Update of /cvsroot/monetdb/pathfinder/modules/pftijah
In directory sc8-pr-cvs16.sourceforge.net:/tmp/cvs-serv23686
Modified Files:
pftijah.mx
Log Message:
the introduced new index BAT was not updated correctly when using
incremental indexing.
- the added mergeindex2 operator now merges not only the TID index, but
also the new SIZE index.
Index: pftijah.mx
===================================================================
RCS file: /cvsroot/monetdb/pathfinder/modules/pftijah/pftijah.mx,v
retrieving revision 1.161
retrieving revision 1.162
diff -u -d -r1.161 -r1.162
--- pftijah.mx 14 Jan 2008 14:30:36 -0000 1.161
+++ pftijah.mx 15 Jan 2008 10:55:57 -0000 1.162
@@ -124,6 +124,19 @@
the operation merges a new sorted tid_pre with an existing offset index.
"
+.COMMAND mergeindex2( BAT[oid,oid] tid_pre, BAT[oid,int] tid_size,
BAT[void,oid] index, BAT[void, oid] pre, BAT[void, int] size, int indsize)
+ : BAT[void,bat] = CMDmergeindex2;
+"PARAMETERS:
+BAT[oid,oid] - new tid_pre to merge with the old index
+BAT[oid,int] - sizes of new elements to merge with the old index (synched with
tid_pre)
+BAT[void,oid] - old index bat with value-offset
+BAT[void,oid] - old posting lists (pre order lists)
+BAT[void,int] - corresponding size lists (pre order lists)
+int - size of new (dense) offset index.
+DESCRIPTION:
+the operation merges a new sorted tid_pre with an existing offset index.
+"
+
.COMMAND indexfetchjoin( BAT[any,oid] tid, BAT[void,oid] index, BAT[void, oid]
pre)
: BAT[void,oid] = CMDindexfetchjoin;
"PARAMETERS:
@@ -2588,17 +2601,18 @@
tmp := tids.semijoin(bat("tj_" + ftiName + "_pfpre"));
tmp := tmp.reverse().ssort();
-
- i := mergeindex(tmp, collBat.find("_TagIndex"),
- collBat.find("_Tags"),
- collBat.find("_globalTags").count() + 1);
- var tagsize := i.fetch(1).leftfetchjoin(sizes);
+ var tmpsize := tmp.leftfetchjoin(sizes);
+ i := mergeindex2(tmp, tmpsize,
+ collBat.find("_TagIndex"),
+ collBat.find("_Tags"),
+ collBat.find("_TagSize"),
+ collBat.find("_globalTags").count() + 1);
collBat.replace("_TagIndex", i.fetch(0));
collBat.replace("_Tags", i.fetch(1));
- collBat.replace("_TagSize", tagsize);
+ collBat.replace("_TagSize", i.fetch(2));
i := nil;
tmp := nil;
- tagsize := nil;
+ tmpsize := nil;
replaceBats.insert("_TagIndex", "tj_" + ftiName + "_TagIndex");
replaceBats.insert("_Tags", "tj_" + ftiName + "_Tags");
replaceBats.insert("_TagSize", "tj_" + ftiName + "_TagSize");
@@ -3363,6 +3377,169 @@
return GDK_SUCCEED;
}
+int CMDmergeindex2 ( BAT** result, BAT* tidpre, BAT* tidsize, BAT* oldindex,
BAT* oldpre, BAT* oldsize, int* indsize )
+{
+ char *name = "TJmergeindex2";
+ BAT *res = NULL;
+ BAT *newindex = NULL;
+ BAT *newpre = NULL;
+ BAT *newsize = NULL;
+ int i,j, ressize = 0;
+ BUN lst_tidpre, lst_oldindex, lst_copy, lst_res, cur_tidpre,
cur_tidsize, cur_oldindex, cur_oldpre, cur_oldsize;
+ BATiter tidprei, tidsizei, oldindexi, oldprei, oldsizei;
+ oid *s_newindex, *lst_newindex, *s_newpre, *lst_newpre;
+ int *s_newsize, *lst_newsize;
+ oid tid;
+
+ /* --------------------------- checks
---------------------------------- */
+
+ BATcheck(tidpre, name);
+ BATcheck(tidsize, name);
+ BATcheck(oldindex, name);
+ BATcheck(oldpre, name);
+ BATcheck(oldsize, name);
+
+ if (!(BAThordered(tidpre) & 1))
+ {
+ GDKerror("%s: term-bat must be ordered on tail.\n", name);
+ return GDK_FAIL;
+ }
+
+ /* ---------------------------- inits
---------------------------------- */
+
+ ressize = 3;
+ res = BATnew(TYPE_void, TYPE_bat, ressize);
+ if (res == NULL)
+ {
+ GDKerror("%s: could not allocate a result BAT[void,oid] of size
%d.\n", name, ressize);
+ return(GDK_FAIL);
+ }
+
+ ressize = BATcount(tidpre) + BATcount(oldpre);
+ newpre = BATnew(TYPE_void, TYPE_oid, ressize);
+ if (res == NULL)
+ {
+ GDKerror("%s: could not allocate a result BAT[void,oid] of size
%d.\n", name, ressize);
