-/*
- * Copyright (c) 1996-1998, Index Data.
- * See the file LICENSE for details.
- * Sebastian Hammer, Adam Dickmeiss, Heikki Levanto
- *
- */
+/* $Id: merge.c,v 1.21 2002-08-02 19:26:56 adam Exp $
+ Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002
+ Index Data Aps
+
+This file is part of the Zebra server.
+
+Zebra 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, or (at your option) any later
+version.
+
+Zebra 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 Zebra; see the file LICENSE.zebra. If not, write to the
+Free Software Foundation, 59 Temple Place - Suite 330, Boston, MA
+02111-1307, USA.
+*/
+
+
#include <stdlib.h>
#include <assert.h>
#include <string.h>
#include <stdio.h>
-#include <log.h>
+#include <yaz/log.h>
#include "isamc-p.h"
-#include "isamh-p.h"
struct isc_merge_block {
int offset; /* offset in r_buf */
return cat + firstpos * 8;
}
-static char *hexdump(unsigned char *p, int len, char *buff) {
- static char localbuff[128];
- char bytebuff[8];
- if (!buff) buff=localbuff;
- *buff='\0';
- while (len--) {
- sprintf(bytebuff,"%02x",*p);
- p++;
- strcat(buff,bytebuff);
- if (len) strcat(buff,",");
- }
- return buff;
-}
-
-
-/* isamh - heikki's append-only isam
- * Idea: When allocating a new block, allocate memory for a very large block
- * (maximal blocksize). When done, see if you can shrink it to some
- * smaller size. First-time indexing will go in optimal blocks, and
- * following small additions will go to the end of the last of the
- * maximal ones. Only later, when new blocks need to be allocated, it
- * may make sense to reserve some extra space...
-*/
-
-ISAMC_P isamh_append (ISAMH is, ISAMH_P ipos, ISAMH_I data)
-{
-
- ISAMH_PP pp;
-
- char i_item[128];
- char *i_item_ptr;
- int i_more=1, i_mode;
-
- char codebuffer[128];
- char *codeptr;
- char *bufptr;
- int codelen;
-
- ISAMH_PP firstpp;
- void *r_clientData; /* encode client data */
- int newblock;
- int newcat;
- int maxkeys;
- int maxsize;
- int retval;
-
- pp = firstpp = isamh_pp_open (is, ipos);
- assert (*is->method->code_reset);
-
- if ( 0==ipos)
- { /* new block */
- pp->cat=0; /* start small... */
- pp->pos = isamh_alloc_block(is,pp->cat);
- pp->size= pp->offset = ISAMH_BLOCK_OFFSET_1 ;
- r_clientData = (*is->method->code_start)(ISAMH_ENCODE);
- if (pp->is->method->debug > 2)
- logf(LOG_LOG,"isamh_append: starting with new block %d",pp->pos);
- }
- else
- { /* existing block */
- if (isamh_block(firstpp->lastblock) == firstpp->pos)
- { /* only one block, we have it already */
- pp->offset=ISAMH_BLOCK_OFFSET_1;
- if (pp->is->method->debug > 2)
- logf(LOG_LOG,"isamh_append: starting with one block %d",pp->pos);
- }
- else
- {
- if (pp->is->method->debug > 2)
- logf(LOG_LOG,"isamh_append: starting with multiple blocks %d>%d>%d",
- firstpp->pos,isamh_block(firstpp->next),isamh_block(firstpp->lastblock));
- pp=isamh_pp_open(is,firstpp->lastblock);
- /* dirty, but this can also read a N-block. Just clear extra values*/
- pp->lastblock=0;
- pp->offset=ISAMH_BLOCK_OFFSET_N;
- } /* get last */
- r_clientData = (*is->method->code_start)(ISAMH_ENCODE);
- if (pp->is->method->debug > 3)
- logf(LOG_LOG,"isamh_append: scanning to end of block %d %d->%d",
- pp->pos, pp->offset, pp->size);
- codeptr=codebuffer;
- while (pp->offset<pp->size) {
- codeptr=codebuffer;
- bufptr=pp->buf + pp->offset;
- (*is->method->code_item)(ISAMH_DECODE, r_clientData, &codeptr, &bufptr);
- codelen = bufptr - (pp->buf+pp->offset) ;
- if (pp->is->method->debug > 3)
- logf(LOG_LOG,"isamh_append: dec at %d %d/%d:%s",
- pp->offset, codelen, codeptr-codebuffer,
- hexdump(codebuffer,codeptr-codebuffer,0) );
- pp->offset += codelen;
- }
- } /* existing block */
-
-
- i_item_ptr = i_item;
- i_more = (*data->read_item)(data->clientData,&i_item_ptr,&i_mode);
- if (pp->is->method->debug > 3)
- logf(LOG_LOG,"isamh_append 1: m=%d l=%d %s",
- i_mode, i_item_ptr-i_item, hexdump(i_item,i_item_ptr-i_item,0));
