!(keyseg->flag & HA_BLOB_PART))
Add 0x before pointers (to help with debugging) Add support for VARCHAR with 1 or 2 length bytes Enable VARCHAR packing in MyISAM files (previous patch didn't pack data properly) Give error if we got problems in temporary tables during a SELECT Don't use new table generated by ALTER TABLE if index generation fails Fixed wrong call by range_end() (Could cause an ASSERT in debug mode)
2004-12-18 04:19:21 +01:00
{
/* Make a flag that this is a VARCHAR */
Add support for up to VARCHAR (size up to 65535) Renamed HA_VAR_LENGTH to HA_VAR_LENGTH_PART Renamed in all files FIELD_TYPE_STRING and FIELD_TYPE_VAR_STRING to MYSQL_TYPE_STRING and MYSQL_TYPE_VAR_STRING to make it easy to catch all possible errors Added support for VARCHAR KEYS to heap Removed support for ISAM Now only long VARCHAR columns are changed to TEXT on demand (not CHAR) Internal temporary files can now use fixed length tables if the used VARCHAR columns are short
2004-12-06 01:00:37 +01:00
keyseg->flag|= HA_VAR_LENGTH_PART;
Add 0x before pointers (to help with debugging) Add support for VARCHAR with 1 or 2 length bytes Enable VARCHAR packing in MyISAM files (previous patch didn't pack data properly) Give error if we got problems in temporary tables during a SELECT Don't use new table generated by ALTER TABLE if index generation fails Fixed wrong call by range_end() (Could cause an ASSERT in debug mode)
2004-12-18 04:19:21 +01:00
/* Store in bit_start number of bytes used to pack the length */
keyseg->bit_start= ((keyseg->type == HA_KEYTYPE_VARTEXT1 ||
keyseg->type == HA_KEYTYPE_VARBINARY1) ?
1 : 2);
}
Add support for up to VARCHAR (size up to 65535) Renamed HA_VAR_LENGTH to HA_VAR_LENGTH_PART Renamed in all files FIELD_TYPE_STRING and FIELD_TYPE_VAR_STRING to MYSQL_TYPE_STRING and MYSQL_TYPE_VAR_STRING to make it easy to catch all possible errors Added support for VARCHAR KEYS to heap Removed support for ISAM Now only long VARCHAR columns are changed to TEXT on demand (not CHAR) Internal temporary files can now use fixed length tables if the used VARCHAR columns are short
2004-12-06 01:00:37 +01:00
break;
Fixed bug when having many binary log files Fixed bug when deleting keys with possible NULL values many changes to mysqltest
2001-01-07 23:04:30 +01:00
default:
break;
Import changeset
2000-07-31 21:29:14 +02:00
}
Fixed bug when having many binary log files Fixed bug when deleting keys with possible NULL values many changes to mysqltest
2001-01-07 23:04:30 +01:00
if (keyseg->flag & HA_SPACE_PACK)
{
Add support for up to VARCHAR (size up to 65535) Renamed HA_VAR_LENGTH to HA_VAR_LENGTH_PART Renamed in all files FIELD_TYPE_STRING and FIELD_TYPE_VAR_STRING to MYSQL_TYPE_STRING and MYSQL_TYPE_VAR_STRING to make it easy to catch all possible errors Added support for VARCHAR KEYS to heap Removed support for ISAM Now only long VARCHAR columns are changed to TEXT on demand (not CHAR) Internal temporary files can now use fixed length tables if the used VARCHAR columns are short
2004-12-06 01:00:37 +01:00
DBUG_ASSERT(!(keyseg->flag & HA_VAR_LENGTH_PART));
Fixed bug when having many binary log files Fixed bug when deleting keys with possible NULL values many changes to mysqltest
2001-01-07 23:04:30 +01:00
keydef->flag |= HA_SPACE_PACK_USED | HA_VAR_LENGTH_KEY;
options|=HA_OPTION_PACK_KEYS; /* Using packed keys */
length++; /* At least one length byte */
just tried to find all 'skipp' and replace it with 'skip'.
2004-02-02 17:25:39 +01:00
min_key_length_skip+=keyseg->length;
Fixed bug when having many binary log files Fixed bug when deleting keys with possible NULL values many changes to mysqltest
2001-01-07 23:04:30 +01:00
if (keyseg->length >= 255)
{ /* prefix may be 3 bytes */
just tried to find all 'skipp' and replace it with 'skip'.
2004-02-02 17:25:39 +01:00
min_key_length_skip+=2;
Fixed bug when having many binary log files Fixed bug when deleting keys with possible NULL values many changes to mysqltest
2001-01-07 23:04:30 +01:00
length+=2;
}
}
Add support for up to VARCHAR (size up to 65535) Renamed HA_VAR_LENGTH to HA_VAR_LENGTH_PART Renamed in all files FIELD_TYPE_STRING and FIELD_TYPE_VAR_STRING to MYSQL_TYPE_STRING and MYSQL_TYPE_VAR_STRING to make it easy to catch all possible errors Added support for VARCHAR KEYS to heap Removed support for ISAM Now only long VARCHAR columns are changed to TEXT on demand (not CHAR) Internal temporary files can now use fixed length tables if the used VARCHAR columns are short
2004-12-06 01:00:37 +01:00
if (keyseg->flag & (HA_VAR_LENGTH_PART | HA_BLOB_PART))
Fixed bug when having many binary log files Fixed bug when deleting keys with possible NULL values many changes to mysqltest
2001-01-07 23:04:30 +01:00
{
Add 0x before pointers (to help with debugging) Add support for VARCHAR with 1 or 2 length bytes Enable VARCHAR packing in MyISAM files (previous patch didn't pack data properly) Give error if we got problems in temporary tables during a SELECT Don't use new table generated by ALTER TABLE if index generation fails Fixed wrong call by range_end() (Could cause an ASSERT in debug mode)
2004-12-18 04:19:21 +01:00
DBUG_ASSERT(!test_all_bits(keyseg->flag,
(HA_VAR_LENGTH_PART | HA_BLOB_PART)));
Fixed bug when having many binary log files Fixed bug when deleting keys with possible NULL values many changes to mysqltest
2001-01-07 23:04:30 +01:00
keydef->flag|=HA_VAR_LENGTH_KEY;
length++; /* At least one length byte */
options|=HA_OPTION_PACK_KEYS; /* Using packed keys */
just tried to find all 'skipp' and replace it with 'skip'.
2004-02-02 17:25:39 +01:00
min_key_length_skip+=keyseg->length;
Fixed bug when having many binary log files Fixed bug when deleting keys with possible NULL values many changes to mysqltest
2001-01-07 23:04:30 +01:00
if (keyseg->length >= 255)
{ /* prefix may be 3 bytes */
just tried to find all 'skipp' and replace it with 'skip'.
2004-02-02 17:25:39 +01:00
min_key_length_skip+=2;
Fixed bug when having many binary log files Fixed bug when deleting keys with possible NULL values many changes to mysqltest
2001-01-07 23:04:30 +01:00
length+=2;
}
}
key_length+= keyseg->length;
if (keyseg->null_bit)
{
key_length++;
options|=HA_OPTION_PACK_KEYS;
keyseg->flag|=HA_NULL_PART;
keydef->flag|=HA_VAR_LENGTH_KEY | HA_NULL_PART_KEY;
Import changeset
2000-07-31 21:29:14 +02:00
}
}
} /* if HA_FULLTEXT */
key_segs+=keydef->keysegs;
if (keydef->keysegs > MI_MAX_KEY_SEG)
{
my_errno=HA_WRONG_CREATE_OPTION;
goto err;
}
Allow keys with 0 segements in MyISAM This fixed a bug in SELECT DISTINCT when all selected parts where constants. (Bug #3203)
2004-03-26 13:44:52 +01:00
/*
key_segs may be 0 in the case when we only want to be able to
add on row into the table. This can happen with some DISTINCT queries
in MySQL
*/
if ((keydef->flag & (HA_NOSAME | HA_NULL_PART_KEY)) == HA_NOSAME &&
key_segs)
Import changeset
2000-07-31 21:29:14 +02:00
share.state.rec_per_key_part[key_segs-1]=1L;
length+=key_length;
Added support for key_block_size for key and table level (WL#602) Added support for key_block_size to MyISAM. Simplify interface to 'new Key' to make it easier to add new key options. mysqld option --new is used to define where key options are printed. (In 5.3 we should move all key options to after key part definition to avoid problem with reserved names) Fixed some compiler warnings and a memory leak in ssl
2006-05-03 14:59:17 +02:00
/* Get block length for key, if defined by user */
block_length= (keydef->block_length ?
my_round_up_to_next_power(keydef->block_length) :
myisam_block_size);
block_length= max(block_length, MI_MIN_KEY_BLOCK_LENGTH);
block_length= min(block_length, MI_MAX_KEY_BLOCK_LENGTH);
Fixed portability issue in my_thr_init.c (was added in my last push) Fixed compiler warnings (detected by VC++): - Removed not used variables - Added casts - Fixed wrong assignments to bool - Fixed wrong calls with bool arguments - Added missing argument to store(longlong), which caused wrong store method to be called.
