Added Unique class to be used for duplicate removal in multi-table delete.

mysys/ptr_cmp.c:
  Removed old comments
pstack/bucomm.c:
  Change to use mkstemp.
sql/Makefile.am:
  Added Unique class
sql/filesort.cc:
  Changes to let the Unique class use the old filesort code.
sql/sql_class.h:
  Added Unique class
sql/sql_handler.cc:
  Removed warning.
sql/sql_select.cc:
  Cleaned up typo.

mysys/ptr_cmp.c

@@ -19,9 +19,6 @@
   get_ptr_compare(len) returns a pointer to a optimal byte-compare function
   for a array of stringpointer where all strings have size len.
   The bytes are compare as unsigned chars.
-  Because the size is saved in a static variable.
-  When using threads the program must have called my_init and the thread
-  my_init_thread()
   */
 
 #include "mysys_priv.h"

pstack/bucomm.c

@@ -205,7 +205,7 @@ make_tempname (filename)
       strcpy (tmpname, filename);
       strcat (tmpname, "/");
       strcat (tmpname, template);
-      mktemp (tmpname);
+      mkstemp (tmpname);
       *slash = c;
     }
   else

sql/Makefile.am

@@ -56,7 +56,7 @@ noinst_HEADERS =	item.h item_func.h item_sum.h item_cmpfunc.h \
 			sql_select.h structs.h table.h sql_udf.h hash_filo.h\
 			lex.h lex_symbol.h sql_acl.h sql_crypt.h md5.h \
                         log_event.h mini_client.h sql_repl.h slave.h \
-			stacktrace.h
+			stacktrace.h sql_sort.h
 mysqld_SOURCES =	sql_lex.cc sql_handler.cc \
 			item.cc item_sum.cc item_buff.cc item_func.cc \
 			item_cmpfunc.cc item_strfunc.cc item_timefunc.cc \
@@ -67,7 +67,7 @@ mysqld_SOURCES =	sql_lex.cc sql_handler.cc \
 			mysqld.cc password.c hash_filo.cc hostname.cc \
 			convert.cc sql_parse.cc sql_yacc.yy \
 			sql_base.cc table.cc sql_select.cc sql_insert.cc \
-			sql_update.cc sql_delete.cc \
+			sql_update.cc sql_delete.cc uniques.cc \
 			procedure.cc item_uniq.cc sql_test.cc \
 			log.cc log_event.cc init.cc derror.cc sql_acl.cc \
 			unireg.cc \

sql/filesort.cc

@@ -22,13 +22,11 @@
 #include <stddef.h>			/* for macro offsetof */
 #endif
 #include <m_ctype.h>
+#include "sql_sort.h"
+
 #ifndef THREAD
 #define SKIPP_DBUG_IN_FILESORT
 #endif
-	/* static variabels */
-
-#define MERGEBUFF  7
-#define MERGEBUFF2 15
 
 	/* How to write record_ref. */
 
@@ -36,28 +34,6 @@
 if (my_b_write((file),(byte*) (from),param->ref_length)) \
   DBUG_RETURN(1);
 
-typedef struct st_buffpek {		/* Struktur om sorteringsbuffrarna */
-  my_off_t file_pos;			/* Where we are in the sort file */
-  uchar *base,*key;			/* key pointers */
-  ha_rows count;			/* Number of rows in table */
-  ulong mem_count;			/* numbers of keys in memory */
-  ulong max_keys;			/* Max keys in buffert */
-} BUFFPEK;
-
-
-typedef struct st_sort_param {
-  uint sort_length;			/* Length of sortarg */
-  uint keys;				/* Max antal nycklar / buffert */
-  uint ref_length;			/* Length of record ref. */
-  ha_rows max_rows,examined_rows;
-  TABLE *sort_form;			/* For quicker make_sortkey */
-  SORT_FIELD *local_sortorder;
-  SORT_FIELD *end;
-#ifdef USE_STRCOLL
-  char* tmp_buffer;
-#endif
-} SORTPARAM;
-
 	/* functions defined in this file */
 
