X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=index%2Fextract.c;h=7c325cafb2e346358c09cb45493b17c7973319f9;hb=c41c84a497ae744aa825a90f144c85b54f1cd4bb;hp=4309b60e2a99673b0ce50f5b6997affdda740926;hpb=ef696645cc3b7e0f4027008d1dc589c0f0f90c1f;p=idzebra-moved-to-github.git

diff --git a/index/extract.c b/index/extract.c
index 4309b60..7c325ca 100644
--- a/index/extract.c
+++ b/index/extract.c
@@ -4,7 +4,21 @@
  * Sebastian Hammer, Adam Dickmeiss
  *
  * $Log: extract.c,v $
- * Revision 1.96  1999-05-26 07:49:13  adam
+ * Revision 1.100  2000-03-20 19:08:36  adam
+ * Added remote record import using Z39.50 extended services and Segment
+ * Requests.
+ *
+ * Revision 1.99  2000/02/24 10:57:02  adam
+ * Sequence number incremented after each incomplete-field.
+ *
+ * Revision 1.98  1999/09/07 07:19:21  adam
+ * Work on character mapping. Implemented replace rules.
+ *
+ * Revision 1.97  1999/07/06 12:28:04  adam
+ * Updated record index structure. Format includes version ID. Compression
+ * algorithm ID is stored for each record block.
+ *
+ * Revision 1.96  1999/05/26 07:49:13  adam
  * C++ compilation.
  *
  * Revision 1.95  1999/05/21 12:00:17  adam
@@ -406,6 +420,8 @@ int key_open (struct recordGroup *rGroup, int mem)
     BFiles bfs = rGroup->bfs;
     int rw = rGroup->flagRw;
     data1_handle dh = rGroup->dh;
+    char *recordCompression;
+    int record_compression = REC_COMPRESS_NONE;
     if (!mem)
         mem = atoi(res_get_def (common_resource, "memMax", "4"))*1024*1024;
     if (mem < 50000)
@@ -423,7 +439,13 @@ int key_open (struct recordGroup *rGroup, int mem)
 	return -1;
     }
     assert (!records);
-    records = rec_open (bfs, rw);
+    recordCompression = res_get_def (common_resource,
+				     "recordCompression", "none");
+    if (!strcmp (recordCompression, "none"))
+	record_compression = REC_COMPRESS_NONE;
+    if (!strcmp (recordCompression, "bzip2"))
+	record_compression = REC_COMPRESS_BZIP2;
+    records = rec_open (bfs, rw, record_compression);
     if (!records)
     {
 	dict_close (matchDict);
@@ -544,7 +566,7 @@ void key_flush (void)
     logf (LOG_LOG, "sorting section %d", key_file_no);
 #if !SORT_EXTRA
     qsort (key_buf + ptr_top-ptr_i, ptr_i, sizeof(char*), key_qsort_compare);
-    getFnameTmp (out_fname, key_file_no);
+    getFnameTmp (common_resource, out_fname, key_file_no);
 
     if (!(outf = fopen (out_fname, "wb")))
     {
@@ -619,7 +641,7 @@ int key_close (struct recordGroup *rGroup)
     int rw = rGroup->flagRw;
     if (rw)
 	zebraExplain_runNumberIncrement (zti, 1);
-    zebraExplain_close (zti, rw, 0);
+    zebraExplain_close (zti, rw);
     key_flush ();
     xfree (key_buf);
     rec_close (&records);
@@ -637,6 +659,7 @@ static void wordInit (struct recExtractCtrl *p, RecWord *w)
     w->attrSet = VAL_BIB1;
     w->attrUse = 1016;
     w->reg_type = 'w';
+    w->extractCtrl = p;
 }
 
 static struct sortKey {
@@ -794,6 +817,7 @@ static void addIncompleteField (RecWord *p)
 	    return;
 	addString (p, buf, i);
     }
+    (p->seqnos[p->reg_type])++; /* to separate this from next one  */
 }
 
