show docids of corrupted title recs found.

show key range of each dump to disk.
fix 'sentToDiffbot' bug for unchanged docs in status docs.
make sure firstKeyInQueue is set properly from current key,
so reset list ptr before doing that in RdbDump.cpp.

RdbDump.cpp

@@ -412,7 +412,10 @@ bool RdbDump::dumpTree ( bool recall ) {
 			char *s = "none";
 			if ( m_rdb ) s = getDbnameFromId(m_rdb->m_rdbId);
 			log("dump: verifying list before dumping (rdb=%s "
-			    "collnum=%i)",s,(int)m_collnum);
+			    "collnum=%i k1=%s k2=%s)",s,
+			    (int)m_collnum,
+			    KEYSTR(m_firstKeyInQueue,m_list->m_ks),
+			    KEYSTR(m_lastKeyInQueue ,m_list->m_ks));
 			m_list->checkList_r ( false , // removeNegRecs?
 					      false , // sleep on problem?
 					      m_rdb->m_rdbId );
@@ -460,6 +463,10 @@ bool RdbDump::dumpTree ( bool recall ) {
 	// . this doesn't work if you're doing an unordered dump, but we should
 	//   not allow adds when closing
 	m_lastKeyInQueue  = m_list->getLastKey();
+
+	// ensure we are getting the first key of the list
+	m_list->resetListPtr();
+
 	//m_firstKeyInQueue = m_list->getCurrentKey();
 	m_list->getCurrentKey(m_firstKeyInQueue);
 	// . write this list to disk

RdbList.cpp

@@ -742,7 +742,11 @@ bool RdbList::checkList_r ( bool removeNegRecs , bool sleepOnProblem ,
 			if ( data && 
 			     (*(int32_t *)data < 0 || 
 			      *(int32_t *)data > 100000000 ) ) {
-				char *xx = NULL; *xx = 0; }
+				log("rdblist: bad titlerec data for docid "
+				    "%"INT64,
+				    g_titledb.getDocIdFromKey((key_t *)k));
+				char *xx = NULL; *xx = 0; 
+			}
 		}		
 		// tagrec?
 		if ( rdbId == RDB_TAGDB && ! KEYNEG(k) ) {

RdbTree.cpp

@@ -1187,18 +1187,19 @@ bool RdbTree::fixTree ( ) {
 		// skip if empty
 		if ( m_parents[i] <= -2 ) continue;
 
-
+		char *key = &m_keys[i*m_ks];
 		if ( isTitledb && m_data[i] ) {
 			char *data = m_data[i];
 			int32_t ucompSize = *(int32_t *)data;
 			if ( ucompSize < 0 || ucompSize > 100000000 ) {
 				log("db: removing titlerec with uncompressed "
-				     "size of %i from tree",(int)ucompSize);
+				     "size of %i from tree (docid=%"INT64"",
+				    (int)ucompSize,
+				    g_titledb.getDocIdFromKey((key_t *)key));
 				continue;
 			}
 		}
 
-		char *key = &m_keys[i*m_ks];
 		if ( isSpiderdb && m_data[i] &&
 		     g_spiderdb.isSpiderRequest ( (SPIDERDBKEY *)key ) ) {
 			char *data = m_data[i];

XmlDoc.cpp

@@ -4659,6 +4659,11 @@ int32_t *XmlDoc::getIndexCode2 ( ) {
 		if ( *ch32 == od->m_contentHash32 ) {
 			m_indexCode = EDOCUNCHANGED;
 			m_indexCodeValid = true;
+			// hack these values on or off.
+			// really should be function calls.
+			// but it never gets set when it should if the
+			// doc is unchanged.
+			m_sentToDiffbot = od->m_sentToDiffbot;
 			return &m_indexCode;
 		}
 	}