Update of /cvsroot/monetdb/pathfinder/modules/pftijah
In directory sc8-pr-cvs16.sourceforge.net:/tmp/cvs-serv27498/modules/pftijah
Modified Files:
Tag: GDK-2
nexi.c nexi.h nexi_generate_mil.c pftijah.mx
Log Message:
propagated changes of Tuesday Aug 14 2007 - Wednesday Aug 15 2007
from the development trunk to the GDK-2 branch
Index: nexi_generate_mil.c
===================================================================
RCS file: /cvsroot/monetdb/pathfinder/modules/pftijah/nexi_generate_mil.c,v
retrieving revision 1.35
retrieving revision 1.35.2.1
diff -u -d -r1.35 -r1.35.2.1
--- nexi_generate_mil.c 15 Jun 2007 14:41:34 -0000 1.35
+++ nexi_generate_mil.c 15 Aug 2007 22:23:36 -0000 1.35.2.1
@@ -625,10 +625,12 @@
case P_PRIOR:
com_nr_left = p_com->left->number;
-
- MILPRINTF(MILOUT, "R%d := R%d.prior_ls(%d,qenv);\n", com_num,
com_nr_left, txt_retr_model->size_type);
-
- break;
+ if (txt_retr_model->prior_type == LOG_LENGTH_PRIOR) {
+ MILPRINTF(MILOUT, "R%d := R%d.prior_logls(%d,qenv);\n", com_num,
com_nr_left, txt_retr_model->size_type);
+ } else {
+ MILPRINTF(MILOUT, "R%d := R%d.prior_ls(%d,qenv);\n", com_num,
com_nr_left, txt_retr_model->size_type);
+ }
+ break;
case MUST_CONTAIN_T:
@@ -1058,6 +1060,9 @@
MILPRINTF(MILOUT, "R%d := collect;\n",com_num);
MILPRINTF(MILOUT, "collect := nil;\n");
}
+ if(txt_retr_model->rmoverlap){
+ MILPRINTF(MILOUT, "R%d := rm_overlap(R%d%s);\n", com_num,
com_num,parserCtx->flastPfx);
+ }
MILPRINTF(MILOUT, "R%d := R%d.tsort_rev();\n", com_num, com_num);
#if 0
MILPRINTF(MILOUT, "if ( retNum >= 0 ) { R%d := R%d.slice(0, retNum - 1);
}\n", com_num, com_num);
Index: nexi.h
===================================================================
RCS file: /cvsroot/monetdb/pathfinder/modules/pftijah/nexi.h,v
retrieving revision 1.17
retrieving revision 1.17.4.1
diff -u -d -r1.17 -r1.17.4.1
--- nexi.h 4 Apr 2007 13:52:05 -0000 1.17
+++ nexi.h 15 Aug 2007 22:23:36 -0000 1.17.4.1
@@ -120,6 +120,7 @@
#define NO_PRIOR 1
#define LENGTH_PRIOR 2
#define LOG_NORMAL_PRIOR 3
+#define LOG_LENGTH_PRIOR 4
#define PREF_ELEM_SIZE 1000
@@ -269,6 +270,7 @@
int param3;
int prior_type;
int prior_size;
+ bool rmoverlap;
char context[TERM_LENGTH];
float extra;
struct_RMT *next;
Index: nexi.c
===================================================================
RCS file: /cvsroot/monetdb/pathfinder/modules/pftijah/nexi.c,v
retrieving revision 1.69.2.1
retrieving revision 1.69.2.2
diff -u -d -r1.69.2.1 -r1.69.2.2
--- nexi.c 14 Aug 2007 14:03:48 -0000 1.69.2.1
+++ nexi.c 15 Aug 2007 22:23:35 -0000 1.69.2.2
@@ -217,6 +217,7 @@
//txt_retr_model->prior_type = LENGTH_PRIOR;
//txt_retr_model->prior_size = 0;
txt_retr_model->prior_size = 2;
+ txt_retr_model->rmoverlap = FALSE;
strcpy(txt_retr_model->context, "");
txt_retr_model->extra = 0.0;
txt_retr_model->next = NULL;
@@ -462,11 +463,19 @@
} else if (strcmp(optName, "prior") == 0) {
if (strcasecmp(optVal, "LENGTH_PRIOR") == 0) {
txt_retr_model->prior_type = LENGTH_PRIOR;
+ } else if (strcasecmp(optVal, "LOG_LENGTH_PRIOR") == 0) {
+ txt_retr_model->prior_type = LOG_LENGTH_PRIOR;
