From 2ebc5f0d751cdd9f6fb089b17a361d606da1cfa5 Mon Sep 17 00:00:00 2001
From: Eric Coissac <eric@coissac.eu>
Date: Mon, 1 Jun 2026 15:18:12 +0200
Subject: [PATCH] chore: add logging infrastructure to merge routine

Adds comprehensive logging for source metadata, merge modes, and forced approximation detection. Introduces `format_evidence` and `is_trivial` helpers to format `IndexMode` variants and identify single-genome presence indices. The core merge algorithm remains unmodified, with all changes focused on enhanced runtime observability.
---
 src/obikindex/src/merge.rs | 38 +++++++++++++++++++++++++++++++++++++-
 1 file changed, 37 insertions(+), 1 deletion(-)

diff --git a/src/obikindex/src/merge.rs b/src/obikindex/src/merge.rs
index 68d8db7..39b4395 100644
--- a/src/obikindex/src/merge.rs
+++ b/src/obikindex/src/merge.rs
@@ -63,8 +63,36 @@ impl KmerIndex {
             }
         }
 
-        // ── Choose base: largest source in the output evidence mode ───────────
+        // ── Log source characteristics and choose base ────────────────────────
+        let mode_str = if mode == MergeMode::Presence { "presence" } else { "count" };
+        info!(
+            "merge: {} source(s), smer-size={}, mode={}",
+            sources.len(), sources[0].kmer_size(), mode_str,
+        );
+        for (i, src) in sources.iter().enumerate() {
+            let genome_str = if src.meta.genomes.len() == 1 { "mono-genome".to_string() }
+                             else { format!("{} genomes", src.meta.genomes.len()) };
+            let trivial_str = if is_trivial(src, mode) { " [trivial: no data approximation]" } else { "" };
+            info!(
+                "  [{}] {} — {}, {}, {}{}",
+                i, src.root_path.display(),
+                format_evidence(&src.meta.config.evidence),
+                genome_str, mode_str, trivial_str,
+            );
+        }
+
         let base_idx = choose_base(sources, mode);
+        let needs_approx = sources.iter().any(|src| {
+            !is_trivial(src, mode)
+                && matches!(src.meta.config.evidence, IndexMode::Approx { .. } | IndexMode::Hybrid { .. })
+        });
+        info!(
+            "output evidence: {} ({}base: [{}] {})",
+            format_evidence(&sources[base_idx].meta.config.evidence),
+            if needs_approx { "forced approx — " } else { "" },
+            base_idx, sources[base_idx].root_path.display(),
+        );
+
         let mut ordered: Vec<&KmerIndex> = Vec::with_capacity(sources.len());
         ordered.push(sources[base_idx]);
         for (i, &src) in sources.iter().enumerate() {
@@ -272,6 +300,14 @@ fn partition_bar(n: u64) -> ProgressBar {
     pb
 }
 
+fn format_evidence(ev: &IndexMode) -> String {
+    match ev {
+        IndexMode::Exact             => "exact".to_string(),
+        IndexMode::Approx { b, z }   => format!("approx (b={b}, z={z})"),
+        IndexMode::Hybrid { b, z }   => format!("hybrid (b={b}, z={z})"),
+    }
+}
+
 /// A source is "trivial" if its presence/count values carry no approximation:
 /// single-genome presence index (SetMembership — all values are 1 by construction).
 fn is_trivial(src: &KmerIndex, mode: MergeMode) -> bool {