 static void addCompleteField (RecWord *p)
@@ -844,6 +868,13 @@ static void addCompleteField (RecWord *p)
 
 static void addRecordKey (RecWord *p)
 {
+    WRBUF wrbuf;
+    if ((wrbuf = zebra_replace(p->zebra_maps, p->reg_type, 0,
+			       p->string, p->length)))
+    {
+	p->string = wrbuf_buf(wrbuf);
+	p->length = wrbuf_len(wrbuf);
+    }
     if (zebra_maps_is_complete (p->zebra_maps, p->reg_type))
 	addCompleteField (p);
     else
@@ -1279,8 +1310,8 @@ static int recordExtract (SYSNO *sysno, const char *fname,
 	extractCtrl.fh = fi;
 	extractCtrl.subType = subType;
 	extractCtrl.init = wordInit;
-	extractCtrl.addWord = addRecordKey;
-	extractCtrl.addSchema = addSchema;
+	extractCtrl.tokenAdd = addRecordKey;
+	extractCtrl.schemaAdd = addSchema;
 	extractCtrl.dh = rGroup->dh;
 	for (i = 0; i<256; i++)
 	{
@@ -1620,7 +1651,7 @@ int fileExtract (SYSNO *sysno, const char *fname,
     {
         if (zebraExplain_newDatabase (zti, rGroup->databaseName,
 				      rGroup->explainDatabase))
-            abort ();
+	    return 0;
     }
 
     if (rGroup->flagStoreData == -1)
@@ -1654,59 +1685,6 @@ int fileExtract (SYSNO *sysno, const char *fname,
     if (rGroup->flagStoreKeys == -1)
         rGroup->flagStoreKeys = 0;
 
-#if ZEBRASDR
-    if (rGroup->useSDR)
-    {
-	ZebraSdrHandle h;
-	char xname[128], *xp;
-
-	strncpy (xname, fname, 127);
-	if (!(xp = strchr (xname, '.')))
-	    return 0;
-	*xp = '\0';
-	if (strcmp (xp+1, "sdr.bits"))
-	    return 0;
-
-        h = zebraSdr_open (xname);
-	if (!h)
-	{
-	    logf (LOG_WARN, "sdr open %s", xname);
-	    return 0;
-	}
-	for (;;)
-	{
-    	    unsigned char *buf;
-	    char sdr_name[128];
-	    int r, segmentno;
-
-	    segmentno = zebraSdr_segment (h, 0);
-	    sprintf (sdr_name, "%%%s.%d", xname, segmentno);
-
-#if 0
-	    if (segmentno > 20)
-		break;
-#endif
-	    r = zebraSdr_read (h, &buf);
-
-    	    if (!r)
-		break;
-
-            fi = file_read_start (0);
-	    fi->sdrbuf = buf;
-	    fi->sdrmax = r;
-	    do
-	    {
-		file_begin (fi);
-		r = recordExtract (sysno, sdr_name, rGroup, deleteFlag, fi,
-                           recType, subType);
-	    } while (r && !sysno && fi->file_more);
-	    file_read_stop (fi);
-	    free (buf);
-	}
-	zebraSdr_close (h);
-	return 1;
-    }
-#endif
     if (sysno && deleteFlag)
         fd = -1;
     else
@@ -1748,8 +1726,8 @@ static int explain_extract (void *handle, Record rec, data1_node *n)
     reckeys.prevSeqNo = 0;
     
     extractCtrl.init = wordInit;
-    extractCtrl.addWord = addRecordKey;
-    extractCtrl.addSchema = addSchema;
+    extractCtrl.tokenAdd = addRecordKey;
+    extractCtrl.schemaAdd = addSchema;
     extractCtrl.dh = rGroup->dh;
     for (i = 0; i<256; i++)
 	extractCtrl.seqno[i] = 0;