+ return(GDK_FAIL);
+ }
+
+ newsize = BATnew(TYPE_void, TYPE_int, ressize);
+ if (res == NULL)
+ {
+ GDKerror("%s: could not allocate a result BAT[void,int] of size
%d.\n", name, ressize);
+ return(GDK_FAIL);
+ }
+
+ ressize = *indsize;
+ newindex = BATnew(TYPE_void, TYPE_oid, ressize);
+ if (res == NULL)
+ {
+ GDKerror("%s: could not allocate a result BAT[void,oid] of size
%d.\n", name, ressize);
+ return(GDK_FAIL);
+ }
+
+ lst_tidpre = BUNlast(tidpre);
+ lst_oldindex = BUNlast(oldindex) - 1; /* last index is not a real term
*/
+ s_newindex = lst_newindex = (oid*)Tloc(newindex, BUNlast(newindex));
+ s_newpre = lst_newpre = (oid*)Tloc(newpre, BUNlast(newpre));
+ s_newsize = lst_newsize = (int*)Tloc(newsize, BUNlast(newsize));
+ lst_res = BUNlast(res);
+
+ cur_tidpre = BUNfirst(tidpre);
+ cur_tidsize = BUNfirst(tidsize);
+ cur_oldindex = BUNfirst(oldindex);
+ cur_oldpre = BUNfirst(oldpre);
+ cur_oldsize = BUNfirst(oldsize);
+
+ tidprei = bat_iterator(tidpre);
+ tidsizei = bat_iterator(tidsize);
+ oldindexi = bat_iterator(oldindex);
+ oldprei = bat_iterator(oldpre);
+ oldsizei = bat_iterator(oldsize);
+
+ /* ----------------------------- main
---------------------------------- */
+
+ j = *indsize - 1;
+ for(i = 0; i < j; i++)
+ {
+ tid = (oid) i;
+ *lst_newindex++ = lst_newpre - s_newpre;
+
+ /* copy old nodes to new index */
+ if (cur_oldindex < lst_oldindex && tid ==
*(oid*)BUNhead(oldindexi, cur_oldindex))
+ {
+ lst_copy = *(int*) BUNtail(oldindexi, cur_oldindex + 1);
+ while (cur_oldpre < lst_copy)
+ {
+ *lst_newpre++ = *(oid*)BUNtail(oldprei,
cur_oldpre);
+ *lst_newsize++ = *(int*)BUNtail(oldsizei,
cur_oldsize);
+ cur_oldpre++;
+ cur_oldsize++;
+ }
+ cur_oldindex++;
+ }
+ /* merge-in new nodes */
+ while(cur_tidpre < lst_tidpre && tid ==
*(oid*)BUNhead(tidprei, cur_tidpre))
+ {
+ *lst_newpre++ = *(oid*)BUNtail(tidprei, cur_tidpre);
+ *lst_newsize++ = *(int*)BUNtail(tidsizei, cur_tidsize);
+ cur_tidpre++;
+ cur_tidsize++;
+ }
+ }
+
+ /* write limit of index as last item to index bat */
+ *lst_newindex = lst_newpre - s_newpre;
+ lst_newindex++;
+
+ /* ---------------------------- tidy up
--------------------------------- */
+
+ BATsetcount(newindex, lst_newindex - s_newindex);
+ newindex->batDirty = TRUE;
+ newindex->hsorted = GDK_SORTED;
+ newindex->tsorted = GDK_SORTED;
+ BATkey(newindex, TRUE);
+ BATkey(BATmirror(newindex), FALSE);
+ BATseqbase(newindex, (oid)0);
+
+ BATsetcount(newpre, lst_newpre - s_newpre);
+ newpre->batDirty = TRUE;
+ newpre->hsorted = GDK_SORTED;
+ newpre->tsorted = FALSE;
+ BATkey(newpre, TRUE);
+ BATkey(BATmirror(newpre), TRUE);
+ BATseqbase(newpre, (oid)0);
+
+ BATsetcount(newsize, lst_newsize - s_newsize);
+ newsize->batDirty = TRUE;
+ newsize->hsorted = GDK_SORTED;
+ newsize->tsorted = FALSE;
+ BATkey(newsize, TRUE);
+ BATseqbase(newsize, (oid)0);
+
+ /* insert bats in result */
+ BATseqbase(res, (oid)0);
+ voidfix_bunfastins_nocheck_noinc(res, lst_res, 0,
(void*)&newindex->batCacheid);
+ BBPunfix(newindex->batCacheid);
+ lst_res++;
+ voidfix_bunfastins_nocheck_noinc(res, lst_res, 0,
(void*)&newpre->batCacheid);
+ BBPunfix(newpre->batCacheid);
+ lst_res++;
+ voidfix_bunfastins_nocheck_noinc(res, lst_res, 0,
(void*)&newsize->batCacheid);
+ BBPunfix(newsize->batCacheid);
+ lst_res++;
+
+ BATsetcount(res, 3);
+ res->batDirty = TRUE;
+ BATkey(res, TRUE);
+ BATkey(BATmirror(res), TRUE);
+ res->hsorted = GDK_SORTED;
+ res->tsorted = FALSE;
+
+ *result = res;
+ return GDK_SUCCEED;
+}
+
int CMDindexfetchjoin ( BAT** result, BAT* tid, BAT* index, BAT* pre )
{
char *name = "TJindexfetchjoin";
-------------------------------------------------------------------------
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace
_______________________________________________
Monetdb-pf-checkins mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/monetdb-pf-checkins