2006-11-30 17:25:05 +01:00
keydef->block_length= (uint16) MI_BLOCK_SIZE(length-real_length_diff,
Merge mysql.com:/home/my/mysql-5.0 into mysql.com:/home/my/mysql-5.1
2006-11-30 21:38:12 +01:00
pointer,MI_MAX_KEYPTR_SIZE,
block_length);
max_key_length MyISAM bugfix
2002-09-10 21:40:01 +02:00
if (keydef->block_length > MI_MAX_KEY_BLOCK_LENGTH ||
less restrictive test for max key length in mi_create
2002-10-24 00:14:59 +02:00
length >= MI_MAX_KEY_BUFF)
Import changeset
2000-07-31 21:29:14 +02:00
{
my_errno=HA_WRONG_CREATE_OPTION;
goto err;
}
set_if_bigger(max_key_block_length,keydef->block_length);
keydef->keylength= (uint16) key_length;
just tried to find all 'skipp' and replace it with 'skip'.
2004-02-02 17:25:39 +01:00
keydef->minlength= (uint16) (length-min_key_length_skip);
Import changeset
2000-07-31 21:29:14 +02:00
keydef->maxlength= (uint16) length;
if (length > max_key_length)
max_key_length= length;
tot_length+= (max_rows/(ulong) (((uint) keydef->block_length-5)/
(length*2)))*
(ulong) keydef->block_length;
}
Cleaned up SSL documentation Fixes for embedded server Made key_cache more configurable Fixed that one can change key blocksize in MyISAM A lot of optimizations to make MyISAM slightly faster
2001-10-02 04:53:00 +02:00
for (i=max_key_block_length/MI_MIN_KEY_BLOCK_LENGTH ; i-- ; )
Import changeset
2000-07-31 21:29:14 +02:00
key_del[i]=HA_OFFSET_ERROR;
unique_key_parts=0;
offset=reclength-uniques*MI_UNIQUE_HASH_LENGTH;
for (i=0, uniquedef=uniquedefs ; i < uniques ; i++ , uniquedef++)
{
uniquedef->key=keys+i;
unique_key_parts+=uniquedef->keysegs;
share.state.key_root[keys+i]= HA_OFFSET_ERROR;
bugfix: uniques did not contributed to MYI tot_length estimation, which resulted in "Index file full" error on huge SELECT DISTINCT's
2002-07-24 21:56:33 +02:00
tot_length+= (max_rows/(ulong) (((uint) myisam_block_size-5)/
((MI_UNIQUE_HASH_LENGTH + pointer)*2)))*
(ulong) myisam_block_size;
Import changeset
2000-07-31 21:29:14 +02:00
}
keys+=uniques; /* Each unique has 1 key */
key_segs+=uniques; /* Each unique has 1 key seg */
base_pos=(MI_STATE_INFO_SIZE + keys * MI_STATE_KEY_SIZE +
Cleaned up SSL documentation Fixes for embedded server Made key_cache more configurable Fixed that one can change key blocksize in MyISAM A lot of optimizations to make MyISAM slightly faster
2001-10-02 04:53:00 +02:00
max_key_block_length/MI_MIN_KEY_BLOCK_LENGTH*
MI_STATE_KEYBLOCK_SIZE+
Import changeset
2000-07-31 21:29:14 +02:00
key_segs*MI_STATE_KEYSEG_SIZE);
info_length=base_pos+(uint) (MI_BASE_INFO_SIZE+
keys * MI_KEYDEF_SIZE+
uniques * MI_UNIQUEDEF_SIZE +
MI_KEYSEG -> HA_KEYSEG _mi_key_cmp -> ha_key_cmp
2002-04-25 12:10:29 +02:00
(key_segs + unique_key_parts)*HA_KEYSEG_SIZE+
Import changeset
2000-07-31 21:29:14 +02:00
columns*MI_COLUMNDEF_SIZE);
Bug#11824 - internal /tmp/*.{MYD,MYI} files remain, causing subsequent queries to fail Very complex select statements can create temporary tables that are too big to be represented as a MyISAM table. This was not checked at table creation time, but only at open time. The result was an attempt to delete the "impossible" table. But if the server is built --with-raid, MyISAM tries to open the table before deleting the files. It needs to find out if the table uses the raid support and how many raid chunks there are. This is done with an open "for repair", which will almost always succeed. But in this case we have an "impossible" table. The open failed. Hence the files were not deleted. Also the error message was a bit unspecific. I turned an open error in this situation into the assumption of having no raid support on the table. Thus the normal data file is tried to be deleted. This may however leave existing raid chunks behind. I also added a check in mi_create() to prevent the creation of an "impossible" table. A more decriptive error message is given in this case. No test case. The required select statement is way too large for the test suite. I added a test script to the bug report.
2006-06-27 11:26:41 +02:00
DBUG_PRINT("info", ("info_length: %u", info_length));
/* There are only 16 bits for the total header length. */
if (info_length > 65535)
{
my_printf_error(0, "MyISAM table '%s' has too many columns and/or "
"indexes and/or unique constraints.",
MYF(0), name + dirname_length(name));
my_errno= HA_WRONG_CREATE_OPTION;
goto err;
}
Import changeset
2000-07-31 21:29:14 +02:00
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
bmove(share.state.header.file_version,(uchar*) myisam_file_magic,4);
Import changeset
2000-07-31 21:29:14 +02:00
ci->old_options=options| (ci->old_options & HA_OPTION_TEMP_COMPRESS_RECORD ?
HA_OPTION_COMPRESS_RECORD |
HA_OPTION_TEMP_COMPRESS_RECORD: 0);
mi_int2store(share.state.header.options,ci->old_options);
mi_int2store(share.state.header.header_length,info_length);
mi_int2store(share.state.header.state_info_length,MI_STATE_INFO_SIZE);
mi_int2store(share.state.header.base_info_length,MI_BASE_INFO_SIZE);
mi_int2store(share.state.header.base_pos,base_pos);
share.state.header.language= (ci->language ?