 static char **make_char_array(register uint fields, uint length, myf my_flag);
@@ -70,20 +46,11 @@ static int write_keys(SORTPARAM *param,uchar * *sort_keys,
 		      IO_CACHE *tempfile);
 static void make_sortkey(SORTPARAM *param,uchar *to,
 			 byte *ref_pos);
-static bool save_index(SORTPARAM *param,uchar **sort_keys, uint count);
-static int merge_many_buff(SORTPARAM *param,uchar * *sort_keys,
-			   BUFFPEK *buffpek,
-			   uint *maxbuffer, IO_CACHE *t_file);
-static uint read_to_buffer(IO_CACHE *fromfile,BUFFPEK *buffpek,
-			   uint sort_length);
-static int merge_buffers(SORTPARAM *param,IO_CACHE *from_file,
-			 IO_CACHE *to_file,uchar * *sort_keys,
-			 BUFFPEK *lastbuff,BUFFPEK *Fb,
-			 BUFFPEK *Tb,int flag);
-static int merge_index(SORTPARAM *param,uchar * *sort_keys,
+static int merge_index(SORTPARAM *param,uchar *sort_buffer,
 		       BUFFPEK *buffpek,
 		       uint maxbuffer,IO_CACHE *tempfile,
 		       IO_CACHE *outfile);
+static bool save_index(SORTPARAM *param,uchar **sort_keys, uint count);
 static uint sortlength(SORT_FIELD *sortorder,uint length);
 
 	/* Makes a indexfil of recordnumbers of a sorted database */
@@ -225,12 +192,14 @@ ha_rows filesort(TABLE **table, SORT_FIELD *sortorder, uint s_length,
 
     param.keys=((param.keys*(param.sort_length+sizeof(char*))) /
 		param.sort_length-1);
-    if (merge_many_buff(&param,sort_keys,buffpek,&maxbuffer,&tempfile))
+    if (merge_many_buff(&param,(uchar*) sort_keys,buffpek,&maxbuffer,
+			&tempfile))
       goto err;
     if (flush_io_cache(&tempfile) ||
 	reinit_io_cache(&tempfile,READ_CACHE,0L,0,0))
       goto err;
-    if (merge_index(&param,sort_keys,buffpek,maxbuffer,&tempfile,outfile))
+    if (merge_index(&param,(uchar*) sort_keys,buffpek,maxbuffer,&tempfile,
+		    outfile))
       goto err;
   }
   if (records > param.max_rows)
@@ -629,7 +598,7 @@ static bool save_index(SORTPARAM *param, uchar **sort_keys, uint count)
 
 	/* Merge buffers to make < MERGEBUFF2 buffers */
 
-static int merge_many_buff(SORTPARAM *param, uchar **sort_keys,
+int merge_many_buff(SORTPARAM *param, uchar *sort_buffer,
 		    BUFFPEK *buffpek, uint *maxbuffer, IO_CACHE *t_file)
 {
   register int i;
@@ -652,11 +621,11 @@ static int merge_many_buff(SORTPARAM *param, uchar **sort_keys,
     lastbuff=buffpek;
     for (i=0 ; i <= (int) *maxbuffer-MERGEBUFF*3/2 ; i+=MERGEBUFF)
     {
-      if (merge_buffers(param,from_file,to_file,sort_keys,lastbuff++,
+      if (merge_buffers(param,from_file,to_file,sort_buffer,lastbuff++,
 			buffpek+i,buffpek+i+MERGEBUFF-1,0))
 	break;					/* purecov: inspected */
     }
-    if (merge_buffers(param,from_file,to_file,sort_keys,lastbuff++,
+    if (merge_buffers(param,from_file,to_file,sort_buffer,lastbuff++,
 		      buffpek+i,buffpek+ *maxbuffer,0))
       break;					/* purecov: inspected */
     if (flush_io_cache(to_file))
@@ -675,7 +644,7 @@ static int merge_many_buff(SORTPARAM *param, uchar **sort_keys,
 	/* Read data to buffer */
 	/* This returns (uint) -1 if something goes wrong */
 
