[MRG] output unassigned hashes when there are no gather matches (#613)

Fixes #612 - gather fails to output unassigned hashes when no matches. Adjusts behavior so that full query signature is output as unassigned, instead.

sourmash/commands.py

@@ -1018,7 +1018,7 @@ def gather(args):
                                              sourmash_args.get_moltype(query))
 
     # verify signature was computed right.
-    if query.minhash.max_hash == 0:
+    if query.minhash.scaled == 0:
         error('query signature needs to be created with --scaled')
         sys.exit(-1)
 
@@ -1043,6 +1043,8 @@ def gather(args):
 
     found = []
     weighted_missed = 1
+    new_max_hash = query.minhash.max_hash
+    next_query = query
     for result, weighted_missed, new_max_hash, next_query in gather_databases(query, databases, args.threshold_bp, args.ignore_abundance):
         # print interim result & save in a list for later use
         pct_query = '{:.1f}%'.format(result.f_orig_query*100)
@@ -1086,10 +1088,7 @@ def gather(args):
            (1 - weighted_missed) * 100)
     print_results('')
 
-    if not found:
-        sys.exit(0)
-
-    if args.output:
+    if found and args.output:
         fieldnames = ['intersect_bp', 'f_orig_query', 'f_match',
                       'f_unique_to_query', 'f_unique_weighted',
                       'average_abund', 'median_abund', 'std_abund', 'name', 'filename', 'md5']
@@ -1100,15 +1099,13 @@ def gather(args):
             del d['leaf']                 # actual signature not in CSV.
             w.writerow(d)
 
-    if args.save_matches:
+    if found and args.save_matches:
         outname = args.save_matches.name
         notify('saving all matches to "{}"', outname)
         sig.save_signatures([ r.leaf for r in found ], args.save_matches)
 
     if args.output_unassigned:
-        if not found:
-            notify('nothing found - entire query signature unassigned.')
-        elif not len(query.minhash):
+        if not len(query.minhash):
             notify('no unassigned hashes! not saving.')
         else:
             outname = args.output_unassigned.name

tests/test_sourmash.py

@@ -2976,6 +2976,25 @@ def test_gather_metagenome_output_unassigned():
         assert all(('1.3 Mbp       13.6%   28.2%' in out,
                 'NC_011294.1' in out))
 
+
+@utils.in_tempdir
+def test_gather_metagenome_output_unassigned_nomatches(c):
+    # test --output-unassigned when there are no matches
+    query_sig = utils.get_test_data('2.fa.sig')
+    against_sig = utils.get_test_data('47.fa.sig')
+
+    c.run_sourmash('gather', query_sig, against_sig,
+                   '--output-unassigned', 'foo.sig')
+
+    print(c.last_result.out)
+    assert 'found 0 matches total;' in c.last_result.out
+
+    x = sourmash.load_one_signature(query_sig, ksize=31)
+    y = sourmash.load_one_signature(c.output('foo.sig'))
+
+    assert x.minhash == y.minhash
+
+
 def test_gather_metagenome_downsample():
     with utils.TempDirectory() as location:
         testdata_glob = utils.get_test_data('gather/GCF*.sig')