MY_CS_NONTEXT flag for sprintf() incompatible charsets (ucs2 if the only one now) Some default_charset_info were removed
2003-06-03 11:59:17 +02:00
ci->language : default_charset_info->number);
Added support for key_block_size for key and table level (WL#602) Added support for key_block_size to MyISAM. Simplify interface to 'new Key' to make it easier to add new key options. mysqld option --new is used to define where key options are printed. (In 5.3 we should move all key options to after key part definition to avoid problem with reserved names) Fixed some compiler warnings and a memory leak in ssl
2006-05-03 14:59:17 +02:00
share.state.header.max_block_size_index= max_key_block_length/MI_MIN_KEY_BLOCK_LENGTH;
Import changeset
2000-07-31 21:29:14 +02:00
share.state.dellink = HA_OFFSET_ERROR;
share.state.process= (ulong) getpid();
share.state.unique= (ulong) 0;
changed to use IO_CACHE instead of FILE
2000-11-16 02:58:58 +01:00
share.state.update_count=(ulong) 0;
Import changeset
2000-07-31 21:29:14 +02:00
share.state.version= (ulong) time((time_t*) 0);
share.state.sortkey= (ushort) ~0;
share.state.auto_increment=ci->auto_increment;
share.options=options;
share.base.rec_reflength=pointer;
Fix for BUG#4785: * myisampack leaves key_file_length value from original table * myisamchk uses this value when calculating key file pointer length
2004-09-22 14:29:15 +02:00
/* Get estimate for index file length (this may be wrong for FT keys) */
tmp= (tot_length + max_key_block_length * keys *
MI_INDEX_BLOCK_MARGIN) / MI_MIN_KEY_BLOCK_LENGTH;
/*
use maximum of key_file_length we calculated and key_file_length value we
got from MYI file header (see also myisampack.c:save_state)
*/
Import changeset
2000-07-31 21:29:14 +02:00
share.base.key_reflength=
Fix for BUG#4785: * myisampack leaves key_file_length value from original table * myisamchk uses this value when calculating key file pointer length
2004-09-22 14:29:15 +02:00
mi_get_pointer_length(max(ci->key_file_length,tmp),3);
Two-level index structure for FULLTEXT indexes
2003-01-21 19:24:34 +01:00
share.base.keys= share.state.header.keys= keys;
Import changeset
2000-07-31 21:29:14 +02:00
share.state.header.uniques= uniques;
Two-level index structure for FULLTEXT indexes
2003-01-21 19:24:34 +01:00
share.state.header.fulltext_keys= fulltext_keys;
Import changeset
2000-07-31 21:29:14 +02:00
mi_int2store(share.state.header.key_parts,key_segs);
mi_int2store(share.state.header.unique_key_parts,unique_key_parts);
Bug#10932 - Building server with key limit of 128, makes test cases fail This patch allows to configure MyISAM for 128 indexes per table. The main problem is the key_map, wich is implemented as an ulonglong. To get rid of the limit and keep the efficient and flexible implementation, the highest bit is now used for all upper keys. This means that the lower keys can be disabled and enabled individually as usual and the high keys can only be disabled and enabled as a block. That way the existing test suite is still applicable, while more keys work, though slightly less efficient. To really get more than 64 keys, some defines need to be changed. Another patch will address this.
2005-07-19 14:13:56 +02:00
mi_set_all_keys_active(share.state.key_map, keys);
Added support for key_block_size for key and table level (WL#602) Added support for key_block_size to MyISAM. Simplify interface to 'new Key' to make it easier to add new key options. mysqld option --new is used to define where key options are printed. (In 5.3 we should move all key options to after key part definition to avoid problem with reserved names) Fixed some compiler warnings and a memory leak in ssl
2006-05-03 14:59:17 +02:00
aligned_key_start= my_round_up_to_next_power(max_key_block_length ?
max_key_block_length :
myisam_block_size);
share.base.keystart= share.state.state.key_file_length=
MY_ALIGN(info_length, aligned_key_start);
Import changeset
2000-07-31 21:29:14 +02:00
share.base.max_key_block_length=max_key_block_length;
share.base.max_key_length=ALIGN_SIZE(max_key_length+4);
share.base.records=ci->max_rows;
share.base.reloc= ci->reloc_rows;
share.base.reclength=real_reclength;
Fixed length rows should not use row checksum in MyISAM tables Fixed bug in handling of packed MyISAM keys
2001-06-26 09:23:50 +02:00
share.base.pack_reclength=reclength+ test(options & HA_OPTION_CHECKSUM);
Import changeset
2000-07-31 21:29:14 +02:00
share.base.max_pack_length=pack_reclength;
share.base.min_pack_length=min_pack_length;
share.base.pack_bits=packed;
share.base.fields=fields;
share.base.pack_fields=packed;
#ifdef USE_RAID
share.base.raid_type=ci->raid_type;
share.base.raid_chunks=ci->raid_chunks;
share.base.raid_chunksize=ci->raid_chunksize;
#endif
/* max_data_file_length and max_key_file_length are recalculated on open */
if (options & HA_OPTION_TMP_TABLE)
share.base.max_data_file_length=(my_off_t) ci->data_file_length;
share.base.min_block_length=
(share.base.pack_reclength+3 < MI_EXTEND_BLOCK_LENGTH &&
! share.base.blobs) ?
max(share.base.pack_reclength,MI_MIN_BLOCK_LENGTH) :
MI_EXTEND_BLOCK_LENGTH;
if (! (flags & HA_DONT_TOUCH_DATA))
share.state.create_time= (long) time((time_t*) 0);
Update copyright Fixed memory leak on shutdown (Affects the embedded version & MyODBC)
2001-12-06 13:10:51 +01:00
locking issues, test for LONGLONG_MIN in decimal.c
2005-12-23 20:50:28 +01:00
pthread_mutex_lock(&THR_LOCK_myisam);
Bug#26782 - Patch: myisamchk -rq creates .MYI.MYI file on packed MyISAM tables After merge fix In conjunction with the newest 5.1 we always need to create a real path name to be able to detect already open tables.
2007-03-23 10:26:14 +01:00
/*
NOTE: For test_if_reopen() we need a real path name. Hence we need
MY_RETURN_REAL_PATH for every fn_format(filename, ...).
*/
Symlinking of tables should not work. Fixed error number handling bug in mysqltest. Fixed that error number from insert delayed is reported correctly. merged new vio code with old violite code.
2001-06-05 02:38:10 +02:00
if (ci->index_file_name)
{
WL#1324 table name to file name encoding - Encoding itself, implemented as a charset "filename". Originally planned to use '.' as an escape character, but now changed to '@' for two reasons: "ls" does not return file names starting with '.' considering them as a kind of hidden files; some platforms do not allow several dots in a file name. - replacing many calls of my_snprintf() and strnxmov() to the new build_table_filename(). - Adding MY_APPEND_EXT mysys flag, to append an extention rather that replace it. - Replacing all numeric constants in fn_format flag arguments to their mysys definitions, e.g. MY_UNPACK_FILENAME, - Predictability in several function/methods: when a table name can appear with or withot .frm extension. Some functions/methods were changed so accept names strictly with .frm, other - strictly without .frm extensions. Several DBUG_ASSERTs were added to check whether an extension is passed. Many files: table name to file name encoding mysql_priv.h: Prototypes for new table name encoding tools. ctype-utf8.c: Implementing "filename" charset for table name to file name encoding. row0mysql.c: Fixing table name prefix. mf_format.c: Adding MY_APPEND_EXT processing. Many files: Fixing tests. my_sys.h: Adding new flag to append rather than replace an extension. m_ctype.h: Adding "filename" charset definition.
2005-12-31 06:01:26 +01:00
char *iext= strrchr(ci->index_file_name, '.');
int have_iext= iext && !strcmp(iext, MI_NAME_IEXT);
Bug#8706 "temporary table with data directory option fails" myisam should not use user-specified table name when creating temporary tables and use generated connection specific real name. Test included.
2006-07-06 02:18:59 +02:00
if (options & HA_OPTION_TMP_TABLE)
{
char *path;
/* chop off the table name, tempory tables use generated name */
if ((path= strrchr(ci->index_file_name, FN_LIBCHAR)))
*path= '\0';
fn_format(filename, name, ci->index_file_name, MI_NAME_IEXT,
Bug#26782 - Patch: myisamchk -rq creates .MYI.MYI file on packed MyISAM tables After merge fix In conjunction with the newest 5.1 we always need to create a real path name to be able to detect already open tables.
2007-03-23 10:26:14 +01:00
MY_REPLACE_DIR | MY_UNPACK_FILENAME |
MY_RETURN_REAL_PATH | MY_APPEND_EXT);
Bug#8706 "temporary table with data directory option fails" myisam should not use user-specified table name when creating temporary tables and use generated connection specific real name. Test included.
2006-07-06 02:18:59 +02:00
}
else
Merge bodhi.local:/opt/local/work/tmp_merge into bodhi.local:/opt/local/work/mysql-5.1-runtime-merge
2006-07-26 21:33:25 +02:00
{
fn_format(filename, ci->index_file_name, "", MI_NAME_IEXT,
Bug#26782 - Patch: myisamchk -rq creates .MYI.MYI file on packed MyISAM tables After merge fix In conjunction with the newest 5.1 we always need to create a real path name to be able to detect already open tables.
