fix(data status): show committed changes in a new Git repository (#10790)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

Author: skshetry

dvc/repo/data.py

@@ -4,6 +4,8 @@
 from typing import TYPE_CHECKING, Optional, TypedDict, Union
 
 from dvc.fs.callbacks import DEFAULT_CALLBACK
+from dvc.log import logger
+from dvc.scm import RevError
 from dvc.ui import ui
 from dvc_data.index.view import DataIndexView
 
@@ -14,6 +16,8 @@
     from dvc_data.index import BaseDataIndex, DataIndex, DataIndexKey
     from dvc_data.index.diff import Change
 
+logger = logger.getChild(__name__)
+
 
 def posixpath_to_os_path(path: str) -> str:
     return path.replace(posixpath.sep, os.path.sep)
@@ -220,12 +224,18 @@ def _diff_head_to_index(
     filter_keys: Optional[list["DataIndexKey"]] = None,
     granular: bool = False,
 ) -> dict[str, list[str]]:
+    from dvc_data.index import DataIndex
+
     index = repo.index.data["repo"]
     index_view = filter_index(index, filter_keys=filter_keys)
 
-    with repo.switch(head):
-        head_index = repo.index.data["repo"]
-        head_view = filter_index(head_index, filter_keys=filter_keys)
+    try:
+        with repo.switch(head):
+            head_index = repo.index.data["repo"]
+            head_view = filter_index(head_index, filter_keys=filter_keys)
+    except RevError:
+        logger.debug("failed to switch to '%s'", head)
+        head_view = DataIndex()
 
     with ui.progress(desc="Calculating diff between head/index", unit="entry") as pb:
         return _diff(

tests/func/test_data_status.py

@@ -144,9 +144,19 @@ def test_new_empty_git_repo(M, tmp_dir, scm):
         "git": M.dict(is_empty=True, is_dirty=True),
     }
 
+    tmp_dir.gen("foo", "foo")
+    dvc.add(["foo"])
+    assert dvc.data_status() == {
+        **EMPTY_STATUS,
+        "git": M.dict(is_empty=True, is_dirty=True),
+        "committed": {"added": ["foo"]},
+    }
+
 
-def test_noscm_repo(dvc):
+def test_noscm_repo(tmp_dir, dvc):
     assert dvc.data_status() == EMPTY_STATUS
+    tmp_dir.dvc_gen("foo", "foo")
+    assert dvc.data_status() == {**EMPTY_STATUS, "unchanged": ["foo"]}
 
 
 def test_unchanged(M, tmp_dir, dvc, scm):