fix: address tag metadata review feedback

Author: majin1102

docs/src/format/table/branch_tag.md

@@ -121,5 +121,5 @@ Each tag file is a JSON file with the following fields:
 | `version`       | number |          | Version number being tagged within that branch.                          |
 | `createdAt`     | string | Yes      | RFC 3339 timestamp for when the tag was first created.                  |
 | `updatedAt`     | string | Yes      | RFC 3339 timestamp for the latest tag reference update.                 |
-| `manifest_size` | number |          | Size of the manifest file in bytes. Used for efficient manifest loading. |
+| `manifestSize`  | number |          | Size of the manifest file in bytes. Used for efficient manifest loading. |
 | `metadata`      | object | Yes      | String key/value metadata map. If absent, it is treated as an empty object. |

python/python/lance/dataset.py

@@ -5824,6 +5824,11 @@ def update(
     def replace_metadata(self, tag: str, metadata: Dict[str, str]) -> None:
         """
         Replace metadata for an existing tag.
+
+        This replaces the entire metadata map rather than merging with existing
+        keys. It does not change the tag's referenced branch/version, and it
+        does not update ``updated_at``. The ``updated_at`` timestamp only
+        changes when :meth:`update` moves the tag to a different reference.
         """
         self._ds.replace_tag_metadata(tag, metadata)
 

python/python/tests/test_dataset.py

@@ -521,7 +521,14 @@ def test_tag(tmp_path: Path):
     ds.tags.replace_metadata("tag1", {"description": "updated tag"})
     ds = lance.dataset(base_dir, "tag1")
     assert ds.version == 1
-    assert ds.tags.list()["tag1"]["metadata"] == {"description": "updated tag"}
+    replaced_tag1_meta = ds.tags.list()["tag1"]
+    assert replaced_tag1_meta["metadata"] == {"description": "updated tag"}
+    assert replaced_tag1_meta["updated_at"] == tag1_updated_at
+
+    ds.tags.replace_metadata("tag1", {"owner": "ml-team"})
+    replaced_again_tag1_meta = ds.tags.list()["tag1"]
+    assert replaced_again_tag1_meta["metadata"] == {"owner": "ml-team"}
+    assert replaced_again_tag1_meta["updated_at"] == tag1_updated_at
 
     ds.tags.update("tag1", 2)
     updated_tag1_meta = ds.tags.list()["tag1"]
@@ -5266,10 +5273,13 @@ def test_branches(tmp_path: Path):
     branch1.tags.create("main_latest", (None, None))
     branch1.tags.create("main_latest2", ("main", None))
     branch1.create_branch("branch_from_main", ("main", None))
+    ordered_tags = dict(branch1.tags.list_ordered())
     branches_with_main = branch1.branches.list()
     assert branch1.tags.list()["branch1_latest"]["branch"] == "branch1"
     assert branch1.tags.list()["main_latest"]["branch"] is None
     assert branch1.tags.list()["main_latest2"]["branch"] is None
+    assert ordered_tags["branch1_latest"]["branch"] == "branch1"
+    assert ordered_tags["main_latest"]["branch"] is None
     assert branches_with_main["branch_from_main"]["parent_branch"] is None
     assert branches_with_main["branch_from_main"]["branch_identifier"][0][0] == 1
     assert isinstance(

python/src/dataset.rs

@@ -1608,6 +1608,7 @@ impl Dataset {
 
         for (tag_name, tag_content) in tags {
             let dict = PyDict::new(py);
+            dict.set_item("branch", tag_content.branch.clone())?;
             dict.set_item("version", tag_content.version)?;
             dict.set_item("created_at", tag_content.created_at)?;
             dict.set_item("updated_at", tag_content.updated_at)?;