} else {
txt_retr_model->prior_type = NO_PRIOR;
}
- } else if (strcmp(optName, "returnNumber") == 0) {
+ } else if (strcmp(optName, "rmoverlap") == 0) {
+ if (strcasecmp(optVal, "TRUE") == 0) {
+ txt_retr_model->rmoverlap = TRUE;
+ } else {
+ txt_retr_model->rmoverlap=FALSE;
+ }
+ } else if (strcmp(optName, "returnNumber") == 0) {
// ignore, is handled by milprint_summer
} else if (strcmp(optName, "term-proximity") == 0) {
qenv_prox_val = (char*)strdup(optVal);
Index: pftijah.mx
===================================================================
RCS file: /cvsroot/monetdb/pathfinder/modules/pftijah/pftijah.mx,v
retrieving revision 1.151.2.1
retrieving revision 1.151.2.2
diff -u -d -r1.151.2.1 -r1.151.2.2
--- pftijah.mx 14 Aug 2007 14:03:48 -0000 1.151.2.1
+++ pftijah.mx 15 Aug 2007 22:23:37 -0000 1.151.2.2
@@ -2690,6 +2690,62 @@
#"Adds element prior based on element length to the region score.",
#"TIJAH");
+PROC prior_logls(bat region, int size_type, BAT[oid,str] qenv) : bat :=
+{
+
+ var ctx_size := new(oid,dbl,ENTITY_NUM);
+ var res_reg := new(oid,dbl,ENTITY_NUM);
+
+ if (size_type = SIZE_ENTITY) {
+ ctx_size := size_entity(region,qenv);
+ # Make sure sizes are not zero
+ ctx_size := [+](ctx_size, dbl(0.5));
+ } else if (size_type = SIZE_TERM) {
+ ctx_size := size_term(region,qenv);
+ # Make sure sizes are not zero
+ ctx_size := [+](ctx_size, dbl(0.5));
+ }
+
+ res_reg :=
[+]([log](region.select(dbl(0),dbl(nil),false,false)),[log]([log](ctx_size)));
+ res_reg :=
[-](region.select(dbl(0)).project(min(res_reg)),dbl(1)).kunion(res_reg);
+
+ return res_reg;
+
+}
+#ADDHELP("prior_logls", "thijs", "Aug, 2007,
+#"Adds element prior based on log element length to the region score.",
+#"TIJAH");
+
+################################################################################
+# OVERLAP REMOVAL
+################################################################################
+
+PROC rm_overlap(bat ctx,str ind) : bat{
+
+ ctx := ctx.sort();
+ var size := bat("tj_"+qenv.find(QENV_FTINAME)+"_size"+ind);
+ var AD:=anc_desc(ctx,ctx,size);
+ var maxD := AD.join(ctx).sort().{max}();
+ var maxA := AD.reverse().sort().join(ctx).{max}();
+
+ AD:=nil;
+
+ # candidates are nodes that have no descendents or a higher score
+ # than all descendents
+ var candidates := ctx.kdiff(maxD).access(BAT_APPEND);
+ candidates.insert(ctx.[>](maxD).select(true).mirror().join(ctx));
+
+ # keep the candidates without ancestor and w score >= ancestor scores
+ var res := candidates.kdiff(maxA).access(BAT_APPEND);
+ res.insert(candidates.[>=](maxA).select(true).mirror().join(candidates));
+
+ candidates := nil;
+ maxA:=nil;maxD:=nil;
+ return res;
+}
+#ADDHELP("rm_overlap","thijs", "August, 2007",
+#"Removes overlapping elements from result list.",
+#"TIJAH");
################################################################################
-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems? Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Monetdb-pf-checkins mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/monetdb-pf-checkins