-static uint read_to_buffer(IO_CACHE *fromfile, BUFFPEK *buffpek,
+uint read_to_buffer(IO_CACHE *fromfile, BUFFPEK *buffpek,
 		    uint sort_length)
 {
   register uint count;
@@ -697,8 +666,8 @@ static uint read_to_buffer(IO_CACHE *fromfile, BUFFPEK *buffpek,
 
 	/* Merge buffers to one buffer */
 
-static int merge_buffers(SORTPARAM *param, IO_CACHE *from_file,
-			 IO_CACHE *to_file, uchar **sort_keys,
+int merge_buffers(SORTPARAM *param, IO_CACHE *from_file,
+		  IO_CACHE *to_file, uchar *sort_buffer,
 		  BUFFPEK *lastbuff, BUFFPEK *Fb, BUFFPEK *Tb,
 		  int flag)
 {
@@ -711,21 +680,21 @@ static int merge_buffers(SORTPARAM *param, IO_CACHE *from_file,
   BUFFPEK *buffpek,**refpek;
   QUEUE queue;
   volatile bool *killed= &current_thd->killed;
+  qsort2_cmp    cmp;
   DBUG_ENTER("merge_buffers");
 
   statistic_increment(filesort_merge_passes, &LOCK_status);
 
   count=error=0;
-  offset=param->sort_length-param->ref_length;
+  offset=(sort_length=param->sort_length)-param->ref_length;
   maxcount=(ulong) (param->keys/((uint) (Tb-Fb) +1));
   to_start_filepos=my_b_tell(to_file);
-  strpos=(uchar*) sort_keys;
-  sort_length=param->sort_length;
+  strpos=(uchar*) sort_buffer;
   max_rows=param->max_rows;
 
   if (init_queue(&queue,(uint) (Tb-Fb)+1,offsetof(BUFFPEK,key),0,
 		 (int (*) (void *, byte *,byte*))
-		 get_ptr_compare(sort_length),(void*) &sort_length))
+		 (cmp=get_ptr_compare(sort_length)),(void*) &sort_length))
     DBUG_RETURN(1);				/* purecov: inspected */
   for (buffpek= Fb ; buffpek <= Tb ; buffpek++)
   {
@@ -739,6 +708,26 @@ static int merge_buffers(SORTPARAM *param, IO_CACHE *from_file,
     queue_insert(&queue,(byte*) buffpek);
   }
 
+  if (param->unique_buff)
+  {
+    /* 
+       Called by Unique::get()
+       Copy the first argument to param->unique_buff for unique removal.
+       Store it also in 'to_file'.
+
+       This is safe as we know that there is always more than one element
+       in each block to merge (This is guaranteed by the Unique:: algorithm
+    */
+    buffpek=(BUFFPEK*) queue_top(&queue);
+    memcpy(param->unique_buff, buffpek->key, sort_length);
+    if (my_b_write(to_file,(byte*) buffpek->key, sort_length))
+    {
+      error=1; goto err;			/* purecov: inspected */
+    }
+  }
+  else
+    cmp=0;					// Not unique
+
   while (queue.elements > 1)
   {
     if (*killed)
@@ -748,6 +737,12 @@ static int merge_buffers(SORTPARAM *param, IO_CACHE *from_file,
     for (;;)
     {
       buffpek=(BUFFPEK*) queue_top(&queue);
+      if (cmp)					// Remove duplicates
+      {
+	if (!cmp(&sort_length, param->unique_buff, (uchar*) buffpek->key))
+	  goto skip_duplicate;
+	memcpy(param->unique_buff, (uchar*) buffpek->key,sort_length);
+      }
       if (flag == 0)
       {
 	if (my_b_write(to_file,(byte*) buffpek->key, sort_length))
@@ -764,6 +759,8 @@ static int merge_buffers(SORTPARAM *param, IO_CACHE *from_file,
 	error=0;				/* purecov: inspected */
 	goto end;				/* purecov: inspected */
       }
+
+    skip_duplicate:
       buffpek->key+=sort_length;
       if (! --buffpek->mem_count)
       {
@@ -802,7 +799,7 @@ static int merge_buffers(SORTPARAM *param, IO_CACHE *from_file,
     }
   }
   buffpek=(BUFFPEK*) queue_top(&queue);
-  buffpek->base=(uchar *) sort_keys;
+  buffpek->base= sort_buffer;
   buffpek->max_keys=param->keys;
   do
   {
@@ -845,12 +842,12 @@ err:
 