-
- maxsize = is->method->filecat[pp->cat].bsize;
-
- while(i_more) {
- if (i_mode)
- { /* insert key, ignore all delete keys time being... */
- codeptr = codebuffer;
- i_item_ptr=i_item;
- (*is->method->code_item)(ISAMH_ENCODE, r_clientData, &codeptr, &i_item_ptr);
- codelen = codeptr-codebuffer;
-
- assert( (codelen < 128) && (codelen>0));
-
- if (pp->is->method->debug > 3)
- logf(LOG_LOG,"isamh_append: coded into %d:%s",
- codelen,hexdump(codebuffer,codelen,0));
-
- if ( pp->offset + codelen > maxsize )
- { /* oops, block full, do something */
- newcat = pp->cat;
- maxkeys = is->method->filecat[pp->cat].mblocks; /* max keys */
- if (pp->is->method->debug > 2)
- logf(LOG_LOG,"isamh_append: need new block: %d > %d (k:%d/%d)",
- pp->offset + codelen, maxsize, firstpp->numKeys,maxkeys );
- if ( (maxkeys>0) && (firstpp->numKeys > maxkeys) )
- { /* time to increase block size */
- newcat++;
- maxsize = is->method->filecat[newcat].bsize;
- pp->buf=xrealloc(pp->buf,maxsize);
- if (pp->is->method->debug > 2)
- logf(LOG_LOG,"isamh_append: increased to cat %d ",newcat);
- }
-
- newblock = isamh_alloc_block(is,newcat);
- pp->next = isamh_addr(newblock,newcat);
- if (firstpp!=pp)
- { /* not first block, write to disk already now */
- isamh_buildlaterblock(pp);
- isamh_write_block(is,pp->cat,pp->pos,pp->buf);
- }
- else
- { /* we had only one block, allocate a second buffer */
- pp = isamh_pp_open(is,0);
- }
- pp->cat = newcat;
- pp->pos = newblock;
-
- pp->size=pp->offset=ISAMH_BLOCK_OFFSET_N ;
- pp->next=0;
- pp->lastblock=0;
- if (pp->is->method->debug > 2)
- logf(LOG_LOG,"isamh_append: got a new block %d:%d",pp->cat,pp->pos);
-
- /* reset the encoding, and code again */
- (*is->method->code_reset)(r_clientData);
- codeptr = codebuffer;
- i_item_ptr=i_item;
- (*is->method->code_item)(ISAMH_ENCODE, r_clientData, &codeptr, &i_item_ptr);
- codelen = codeptr-codebuffer;
- if (pp->is->method->debug > 3)
- logf(LOG_LOG,"isamh_append: coded again %d:%s",
- codelen,hexdump(codebuffer,codelen,0));
-
- } /* block full */
-
- /* ok, now we can write it */
- memcpy(&(pp->buf[pp->offset]), codebuffer, codelen);
- pp->offset += codelen;
- pp->size += codelen;
- firstpp->numKeys++;
- } /* not a delete */
-
- /* try to read the next element */
- i_item_ptr = i_item;
- i_more = (*data->read_item)(data->clientData,&i_item_ptr,&i_mode);
- if (pp->is->method->debug > 3)
- logf(LOG_LOG,"isamh_append 2: m=%d l=%d %s",
- i_mode, i_item_ptr-i_item, hexdump(i_item,i_item_ptr-i_item,0));
-
- } /* while */
-
- /* Write the last (partial) block, if needed. */
- if (pp!=firstpp)
- {
- pp->next=0; /* just to be sure */
- isamh_buildlaterblock(pp);
- isamh_write_block(is,pp->cat,pp->pos,pp->buf);
- }
-
- /* update first block and write it */
- firstpp->lastblock = isamh_addr(pp->pos,pp->cat);
- isamh_buildfirstblock(firstpp);
- isamh_write_block(is,firstpp->cat,firstpp->pos,firstpp->buf);
-
- /* release the second block, if we allocated one */
- if ( firstpp != pp )
- isamh_pp_close(pp);
-
- /* get return value (before it disappears at close! */
- retval = isamh_addr(firstpp->pos,firstpp->cat);
-
- isamh_pp_close(firstpp);
-
- return retval;
-
-} /* isamh_append */
-
-
/*
* $Log: merge.c,v $
- * Revision 1.15 1999-07-07 09:36:04 heikki
+ * Revision 1.21 2002-08-02 19:26:56 adam
+ * Towards GPL
+ *
+ * Revision 1.20 1999/11/30 13:48:04 adam
+ * Improved installation. Updated for inclusion of YAZ header files.
+ *
+ * Revision 1.19 1999/07/14 12:12:07 heikki
+ * Large-block isam-h (may not work too well... Abandoning for isam-d)
+ *
+ * Revision 1.17 1999/07/13 14:22:17 heikki
+ * Better allocation strategy in isamh_merge
+ *
+ * Revision 1.16 1999/07/08 14:23:27 heikki
+ * Fixed a bug in isamh_pp_read and cleaned up a bit
+ *
+ * Revision 1.15 1999/07/07 09:36:04 heikki
* Fixed an assertion in isamh
*
* Revision 1.13 1999/07/06 09:37:05 heikki