reduce: add more logging, temporarily

1 files changed, 6 insertions, 1 deletions

diff --git a/skate/reduce.go b/skate/reduce.go
index 37de218..992aaec 100644
--- a/skate/reduce.go
+++ b/skate/reduce.go
@@ -451,11 +451,12 @@ func ZippyBrefAugment(bref, raw io.Reader, w io.Writer) error {
 
 			// First, iterate over all matches and sort out duplicates, e.g.
 			// docs that have the same source and target id.
-			log.Printf("group K=%s, G0=%d, G1=%d", g.Key, len(g.G0), len(g.G1))
+			log.Printf("group K=%s, G0=%d (bref), G1=%d (ref)", g.Key, len(g.G0), len(g.G1))
 			matched, err := uniqueMatches(CutBatch(g.G0, 2), &stats)
 			if err != nil {
 				return err
 			}
+			log.Printf("found %v unique matches", len(matched))
 			var refs = make([]*Ref, len(g.G1))
 			for i := 0; i < len(refs); i++ {
 				var (
@@ -469,12 +470,15 @@ func ZippyBrefAugment(bref, raw io.Reader, w io.Writer) error {
 			}
 			// TODO: this slows down this process; be a bit smarter about slices.
 			matched = matchedRefsExtend(matched, refs, &stats)
+			log.Printf("matchedRefsExtend results in %d docs", len(matched))
 			// At this point, we may have duplicates by "_id", e.g. source
 			// release ident and ref index (example:
 			// 4kg2dejsgzaf3cszs2lt5hz4by_9, which appears three times, one
 			// exact match, and twice unmatched).
 			matched = deduplicateBrefs(matched)
+			log.Printf("deduplicateBrefs results in %d docs", len(matched))
 			matched = removeSelfLinks(matched)
+			log.Printf("removeSelfLinks results in %d docs", len(matched))
 			for _, bref := range matched {
 				stats.total++
 				if err := enc.Encode(bref); err != nil {
@@ -533,6 +537,7 @@ func deduplicateBrefs(brefs []*BiblioRef) []*BiblioRef {
 		i    int
 	)
 	for _, v := range brefs {
+		// XXX: Is this enough, do we get rid of too many things here?
 		if seen.Contains(v.Key) {
 			continue
 		}