2007-03-23 10:26:14 +01:00
MY_UNPACK_FILENAME | MY_RETURN_REAL_PATH |
(have_iext ? MY_REPLACE_EXT : MY_APPEND_EXT));
Merge bodhi.local:/opt/local/work/tmp_merge into bodhi.local:/opt/local/work/mysql-5.1-runtime-merge
2006-07-26 21:33:25 +02:00
}
WL#1324 table name to file name encoding - Encoding itself, implemented as a charset "filename". Originally planned to use '.' as an escape character, but now changed to '@' for two reasons: "ls" does not return file names starting with '.' considering them as a kind of hidden files; some platforms do not allow several dots in a file name. - replacing many calls of my_snprintf() and strnxmov() to the new build_table_filename(). - Adding MY_APPEND_EXT mysys flag, to append an extention rather that replace it. - Replacing all numeric constants in fn_format flag arguments to their mysys definitions, e.g. MY_UNPACK_FILENAME, - Predictability in several function/methods: when a table name can appear with or withot .frm extension. Some functions/methods were changed so accept names strictly with .frm, other - strictly without .frm extensions. Several DBUG_ASSERTs were added to check whether an extension is passed. Many files: table name to file name encoding mysql_priv.h: Prototypes for new table name encoding tools. ctype-utf8.c: Implementing "filename" charset for table name to file name encoding. row0mysql.c: Fixing table name prefix. mf_format.c: Adding MY_APPEND_EXT processing. Many files: Fixing tests. my_sys.h: Adding new flag to append rather than replace an extension. m_ctype.h: Adding "filename" charset definition.
2005-12-31 06:01:26 +01:00
fn_format(linkname, name, "", MI_NAME_IEXT,
MY_UNPACK_FILENAME|MY_APPEND_EXT);
Symlinking of tables should not work. Fixed error number handling bug in mysqltest. Fixed that error number from insert delayed is reported correctly. merged new vio code with old violite code.
2001-06-05 02:38:10 +02:00
linkname_ptr=linkname;
/*
Don't create the table if the link or file exists to ensure that one
doesn't accidently destroy another table.
*/
create_flag=0;
}
else
{
Bug#26782 - Patch: myisamchk -rq creates .MYI.MYI file on packed MyISAM tables When fixing the indexes with "myisamchk -rq" after compressing the table with "myisampack", an optionally supplied extension ".MYI" of the index file was not detected. The extension was appended unconditionally. The result was ".MYI.MYI". Now an extension is no longer appended if present already. Thanks to David Shrewsbury for providing this patch. Another problem was a misplaced parenthesis. We did never unpack the file name ("~/..") and always returned a real path. No test case. This is manually tested with the utilities "myisampack" and "myisamchk".
2007-03-07 16:30:13 +01:00
char *iext= strrchr(name, '.');
int have_iext= iext && !strcmp(iext, MI_NAME_IEXT);
WL#1324 table name to file name encoding - Encoding itself, implemented as a charset "filename". Originally planned to use '.' as an escape character, but now changed to '@' for two reasons: "ls" does not return file names starting with '.' considering them as a kind of hidden files; some platforms do not allow several dots in a file name. - replacing many calls of my_snprintf() and strnxmov() to the new build_table_filename(). - Adding MY_APPEND_EXT mysys flag, to append an extention rather that replace it. - Replacing all numeric constants in fn_format flag arguments to their mysys definitions, e.g. MY_UNPACK_FILENAME, - Predictability in several function/methods: when a table name can appear with or withot .frm extension. Some functions/methods were changed so accept names strictly with .frm, other - strictly without .frm extensions. Several DBUG_ASSERTs were added to check whether an extension is passed. Many files: table name to file name encoding mysql_priv.h: Prototypes for new table name encoding tools. ctype-utf8.c: Implementing "filename" charset for table name to file name encoding. row0mysql.c: Fixing table name prefix. mf_format.c: Adding MY_APPEND_EXT processing. Many files: Fixing tests. my_sys.h: Adding new flag to append rather than replace an extension. m_ctype.h: Adding "filename" charset definition.
2005-12-31 06:01:26 +01:00
fn_format(filename, name, "", MI_NAME_IEXT,
Bug#26782 - Patch: myisamchk -rq creates .MYI.MYI file on packed MyISAM tables After merge fix In conjunction with the newest 5.1 we always need to create a real path name to be able to detect already open tables.
2007-03-23 10:26:14 +01:00
MY_UNPACK_FILENAME | MY_RETURN_REAL_PATH |
Bug#26782 - Patch: myisamchk -rq creates .MYI.MYI file on packed MyISAM tables When fixing the indexes with "myisamchk -rq" after compressing the table with "myisampack", an optionally supplied extension ".MYI" of the index file was not detected. The extension was appended unconditionally. The result was ".MYI.MYI". Now an extension is no longer appended if present already. Thanks to David Shrewsbury for providing this patch. Another problem was a misplaced parenthesis. We did never unpack the file name ("~/..") and always returned a real path. No test case. This is manually tested with the utilities "myisampack" and "myisamchk".
2007-03-07 16:30:13 +01:00
(have_iext ? MY_REPLACE_EXT : MY_APPEND_EXT));
Symlinking of tables should not work. Fixed error number handling bug in mysqltest. Fixed that error number from insert delayed is reported correctly. merged new vio code with old violite code.
2001-06-05 02:38:10 +02:00
linkname_ptr=0;
/* Replace the current file */
fixed uninitialized variable introduced by the fix for bug 29325
2007-07-11 11:37:47 +02:00
create_flag=(flags & HA_CREATE_KEEP_FILES) ? 0 : MY_DELETE_OLD;
Symlinking of tables should not work. Fixed error number handling bug in mysqltest. Fixed that error number from insert delayed is reported correctly. merged new vio code with old violite code.
2001-06-05 02:38:10 +02:00
}
Import changeset
2000-07-31 21:29:14 +02:00
Bug#8306 - TRUNCATE leads to index corruption Added a check, if the table, which we are going to create, is open. This can happen if a MERGE mapped table is TRUNCATEd.
2005-03-02 10:35:00 +01:00
/*
If a MRG_MyISAM table is in use, the mapped MyISAM tables are open,
but no entry is made in the table cache for them.
A TRUNCATE command checks for the table in the cache only and could
be fooled to believe, the table is not open.
Pull the emergency brake in this situation. (Bug #8306)
Bug#26782 - Patch: myisamchk -rq creates .MYI.MYI file on packed MyISAM tables After merge fix In conjunction with the newest 5.1 we always need to create a real path name to be able to detect already open tables.
2007-03-23 10:26:14 +01:00
NOTE: The filename is compared against unique_file_name of every
open table. Hence we need a real path here.
Bug#8306 - TRUNCATE leads to index corruption Added a check, if the table, which we are going to create, is open. This can happen if a MERGE mapped table is TRUNCATEd.
2005-03-02 10:35:00 +01:00
*/
if (test_if_reopen(filename))
{
my_printf_error(0, "MyISAM table '%s' is in use "
"(most likely by a MERGE table). Try FLUSH TABLES.",
MYF(0), name + dirname_length(name));
goto err;
}
Fixes for bugs reported by Stefano Di Paola (stefano.dipaola@wisec.it)
2005-03-03 19:51:29 +01:00
if ((file= my_create_with_symlink(linkname_ptr, filename, 0, create_mode,
Symlinking of tables should not work. Fixed error number handling bug in mysqltest. Fixed that error number from insert delayed is reported correctly. merged new vio code with old violite code.