 	/* Do a merge to output-file (save only positions) */
 
-static int merge_index(SORTPARAM *param, uchar **sort_keys,
+static int merge_index(SORTPARAM *param, uchar *sort_buffer,
 		       BUFFPEK *buffpek, uint maxbuffer,
 		       IO_CACHE *tempfile, IO_CACHE *outfile)
 {
   DBUG_ENTER("merge_index");
-  if (merge_buffers(param,tempfile,outfile,sort_keys,buffpek,buffpek,
+  if (merge_buffers(param,tempfile,outfile,sort_buffer,buffpek,buffpek,
 		    buffpek+maxbuffer,1))
     DBUG_RETURN(1);				/* purecov: inspected */
   DBUG_RETURN(0);

sql/sql_class.h

@@ -567,3 +567,32 @@ class user_var_entry
   Item_result type;
 };
 
+/* Class for unique (removing of duplicates) */
+
+class Unique :public Sql_alloc
+{
+  DYNAMIC_ARRAY file_ptrs;
+  ulong max_elements, max_in_memory_size;
+  IO_CACHE file;
+  TREE tree;
+  char *record_pointers;
+  bool flush();
+
+public:
+  ulong elements;
+  Unique(qsort_cmp2 comp_func, uint size, ulong max_in_memory_size_arg);
+  ~Unique();
+  inline bool Unique::unique_add(gptr ptr)
+  {
+    if (tree.elements_in_tree > max_elements && flush())
+      return 1;
+    return tree_insert(&tree,ptr,0);
+  }
+
+  bool get(TABLE *table);
+
+  friend int unique_write_to_file(gptr key, Unique *unique,
+				  element_count count);
+  friend int unique_write_to_ptrs(gptr key, Unique *unique,
+				  element_count count);
+};

sql/sql_handler.cc

@@ -43,7 +43,8 @@
   thd->open_tables=thd->handler_tables; \
   thd->handler_tables=tmp; }
 
-static TABLE **find_table_ptr_by_name(THD *thd, char *db, char *table_name);
+static TABLE **find_table_ptr_by_name(THD *thd, const char *db, 
+				      const char *table_name);
 
 int mysql_ha_open(THD *thd, TABLE_LIST *tables)
 {
@@ -231,14 +232,15 @@ err:
 /* Note: this function differs from find_locked_table() because we're looking
    here for alias, not real table name 
  */
-static TABLE **find_table_ptr_by_name(THD *thd, char *db, char *table_name)
+static TABLE **find_table_ptr_by_name(THD *thd, const char *db,
+				      const char *table_name)
 {
   int dblen;
   TABLE **ptr;
   
-  if (!db || ! *db) db=thd->db;
-  if (!db || ! *db) db="";
-  dblen=strlen(db);  
+  if (!db || ! *db)
+    db= thd->db ? thd->db : "";
+  dblen=strlen(db)+1;  
   ptr=&(thd->handler_tables);
   
   for (TABLE *table=*ptr; table ; table=*ptr)

sql/sql_select.cc

@@ -3928,8 +3928,8 @@ bool create_myisam_from_heap(TABLE *table, TMP_TABLE_PARAM *param, int error,
   table->file->rnd_init();
   if (table->no_rows)
   {
-    new_table->file->extra(HA_EXTRA_NO_ROWS);
-    new_table->no_rows=1;
+    new_table.file->extra(HA_EXTRA_NO_ROWS);
+    new_table.no_rows=1;
   }
 
