pull: Add allow_missing flag.

Allows `pull` to succeed even if some files or directories are missing.

Closes #4746

dvc/commands/data_sync.py

@@ -40,6 +40,7 @@ class CmdDataPull(CmdDataBase):
                 recursive=self.args.recursive,
                 run_cache=self.args.run_cache,
                 glob=self.args.glob,
+                allow_missing=self.args.allow_missing,
             )
             self.log_summary(stats)
         except (CheckoutError, DvcException) as exc:
@@ -197,6 +198,12 @@ def add_parser(subparsers, _parent_parser):
         default=False,
         help=argparse.SUPPRESS,
     )
+    pull_parser.add_argument(
+        "--allow-missing",
+        action="store_true",
+        default=False,
+        help="Ignore errors if some of the files or directories are missing.",
+    )
     pull_parser.set_defaults(func=CmdDataPull)
 
     # Push

dvc/repo/pull.py

@@ -25,6 +25,7 @@ def pull(  # noqa: PLR0913
     run_cache=False,
     glob=False,
     odb: Optional["ObjectDB"] = None,
+    allow_missing=False,
 ):
     if isinstance(targets, str):
         targets = [targets]
@@ -48,6 +49,7 @@ def pull(  # noqa: PLR0913
         with_deps=with_deps,
         force=force,
         recursive=recursive,
+        allow_missing=allow_missing,
     )
 
     stats["fetched"] = processed_files_count

tests/func/test_data_cloud.py

@@ -7,6 +7,7 @@ from flaky.flaky_decorator import flaky
 
 import dvc_data
 from dvc.cli import main
+from dvc.exceptions import CheckoutError
 from dvc.external_repo import clean_repos
 from dvc.stage.exceptions import StageNotFound
 from dvc.testing.remote_tests import TestRemote  # noqa, pylint: disable=unused-import
@@ -561,3 +562,17 @@ def test_target_remote(tmp_dir, dvc, make_remote):
         "6b18131dc289fd37006705affe961ef8.dir",
         "b8a9f715dbb64fd5c56e7783c6820a61",
     }
+
+
+def test_pull_allow_missing(tmp_dir, dvc, local_remote):
+    dvc.stage.add(name="bar", outs=["bar"], cmd="echo bar > bar")
+
+    with pytest.raises(CheckoutError):
+        dvc.pull()
+
+    tmp_dir.dvc_gen("foo", "foo")
+    dvc.push()
+    clean(["foo"], dvc)
+
+    stats = dvc.pull(allow_missing=True)
+    assert stats["fetched"] == 1

tests/unit/command/test_data_sync.py

@@ -58,6 +58,7 @@ def test_pull(mocker):
             "--recursive",
             "--run-cache",
             "--glob",
+            "--allow-missing",
         ]
     )
     assert cli_args.func == CmdDataPull
@@ -79,6 +80,7 @@ def test_pull(mocker):
         recursive=True,
         run_cache=True,
         glob=True,
+        allow_missing=True,
     )