2001-06-05 02:38:10 +02:00
MYF(MY_WME | create_flag))) < 0)
Import changeset
2000-07-31 21:29:14 +02:00
goto err;
errpos=1;
Removed compiler warnings. Added preliminary handling of symlinks in MyISAM. When using myisamchk to check tables with --force, don't repair tables that are marked as 'not closed' if they are ok. Change fn_format() to use my_real_path and my_symlink
2001-05-31 13:07:17 +02:00
Import changeset
2000-07-31 21:29:14 +02:00
if (!(flags & HA_DONT_TOUCH_DATA))
{
#ifdef USE_RAID
if (share.base.raid_type)
{
WL#1324 table name to file name encoding - Encoding itself, implemented as a charset "filename". Originally planned to use '.' as an escape character, but now changed to '@' for two reasons: "ls" does not return file names starting with '.' considering them as a kind of hidden files; some platforms do not allow several dots in a file name. - replacing many calls of my_snprintf() and strnxmov() to the new build_table_filename(). - Adding MY_APPEND_EXT mysys flag, to append an extention rather that replace it. - Replacing all numeric constants in fn_format flag arguments to their mysys definitions, e.g. MY_UNPACK_FILENAME, - Predictability in several function/methods: when a table name can appear with or withot .frm extension. Some functions/methods were changed so accept names strictly with .frm, other - strictly without .frm extensions. Several DBUG_ASSERTs were added to check whether an extension is passed. Many files: table name to file name encoding mysql_priv.h: Prototypes for new table name encoding tools. ctype-utf8.c: Implementing "filename" charset for table name to file name encoding. row0mysql.c: Fixing table name prefix. mf_format.c: Adding MY_APPEND_EXT processing. Many files: Fixing tests. my_sys.h: Adding new flag to append rather than replace an extension. m_ctype.h: Adding "filename" charset definition.
2005-12-31 06:01:26 +01:00
(void) fn_format(filename, name, "", MI_NAME_DEXT,
MY_UNPACK_FILENAME | MY_APPEND_EXT);
Fix for a build bug. Added a missing comma.
2005-03-04 09:30:22 +01:00
if ((dfile=my_raid_create(filename, 0, create_mode,
Import changeset
2000-07-31 21:29:14 +02:00
share.base.raid_type,
share.base.raid_chunks,
share.base.raid_chunksize,
MYF(MY_WME | MY_RAID))) < 0)
goto err;
}
else
#endif
Removed compiler warnings. Added preliminary handling of symlinks in MyISAM. When using myisamchk to check tables with --force, don't repair tables that are marked as 'not closed' if they are ok. Change fn_format() to use my_real_path and my_symlink
2001-05-31 13:07:17 +02:00
{
Symlinking of tables should not work. Fixed error number handling bug in mysqltest. Fixed that error number from insert delayed is reported correctly. merged new vio code with old violite code.
2001-06-05 02:38:10 +02:00
if (ci->data_file_name)
{
WL#1324 table name to file name encoding - Encoding itself, implemented as a charset "filename". Originally planned to use '.' as an escape character, but now changed to '@' for two reasons: "ls" does not return file names starting with '.' considering them as a kind of hidden files; some platforms do not allow several dots in a file name. - replacing many calls of my_snprintf() and strnxmov() to the new build_table_filename(). - Adding MY_APPEND_EXT mysys flag, to append an extention rather that replace it. - Replacing all numeric constants in fn_format flag arguments to their mysys definitions, e.g. MY_UNPACK_FILENAME, - Predictability in several function/methods: when a table name can appear with or withot .frm extension. Some functions/methods were changed so accept names strictly with .frm, other - strictly without .frm extensions. Several DBUG_ASSERTs were added to check whether an extension is passed. Many files: table name to file name encoding mysql_priv.h: Prototypes for new table name encoding tools. ctype-utf8.c: Implementing "filename" charset for table name to file name encoding. row0mysql.c: Fixing table name prefix. mf_format.c: Adding MY_APPEND_EXT processing. Many files: Fixing tests. my_sys.h: Adding new flag to append rather than replace an extension. m_ctype.h: Adding "filename" charset definition.
2005-12-31 06:01:26 +01:00
char *dext= strrchr(ci->data_file_name, '.');
int have_dext= dext && !strcmp(dext, MI_NAME_DEXT);
Bug#8706 "temporary table with data directory option fails" myisam should not use user-specified table name when creating temporary tables and use generated connection specific real name. Test included.
2006-07-06 02:18:59 +02:00
if (options & HA_OPTION_TMP_TABLE)
{
char *path;
/* chop off the table name, tempory tables use generated name */
if ((path= strrchr(ci->data_file_name, FN_LIBCHAR)))
*path= '\0';
fn_format(filename, name, ci->data_file_name, MI_NAME_DEXT,
Post-merge fixes.
2006-07-27 12:12:41 +02:00
MY_REPLACE_DIR | MY_UNPACK_FILENAME | MY_APPEND_EXT);
Bug#8706 "temporary table with data directory option fails" myisam should not use user-specified table name when creating temporary tables and use generated connection specific real name. Test included.
2006-07-06 02:18:59 +02:00
}
else
Merge bodhi.local:/opt/local/work/tmp_merge into bodhi.local:/opt/local/work/mysql-5.1-runtime-merge
2006-07-26 21:33:25 +02:00
{
fn_format(filename, ci->data_file_name, "", MI_NAME_DEXT,
MY_UNPACK_FILENAME |
(have_dext ? MY_REPLACE_EXT : MY_APPEND_EXT));
}
WL#1324 table name to file name encoding - Encoding itself, implemented as a charset "filename". Originally planned to use '.' as an escape character, but now changed to '@' for two reasons: "ls" does not return file names starting with '.' considering them as a kind of hidden files; some platforms do not allow several dots in a file name. - replacing many calls of my_snprintf() and strnxmov() to the new build_table_filename(). - Adding MY_APPEND_EXT mysys flag, to append an extention rather that replace it. - Replacing all numeric constants in fn_format flag arguments to their mysys definitions, e.g. MY_UNPACK_FILENAME, - Predictability in several function/methods: when a table name can appear with or withot .frm extension. Some functions/methods were changed so accept names strictly with .frm, other - strictly without .frm extensions. Several DBUG_ASSERTs were added to check whether an extension is passed. Many files: table name to file name encoding mysql_priv.h: Prototypes for new table name encoding tools. ctype-utf8.c: Implementing "filename" charset for table name to file name encoding. row0mysql.c: Fixing table name prefix. mf_format.c: Adding MY_APPEND_EXT processing. Many files: Fixing tests. my_sys.h: Adding new flag to append rather than replace an extension. m_ctype.h: Adding "filename" charset definition.
2005-12-31 06:01:26 +01:00
fn_format(linkname, name, "",MI_NAME_DEXT,
MY_UNPACK_FILENAME | MY_APPEND_EXT);
Symlinking of tables should not work. Fixed error number handling bug in mysqltest. Fixed that error number from insert delayed is reported correctly. merged new vio code with old violite code.
2001-06-05 02:38:10 +02:00
linkname_ptr=linkname;
create_flag=0;
}
else
{
WL#1324 table name to file name encoding - Encoding itself, implemented as a charset "filename". Originally planned to use '.' as an escape character, but now changed to '@' for two reasons: "ls" does not return file names starting with '.' considering them as a kind of hidden files; some platforms do not allow several dots in a file name. - replacing many calls of my_snprintf() and strnxmov() to the new build_table_filename(). - Adding MY_APPEND_EXT mysys flag, to append an extention rather that replace it. - Replacing all numeric constants in fn_format flag arguments to their mysys definitions, e.g. MY_UNPACK_FILENAME, - Predictability in several function/methods: when a table name can appear with or withot .frm extension. Some functions/methods were changed so accept names strictly with .frm, other - strictly without .frm extensions. Several DBUG_ASSERTs were added to check whether an extension is passed. Many files: table name to file name encoding mysql_priv.h: Prototypes for new table name encoding tools. ctype-utf8.c: Implementing "filename" charset for table name to file name encoding. row0mysql.c: Fixing table name prefix. mf_format.c: Adding MY_APPEND_EXT processing. Many files: Fixing tests. my_sys.h: Adding new flag to append rather than replace an extension. m_ctype.h: Adding "filename" charset definition.
2005-12-31 06:01:26 +01:00
fn_format(filename,name,"", MI_NAME_DEXT,
MY_UNPACK_FILENAME | MY_APPEND_EXT);
Symlinking of tables should not work. Fixed error number handling bug in mysqltest. Fixed that error number from insert delayed is reported correctly. merged new vio code with old violite code.