   /* copy all old rows */

sql/sql_sort.h

+/* Copyright (C) 2000 MySQL AB
+
+   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
+   the Free Software Foundation; either version 2 of the License, or
+   (at your option) any later version.
+
+   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.
+
+   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 */
+
+/* Defines used by filesort and uniques */
+
+#define MERGEBUFF		7
+#define MERGEBUFF2		15
+
+typedef struct st_buffpek {		/* Struktur om sorteringsbuffrarna */
+  my_off_t file_pos;			/* Where we are in the sort file */
+  uchar *base,*key;			/* key pointers */
+  ha_rows count;			/* Number of rows in table */
+  ulong mem_count;			/* numbers of keys in memory */
+  ulong max_keys;			/* Max keys in buffert */
+} BUFFPEK;
+
+
+typedef struct st_sort_param {
+  uint sort_length;			/* Length of sort columns */
+  uint keys;				/* Max keys / buffert */
+  uint ref_length;			/* Length of record ref. */
+  ha_rows max_rows,examined_rows;
+  TABLE *sort_form;			/* For quicker make_sortkey */
+  SORT_FIELD *local_sortorder;
+  SORT_FIELD *end;
+  uchar *unique_buff;
+#ifdef USE_STRCOLL
+  char* tmp_buffer;
+#endif
+} SORTPARAM;
+
+
+int merge_many_buff(SORTPARAM *param, uchar *sort_buffer,
+		    BUFFPEK *buffpek,
+		    uint *maxbuffer, IO_CACHE *t_file);
+uint read_to_buffer(IO_CACHE *fromfile,BUFFPEK *buffpek,
+		    uint sort_length);
+int merge_buffers(SORTPARAM *param,IO_CACHE *from_file,
+		  IO_CACHE *to_file, uchar *sort_buffer,
+		  BUFFPEK *lastbuff,BUFFPEK *Fb,
+		  BUFFPEK *Tb,int flag);

