From 9db02cf87e0a9d2ed869355208a6f2f4726b803d Mon Sep 17 00:00:00 2001
From: Jerome Kelleher <jk@well.ox.ac.uk>
Date: Wed, 17 Apr 2024 13:43:28 +0100
Subject: [PATCH 1/2] Filter out empty partitions

Closes #120
Closes #45
---
 CHANGELOG.md            |  1 +
 bio2zarr/vcf_utils.py   | 18 +++++++++++++++++-
 tests/test_vcf_utils.py |  8 +-------
 3 files changed, 19 insertions(+), 8 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index ba20b30..233a255 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -4,6 +4,7 @@
 - Move making ICF field partition directories into per-partition processing.
   Remove progress on the init mkdirs step.
 - Turn off progress monitor on dexplode-partition
+- Fix empty partition bug
 
 # 0.0.4 2024-04-08
 
diff --git a/bio2zarr/vcf_utils.py b/bio2zarr/vcf_utils.py
index d5cc00c..344d615 100644
--- a/bio2zarr/vcf_utils.py
+++ b/bio2zarr/vcf_utils.py
@@ -435,6 +435,22 @@ def variants(self, region):
             if var.POS >= start:
                 yield var
 
+    def _filter_empty(self, regions):
+        """
+        Return all regions in the specified list that have one or more records.
+
+        Sometimes with Tabix indexes these seem to crop up:
+
+        - https://github.com/sgkit-dev/bio2zarr/issues/45
+        - https://github.com/sgkit-dev/bio2zarr/issues/120
+        """
+        ret = []
+        for region in regions:
+            variants = self.variants(region)
+            if next(variants, None) is not None:
+                ret.append(region)
+        return ret
+
     def partition_into_regions(
         self,
         num_parts: Optional[int] = None,
@@ -511,4 +527,4 @@ def partition_into_regions(
             if self.index.record_counts[ri] > 0:
                 regions.append(Region(self.sequence_names[ri]))
 
-        return regions
+        return self._filter_empty(regions)
diff --git a/tests/test_vcf_utils.py b/tests/test_vcf_utils.py
index 03cac42..898dcd6 100644
--- a/tests/test_vcf_utils.py
+++ b/tests/test_vcf_utils.py
@@ -9,13 +9,7 @@
 
 
 def assert_part_counts_non_zero(part_counts, index_file):
-    # We may have one zero count value at the end in Tabix indexes.
-    # Should probably try to get rid of it, but probably no harm
-    # https://github.com/jeromekelleher/bio2zarr/issues/45
-    if index_file.endswith(".tbi"):
-        assert np.all(part_counts[:-1] > 0)
-    else:
-        assert np.all(part_counts > 0)
+    assert np.all(part_counts > 0)
 
 
 class TestIndexedVcf:

From d17ee5dbdbc262d56d73ec7f16017871ce80d7be Mon Sep 17 00:00:00 2001
From: Jerome Kelleher <jk@well.ox.ac.uk>
Date: Wed, 17 Apr 2024 13:45:33 +0100
Subject: [PATCH 2/2] Update CHANGELOG for 0.0.5

---
 CHANGELOG.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 233a255..2b328c1 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,4 +1,4 @@
-# 0.0.5 2024-04-XX
+# 0.0.5 2024-04-17
 
 - Fix bug in schema handling (compressor settings ignored)
 - Move making ICF field partition directories into per-partition processing.