2001-06-05 02:38:10 +02:00
linkname_ptr=0;
fixed uninitialized variable introduced by the fix for bug 29325
2007-07-11 11:37:47 +02:00
create_flag=(flags & HA_CREATE_KEEP_FILES) ? 0 : MY_DELETE_OLD;
Symlinking of tables should not work. Fixed error number handling bug in mysqltest. Fixed that error number from insert delayed is reported correctly. merged new vio code with old violite code.
2001-06-05 02:38:10 +02:00
}
Two-level index structure for FULLTEXT indexes
2003-01-21 19:24:34 +01:00
if ((dfile=
Fixes for bugs reported by Stefano Di Paola (stefano.dipaola@wisec.it)
2005-03-03 19:51:29 +01:00
my_create_with_symlink(linkname_ptr, filename, 0, create_mode,
Symlinking of tables should not work. Fixed error number handling bug in mysqltest. Fixed that error number from insert delayed is reported correctly. merged new vio code with old violite code.
2001-06-05 02:38:10 +02:00
MYF(MY_WME | create_flag))) < 0)
Removed compiler warnings. Added preliminary handling of symlinks in MyISAM. When using myisamchk to check tables with --force, don't repair tables that are marked as 'not closed' if they are ok. Change fn_format() to use my_real_path and my_symlink
2001-05-31 13:07:17 +02:00
goto err;
}
Import changeset
2000-07-31 21:29:14 +02:00
errpos=3;
}
Bug#11824 - internal /tmp/*.{MYD,MYI} files remain, causing subsequent queries to fail Very complex select statements can create temporary tables that are too big to be represented as a MyISAM table. This was not checked at table creation time, but only at open time. The result was an attempt to delete the "impossible" table. But if the server is built --with-raid, MyISAM tries to open the table before deleting the files. It needs to find out if the table uses the raid support and how many raid chunks there are. This is done with an open "for repair", which will almost always succeed. But in this case we have an "impossible" table. The open failed. Hence the files were not deleted. Also the error message was a bit unspecific. I turned an open error in this situation into the assumption of having no raid support on the table. Thus the normal data file is tried to be deleted. This may however leave existing raid chunks behind. I also added a check in mi_create() to prevent the creation of an "impossible" table. A more decriptive error message is given in this case. No test case. The required select statement is way too large for the test suite. I added a test script to the bug report.
2006-06-27 11:26:41 +02:00
DBUG_PRINT("info", ("write state info and base info"));
Import changeset
2000-07-31 21:29:14 +02:00
if (mi_state_info_write(file, &share.state, 2) ||
mi_base_info_write(file, &share.base))
goto err;
#ifndef DBUG_OFF
if ((uint) my_tell(file,MYF(0)) != base_pos+ MI_BASE_INFO_SIZE)
{
uint pos=(uint) my_tell(file,MYF(0));
DBUG_PRINT("warning",("base_length: %d != used_length: %d",
base_pos+ MI_BASE_INFO_SIZE, pos));
}
#endif
/* Write key and keyseg definitions */
Bug#11824 - internal /tmp/*.{MYD,MYI} files remain, causing subsequent queries to fail Very complex select statements can create temporary tables that are too big to be represented as a MyISAM table. This was not checked at table creation time, but only at open time. The result was an attempt to delete the "impossible" table. But if the server is built --with-raid, MyISAM tries to open the table before deleting the files. It needs to find out if the table uses the raid support and how many raid chunks there are. This is done with an open "for repair", which will almost always succeed. But in this case we have an "impossible" table. The open failed. Hence the files were not deleted. Also the error message was a bit unspecific. I turned an open error in this situation into the assumption of having no raid support on the table. Thus the normal data file is tried to be deleted. This may however leave existing raid chunks behind. I also added a check in mi_create() to prevent the creation of an "impossible" table. A more decriptive error message is given in this case. No test case. The required select statement is way too large for the test suite. I added a test script to the bug report.
2006-06-27 11:26:41 +02:00
DBUG_PRINT("info", ("write key and keyseg definitions"));
Import changeset
2000-07-31 21:29:14 +02:00
for (i=0 ; i < share.base.keys - uniques; i++)
{
This ChangeSet adds RTREE support into myisam library. RTREEs will be used for GIS extension in MySQL
2002-02-20 11:11:21 +01:00
uint sp_segs=(keydefs[i].flag & HA_SPATIAL) ? 2*SPDIMS : 0;
Import changeset
2000-07-31 21:29:14 +02:00
if (mi_keydef_write(file, &keydefs[i]))
goto err;
Two-level index structure for FULLTEXT indexes
2003-01-21 19:24:34 +01:00
for (j=0 ; j < keydefs[i].keysegs-sp_segs ; j++)
Import changeset
2000-07-31 21:29:14 +02:00
if (mi_keyseg_write(file, &keydefs[i].seg[j]))
This ChangeSet adds RTREE support into myisam library. RTREEs will be used for GIS extension in MySQL
2002-02-20 11:11:21 +01:00
goto err;
WL#1163 To make spatial code optional myisam spatial code isolated
2004-03-12 15:51:03 +01:00
#ifdef HAVE_SPATIAL
This ChangeSet adds RTREE support into myisam library. RTREEs will be used for GIS extension in MySQL
2002-02-20 11:11:21 +01:00
for (j=0 ; j < sp_segs ; j++)
{
MI_KEYSEG -> HA_KEYSEG _mi_key_cmp -> ha_key_cmp
2002-04-25 12:10:29 +02:00
HA_KEYSEG sseg;
This ChangeSet adds RTREE support into myisam library. RTREEs will be used for GIS extension in MySQL
2002-02-20 11:11:21 +01:00
sseg.type=SPTYPE;
Fixed memory reference errors found by valgrind
2005-01-12 02:38:53 +01:00
sseg.language= 7; /* Binary */
This ChangeSet adds RTREE support into myisam library. RTREEs will be used for GIS extension in MySQL
2002-02-20 11:11:21 +01:00
sseg.null_bit=0;
sseg.bit_start=0;
sseg.bit_end=0;
Fixed memory reference errors found by valgrind
2005-01-12 02:38:53 +01:00
sseg.bit_length= 0;
sseg.bit_pos= 0;
This ChangeSet adds RTREE support into myisam library. RTREEs will be used for GIS extension in MySQL
2002-02-20 11:11:21 +01:00
sseg.length=SPLEN;
sseg.null_pos=0;
sseg.start=j*SPLEN;
sseg.flag= HA_SWAP_KEY;
if (mi_keyseg_write(file, &sseg))
goto err;
}
WL#1163 To make spatial code optional myisam spatial code isolated
2004-03-12 15:51:03 +01:00
#endif
Import changeset
2000-07-31 21:29:14 +02:00
}
/* Create extra keys for unique definitions */
offset=reclength-uniques*MI_UNIQUE_HASH_LENGTH;
bzero((char*) &tmp_keydef,sizeof(tmp_keydef));
bzero((char*) &tmp_keyseg,sizeof(tmp_keyseg));
for (i=0; i < uniques ; i++)
{
tmp_keydef.keysegs=1;
tmp_keydef.flag= HA_UNIQUE_CHECK;
fixes for compiler warnings from VC6
2005-06-06 18:34:52 +02:00
tmp_keydef.block_length= (uint16)myisam_block_size;
Import changeset
2000-07-31 21:29:14 +02:00
tmp_keydef.keylength= MI_UNIQUE_HASH_LENGTH + pointer;
tmp_keydef.minlength=tmp_keydef.maxlength=tmp_keydef.keylength;
bugfix: uniques did not contributed to MYI tot_length estimation, which resulted in "Index file full" error on huge SELECT DISTINCT's
2002-07-24 21:56:33 +02:00
tmp_keyseg.type= MI_UNIQUE_HASH_TYPE;
tmp_keyseg.length= MI_UNIQUE_HASH_LENGTH;
Import changeset
2000-07-31 21:29:14 +02:00
tmp_keyseg.start= offset;
offset+= MI_UNIQUE_HASH_LENGTH;
if (mi_keydef_write(file,&tmp_keydef) ||
mi_keyseg_write(file,(&tmp_keyseg)))
goto err;
}
/* Save unique definition */
Bug#11824 - internal /tmp/*.{MYD,MYI} files remain, causing subsequent queries to fail Very complex select statements can create temporary tables that are too big to be represented as a MyISAM table. This was not checked at table creation time, but only at open time. The result was an attempt to delete the "impossible" table. But if the server is built --with-raid, MyISAM tries to open the table before deleting the files. It needs to find out if the table uses the raid support and how many raid chunks there are. This is done with an open "for repair", which will almost always succeed. But in this case we have an "impossible" table. The open failed. Hence the files were not deleted. Also the error message was a bit unspecific. I turned an open error in this situation into the assumption of having no raid support on the table. Thus the normal data file is tried to be deleted. This may however leave existing raid chunks behind. I also added a check in mi_create() to prevent the creation of an "impossible" table. A more decriptive error message is given in this case. No test case. The required select statement is way too large for the test suite. I added a test script to the bug report.