sql/uniques.cc

+/* Copyright (C) 2001 MySQL AB
+
+   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
+   the Free Software Foundation; either version 2 of the License, or
+   (at your option) any later version.
+
+   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.
+
+   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 */
+
+/*
+  Function to handle quick removal of duplicates
+  This code is used when doing multi-table deletes to find the rows in
+  reference tables that needs to be deleted.
+
+  The basic idea is as follows:
+
+  Store first all strings in a binary tree, ignoring duplicates.
+  When the three uses more memory than 'max_heap_table_size',
+  write the tree (in sorted order) out to disk and start with a new tree.
+  When all data has been generated, merge the trees (removing any found
+  duplicates).
+
+  The unique entries will be returned in sort order, to ensure that we do the
+  deletes in disk order.
+*/
+
+#include "mysql_priv.h"
+#include "sql_sort.h"
+
+Unique::Unique(qsort_cmp2 comp_func, uint size, ulong max_in_memory_size_arg)
+  :max_in_memory_size(max_in_memory_size_arg),elements(0)
+{
+  my_b_clear(&file);
+  init_tree(&tree, max_in_memory_size / 16, size, comp_func, 0, 0);
+  /* If the following fail's the next add will also fail */
+  init_dynamic_array(&file_ptrs, sizeof(BUFFPEK), 16, 16);
+  max_elements= max_in_memory_size / ALIGN_SIZE(sizeof(TREE_ELEMENT)+size);
+}
+
+
+Unique::~Unique()
+{
+  close_cached_file(&file);
+  delete_tree(&tree);
+  delete_dynamic(&file_ptrs);
+}
+
+
+    /* Write tree to disk; clear tree */    
+bool Unique::flush()
+{
+  BUFFPEK file_ptr;
+  elements+= tree.elements_in_tree;
+  file_ptr.count=tree.elements_in_tree;
+  file_ptr.file_pos=my_b_tell(&file);
+  if (tree_walk(&tree, (tree_walk_action) unique_write_to_file,
+		(void*) this, left_root_right) ||
+      insert_dynamic(&file_ptrs, (gptr) &file_ptr))
+    return 1;
+  delete_tree(&tree);
+  return 0;
+}
+
+
+int unique_write_to_file(gptr key, Unique *unique,  element_count count)
+{
+  return my_b_write(&unique->file, key, unique->tree.size_of_element) ? 1 : 0;
+}
+
+int unique_write_to_ptrs(gptr key, Unique *unique,  element_count count)
+{
+  memcpy(unique->record_pointers, key, unique->tree.size_of_element);
+  unique->record_pointers+=unique->tree.size_of_element;
+  return 0;
+}
+
+
+/*
+  Modify the TABLE element so that when one calls init_records()
+  the rows will be read in priority order.
+*/
+
+bool Unique::get(TABLE *table)
+{
+  SORTPARAM sort_param;
+  table->found_records=elements+tree.elements_in_tree;
+
+  if (!my_b_inited(&file))
+  {
+    /* Whole tree is in memory;  Don't use disk if you don't need to */
+    if ((record_pointers=table->record_pointers= (byte*)
+	 my_malloc(tree.size_of_element * tree.elements_in_tree, MYF(0))))
+    {
+      (void) tree_walk(&tree, (tree_walk_action) unique_write_to_ptrs,
+		       this, left_root_right);
+      return 0;
+    }
+  }
+  /* Not enough memory; Save the result to file */
+  if (flush())
+    return 1;
+
+  IO_CACHE *outfile=table->io_cache, tempfile;
+  BUFFPEK *file_ptr= (BUFFPEK*) file_ptrs.buffer;
+  uint maxbuffer= file_ptrs.elements;
+  uchar *sort_buffer;
+  my_off_t save_pos;
+  bool error=1;
+
+  my_b_clear(&tempfile);
+
+      /* Open cached file if it isn't open */
+  if (! my_b_inited(outfile) &&
+      open_cached_file(outfile,mysql_tmpdir,TEMP_PREFIX,READ_RECORD_BUFFER,
+		       MYF(MY_WME)))
+    return 1;
+  reinit_io_cache(outfile,WRITE_CACHE,0L,0,0);
+
+  sort_param.keys=elements;
+  sort_param.sort_form=table;
+  sort_param.sort_length=sort_param.ref_length=tree.size_of_element;
+  sort_param.keys= max_in_memory_size / sort_param.sort_length;
+  
+  if (!(sort_buffer=(uchar*) my_malloc((sort_param.keys+1) *
+				      sort_param.sort_length,
+				      MYF(0))))
+    return 1;
+  sort_param.unique_buff= sort_buffer+(sort_param.keys*
+				       sort_param.sort_length);
+
+  /* Merge the buffers to one file, removing duplicates */
+  if (merge_many_buff(&sort_param,sort_buffer,file_ptr,&maxbuffer,&tempfile))
+    goto err;
+  if (flush_io_cache(&tempfile) ||
+      reinit_io_cache(&tempfile,READ_CACHE,0L,0,0))
+    goto err;
+  if (merge_buffers(&sort_param, &tempfile, outfile, sort_buffer, file_ptr,
+		    file_ptr, file_ptr+maxbuffer,0))
+    goto err;                                                                 
+  error=0;
+err:
+  x_free((gptr) sort_buffer);
+  close_cached_file(&tempfile);
+  if (flush_io_cache(outfile))
+    error=1;
+
+  /* Setup io_cache for reading */
+  save_pos=outfile->pos_in_file;
+  if (reinit_io_cache(outfile,READ_CACHE,0L,0,0))
+    error=1;
+  outfile->end_of_file=save_pos;
+  return error;
+}