2006-06-27 11:26:41 +02:00
DBUG_PRINT("info", ("write unique definitions"));
Import changeset
2000-07-31 21:29:14 +02:00
for (i=0 ; i < share.state.header.uniques ; i++)
{
Add 0x before pointers (to help with debugging) Add support for VARCHAR with 1 or 2 length bytes Enable VARCHAR packing in MyISAM files (previous patch didn't pack data properly) Give error if we got problems in temporary tables during a SELECT Don't use new table generated by ALTER TABLE if index generation fails Fixed wrong call by range_end() (Could cause an ASSERT in debug mode)
2004-12-18 04:19:21 +01:00
HA_KEYSEG *keyseg_end;
keyseg= uniquedefs[i].seg;
Import changeset
2000-07-31 21:29:14 +02:00
if (mi_uniquedef_write(file, &uniquedefs[i]))
goto err;
Add 0x before pointers (to help with debugging) Add support for VARCHAR with 1 or 2 length bytes Enable VARCHAR packing in MyISAM files (previous patch didn't pack data properly) Give error if we got problems in temporary tables during a SELECT Don't use new table generated by ALTER TABLE if index generation fails Fixed wrong call by range_end() (Could cause an ASSERT in debug mode)
2004-12-18 04:19:21 +01:00
for (keyseg= uniquedefs[i].seg, keyseg_end= keyseg+ uniquedefs[i].keysegs;
keyseg < keyseg_end;
keyseg++)
Import changeset
2000-07-31 21:29:14 +02:00
{
Add 0x before pointers (to help with debugging) Add support for VARCHAR with 1 or 2 length bytes Enable VARCHAR packing in MyISAM files (previous patch didn't pack data properly) Give error if we got problems in temporary tables during a SELECT Don't use new table generated by ALTER TABLE if index generation fails Fixed wrong call by range_end() (Could cause an ASSERT in debug mode)
2004-12-18 04:19:21 +01:00
switch (keyseg->type) {
case HA_KEYTYPE_VARTEXT1:
case HA_KEYTYPE_VARTEXT2:
case HA_KEYTYPE_VARBINARY1:
case HA_KEYTYPE_VARBINARY2:
if (!(keyseg->flag & HA_BLOB_PART))
{
keyseg->flag|= HA_VAR_LENGTH_PART;
keyseg->bit_start= ((keyseg->type == HA_KEYTYPE_VARTEXT1 ||
keyseg->type == HA_KEYTYPE_VARBINARY1) ?
1 : 2);
}
break;
default:
break;
}
if (mi_keyseg_write(file, keyseg))
Import changeset
2000-07-31 21:29:14 +02:00
goto err;
}
}
Bug#11824 - internal /tmp/*.{MYD,MYI} files remain, causing subsequent queries to fail Very complex select statements can create temporary tables that are too big to be represented as a MyISAM table. This was not checked at table creation time, but only at open time. The result was an attempt to delete the "impossible" table. But if the server is built --with-raid, MyISAM tries to open the table before deleting the files. It needs to find out if the table uses the raid support and how many raid chunks there are. This is done with an open "for repair", which will almost always succeed. But in this case we have an "impossible" table. The open failed. Hence the files were not deleted. Also the error message was a bit unspecific. I turned an open error in this situation into the assumption of having no raid support on the table. Thus the normal data file is tried to be deleted. This may however leave existing raid chunks behind. I also added a check in mi_create() to prevent the creation of an "impossible" table. A more decriptive error message is given in this case. No test case. The required select statement is way too large for the test suite. I added a test script to the bug report.
2006-06-27 11:26:41 +02:00
DBUG_PRINT("info", ("write field definitions"));
Import changeset
2000-07-31 21:29:14 +02:00
for (i=0 ; i < share.base.fields ; i++)
if (mi_recinfo_write(file, &recinfo[i]))
goto err;
#ifndef DBUG_OFF
if ((uint) my_tell(file,MYF(0)) != info_length)
{
uint pos= (uint) my_tell(file,MYF(0));
DBUG_PRINT("warning",("info_length: %d != used_length: %d",
info_length, pos));
}
#endif
/* Enlarge files */
Bug#11824 - internal /tmp/*.{MYD,MYI} files remain, causing subsequent queries to fail Very complex select statements can create temporary tables that are too big to be represented as a MyISAM table. This was not checked at table creation time, but only at open time. The result was an attempt to delete the "impossible" table. But if the server is built --with-raid, MyISAM tries to open the table before deleting the files. It needs to find out if the table uses the raid support and how many raid chunks there are. This is done with an open "for repair", which will almost always succeed. But in this case we have an "impossible" table. The open failed. Hence the files were not deleted. Also the error message was a bit unspecific. I turned an open error in this situation into the assumption of having no raid support on the table. Thus the normal data file is tried to be deleted. This may however leave existing raid chunks behind. I also added a check in mi_create() to prevent the creation of an "impossible" table. A more decriptive error message is given in this case. No test case. The required select statement is way too large for the test suite. I added a test script to the bug report.
2006-06-27 11:26:41 +02:00
DBUG_PRINT("info", ("enlarge to keystart: %lu", (ulong) share.base.keystart));
Lots of code fixes to the replication code (especially the binary logging and index log file handling) Fixed bugs in my last changeset that made MySQL hard to compile. Added mutex around some data that could cause table cache corruptions when using OPTIMIZE TABLE / REPAIR TABLE or automatic repair of MyISAM tables. Added mutex around some data in the slave start/stop code that could cause THD linked list corruptions Extended my_chsize() to allow one to specify a filler character. Extend vio_blocking to return the old state (This made some usage of this function much simpler) Added testing for some functions that they caller have got the required mutexes before calling the function. Use setrlimit() to ensure that we can write core file if one specifies --core-file. Added --slave-compressed-protocol Made 2 the minimum length for ft_min_word_len Added variables foreign_key_checks & unique_checks. Less logging from replication code (if not started with --log-warnings) Changed that SHOW INNODB STATUS requre the SUPER privilege More DBUG statements and a lot of new code comments
2002-08-08 02:12:02 +02:00
if (my_chsize(file,(ulong) share.base.keystart,0,MYF(0)))
Import changeset
2000-07-31 21:29:14 +02:00
goto err;
if (! (flags & HA_DONT_TOUCH_DATA))
{
#ifdef USE_RELOC
Lots of code fixes to the replication code (especially the binary logging and index log file handling) Fixed bugs in my last changeset that made MySQL hard to compile. Added mutex around some data that could cause table cache corruptions when using OPTIMIZE TABLE / REPAIR TABLE or automatic repair of MyISAM tables. Added mutex around some data in the slave start/stop code that could cause THD linked list corruptions Extended my_chsize() to allow one to specify a filler character. Extend vio_blocking to return the old state (This made some usage of this function much simpler) Added testing for some functions that they caller have got the required mutexes before calling the function. Use setrlimit() to ensure that we can write core file if one specifies --core-file. Added --slave-compressed-protocol Made 2 the minimum length for ft_min_word_len Added variables foreign_key_checks & unique_checks. Less logging from replication code (if not started with --log-warnings) Changed that SHOW INNODB STATUS requre the SUPER privilege More DBUG statements and a lot of new code comments
2002-08-08 02:12:02 +02:00
if (my_chsize(dfile,share.base.min_pack_length*ci->reloc_rows,0,MYF(0)))
Import changeset
2000-07-31 21:29:14 +02:00
goto err;
#endif
errpos=2;
if (my_close(dfile,MYF(0)))
goto err;
}
errpos=0;
pthread_mutex_unlock(&THR_LOCK_myisam);
if (my_close(file,MYF(0)))
goto err;
New thr_alarm struct for better integration with OS2 Run bootstrap in separate thread Fix bug in FLUSH TABLES table_name
2000-11-28 03:47:47 +01:00
my_free((char*) rec_per_key_part,MYF(0));
Import changeset
2000-07-31 21:29:14 +02:00
DBUG_RETURN(0);
err:
pthread_mutex_unlock(&THR_LOCK_myisam);
save_errno=my_errno;
switch (errpos) {
case 3:
VOID(my_close(dfile,MYF(0)));
/* fall through */
case 2:
/* QQ: T<>nu should add a call to my_raid_delete() here */
Symlinking of tables should not work. Fixed error number handling bug in mysqltest. Fixed that error number from insert delayed is reported correctly. merged new vio code with old violite code.
2001-06-05 02:38:10 +02:00
if (! (flags & HA_DONT_TOUCH_DATA))
WL#1324 table name to file name encoding - Encoding itself, implemented as a charset "filename". Originally planned to use '.' as an escape character, but now changed to '@' for two reasons: "ls" does not return file names starting with '.' considering them as a kind of hidden files; some platforms do not allow several dots in a file name. - replacing many calls of my_snprintf() and strnxmov() to the new build_table_filename(). - Adding MY_APPEND_EXT mysys flag, to append an extention rather that replace it. - Replacing all numeric constants in fn_format flag arguments to their mysys definitions, e.g. MY_UNPACK_FILENAME, - Predictability in several function/methods: when a table name can appear with or withot .frm extension. Some functions/methods were changed so accept names strictly with .frm, other - strictly without .frm extensions. Several DBUG_ASSERTs were added to check whether an extension is passed. Many files: table name to file name encoding mysql_priv.h: Prototypes for new table name encoding tools. ctype-utf8.c: Implementing "filename" charset for table name to file name encoding. row0mysql.c: Fixing table name prefix. mf_format.c: Adding MY_APPEND_EXT processing. Many files: Fixing tests. my_sys.h: Adding new flag to append rather than replace an extension. m_ctype.h: Adding "filename" charset definition.
2005-12-31 06:01:26 +01:00
my_delete_with_symlink(fn_format(filename,name,"",MI_NAME_DEXT,
MY_UNPACK_FILENAME | MY_APPEND_EXT),
Symlinking of tables should not work. Fixed error number handling bug in mysqltest. Fixed that error number from insert delayed is reported correctly. merged new vio code with old violite code.
2001-06-05 02:38:10 +02:00
MYF(0));
Import changeset
2000-07-31 21:29:14 +02:00
/* fall through */
case 1:
VOID(my_close(file,MYF(0)));
if (! (flags & HA_DONT_TOUCH_DATA))
WL#1324 table name to file name encoding - Encoding itself, implemented as a charset "filename". Originally planned to use '.' as an escape character, but now changed to '@' for two reasons: "ls" does not return file names starting with '.' considering them as a kind of hidden files; some platforms do not allow several dots in a file name. - replacing many calls of my_snprintf() and strnxmov() to the new build_table_filename(). - Adding MY_APPEND_EXT mysys flag, to append an extention rather that replace it. - Replacing all numeric constants in fn_format flag arguments to their mysys definitions, e.g. MY_UNPACK_FILENAME, - Predictability in several function/methods: when a table name can appear with or withot .frm extension. Some functions/methods were changed so accept names strictly with .frm, other - strictly without .frm extensions. Several DBUG_ASSERTs were added to check whether an extension is passed. Many files: table name to file name encoding mysql_priv.h: Prototypes for new table name encoding tools. ctype-utf8.c: Implementing "filename" charset for table name to file name encoding. row0mysql.c: Fixing table name prefix. mf_format.c: Adding MY_APPEND_EXT processing. Many files: Fixing tests. my_sys.h: Adding new flag to append rather than replace an extension. m_ctype.h: Adding "filename" charset definition.
2005-12-31 06:01:26 +01:00
my_delete_with_symlink(fn_format(filename,name,"",MI_NAME_IEXT,
MY_UNPACK_FILENAME | MY_APPEND_EXT),
Symlinking of tables should not work. Fixed error number handling bug in mysqltest. Fixed that error number from insert delayed is reported correctly. merged new vio code with old violite code.
2001-06-05 02:38:10 +02:00
MYF(0));
Import changeset
2000-07-31 21:29:14 +02:00
}
New thr_alarm struct for better integration with OS2 Run bootstrap in separate thread Fix bug in FLUSH TABLES table_name
2000-11-28 03:47:47 +01:00
my_free((char*) rec_per_key_part, MYF(0));
Import changeset
2000-07-31 21:29:14 +02:00
DBUG_RETURN(my_errno=save_errno); /* return the fatal errno */
}
uint mi_get_pointer_length(ulonglong file_length, uint def)
{
Don't allow 8bytes for data file pointers for now.
2005-06-30 16:55:47 +02:00
DBUG_ASSERT(def >= 2 && def <= 7);
Import changeset
2000-07-31 21:29:14 +02:00
if (file_length) /* If not default */
{
Don't allow 8bytes for data file pointers for now.
2005-06-30 16:55:47 +02:00
#ifdef NOT_YET_READY_FOR_8_BYTE_POINTERS
Bug#24607 - MyISAM pointer size determined incorrectly The function mi_get_pointer_length() computed too small pointer size for very large tables. Inserted missing 'else' between the branches for very large tables.
2007-01-05 10:26:51 +01:00
if (file_length >= ULL(1) << 56)
Import changeset
2000-07-31 21:29:14 +02:00
def=8;
Bug#24607 - MyISAM pointer size determined incorrectly The function mi_get_pointer_length() computed too small pointer size for very large tables. Inserted missing 'else' between the branches for very large tables.
2007-01-05 10:26:51 +01:00
else
Don't allow 8bytes for data file pointers for now.
2005-06-30 16:55:47 +02:00
#endif
Bug#24607 - MyISAM pointer size determined incorrectly The function mi_get_pointer_length() computed too small pointer size for very large tables. Inserted missing 'else' between the branches for very large tables.
2007-01-05 10:26:51 +01:00
if (file_length >= ULL(1) << 48)
Import changeset
2000-07-31 21:29:14 +02:00
def=7;
Bug#24607 - MyISAM pointer size determined incorrectly The function mi_get_pointer_length() computed too small pointer size for very large tables. Inserted missing 'else' between the branches for very large tables.
2007-01-05 10:26:51 +01:00
else if (file_length >= ULL(1) << 40)
Import changeset
2000-07-31 21:29:14 +02:00
def=6;
Bug#24607 - MyISAM pointer size determined incorrectly The function mi_get_pointer_length() computed too small pointer size for very large tables. Inserted missing 'else' between the branches for very large tables.
2007-01-05 10:26:51 +01:00
else if (file_length >= ULL(1) << 32)
Import changeset
2000-07-31 21:29:14 +02:00
def=5;
Bug#24607 - MyISAM pointer size determined incorrectly The function mi_get_pointer_length() computed too small pointer size for very large tables. Inserted missing 'else' between the branches for very large tables.
2007-01-05 10:26:51 +01:00
else if (file_length >= ULL(1) << 24)
Import changeset
2000-07-31 21:29:14 +02:00
def=4;
Bug#24607 - MyISAM pointer size determined incorrectly The function mi_get_pointer_length() computed too small pointer size for very large tables. Inserted missing 'else' between the branches for very large tables.
2007-01-05 10:26:51 +01:00
else if (file_length >= ULL(1) << 16)
Import changeset
2000-07-31 21:29:14 +02:00
def=3;
else
def=2;
}
return def;
}
Reference in a new issue Copy permalink