get: handle non-DVC repositories

Allows us to `dvc get` from non-DVC source repositories.

Fixes #3089

Author: fabiosantoscode

dvc/exceptions.py

@@ -240,11 +240,6 @@ def __init__(self, ignore_dirname):
         )
 
 
-class UrlNotDvcRepoError(DvcException):
-    def __init__(self, url):
-        super().__init__("URL '{}' is not a dvc repository.".format(url))
-
-
 class GitHookAlreadyExistsError(DvcException):
     def __init__(self, hook_name):
         super().__init__(

dvc/external_repo.py

@@ -33,7 +33,7 @@ def external_repo(url=None, rev=None, rev_lock=None, cache_dir=None):
     repo.close()
 
 
-def cached_clone(url, rev=None, **_ignored_kwargs):
+def cached_clone(url, rev=None, clone_path=None, **_ignored_kwargs):
     """Clone an external git repo to a temporary directory.
 
     Returns the path to a local temporary directory with the specified
@@ -44,7 +44,7 @@ def cached_clone(url, rev=None, **_ignored_kwargs):
 
     """
 
-    new_path = tempfile.mkdtemp("dvc-erepo")
+    new_path = clone_path or tempfile.mkdtemp("dvc-erepo")
 
     # Copy and adjust existing clean clone
     if (url, None, None) in REPO_CACHE:

dvc/repo/get.py

@@ -7,10 +7,9 @@
     DvcException,
     NotDvcRepoError,
     OutputNotFoundError,
-    UrlNotDvcRepoError,
     PathMissingError,
 )
-from dvc.external_repo import external_repo
+from dvc.external_repo import cached_clone
 from dvc.path_info import PathInfo
 from dvc.stage import Stage
 from dvc.utils import resolve_output
@@ -28,8 +27,15 @@ def __init__(self):
         )
 
 
+# Dummy exception raised to signal a plain file copy is needed
+class _DoPlainCopy(DvcException):
+    pass
+
+
 @staticmethod
 def get(url, path, out=None, rev=None):
+    from dvc.repo import Repo
+
     out = resolve_output(path, out)
 
     if Stage.is_valid_filename(out):
@@ -43,7 +49,8 @@ def get(url, path, out=None, rev=None):
     dpath = os.path.dirname(os.path.abspath(out))
     tmp_dir = os.path.join(dpath, "." + str(shortuuid.uuid()))
     try:
-        with external_repo(cache_dir=tmp_dir, url=url, rev=rev) as repo:
+        cached_clone(url, rev=rev, clone_path=tmp_dir)
+        try:
             # Try any links possible to avoid data duplication.
             #
             # Not using symlink, because we need to remove cache after we are
@@ -53,26 +60,24 @@ def get(url, path, out=None, rev=None):
             #
             # Also, we can't use theoretical "move" link type here, because
             # the same cache file might be used a few times in a directory.
+            repo = Repo(tmp_dir)
             repo.cache.local.cache_types = ["reflink", "hardlink", "copy"]
+            output = repo.find_out_by_relpath(path)
+            if not output.use_cache:
+                # Catch this below and go for a plain old fs_copy
+                raise _DoPlainCopy
+            _get_cached(repo, output, out)
 
-            try:
-                output = repo.find_out_by_relpath(path)
-            except OutputNotFoundError:
-                output = None
-
-            if output and output.use_cache:
-                _get_cached(repo, output, out)
-            else:
-                # Either an uncached out with absolute path or a user error
-                if os.path.isabs(path):
-                    raise FileNotFoundError
+        except (NotDvcRepoError, OutputNotFoundError, _DoPlainCopy):
+            # It's an uncached out with absolute path, a non-DVC repo, or a
+            # user error
+            if os.path.isabs(path):
+                raise FileNotFoundError
 
-                fs_copy(os.path.join(repo.root_dir, path), out)
+            fs_copy(os.path.join(tmp_dir, path), out)
 
     except (OutputNotFoundError, FileNotFoundError):
         raise PathMissingError(path, url)
-    except NotDvcRepoError:
-        raise UrlNotDvcRepoError(url)
     finally:
         remove(tmp_dir)
 

tests/func/test_get.py

@@ -5,7 +5,6 @@
 
 from dvc.cache import Cache
 from dvc.config import Config
-from dvc.exceptions import UrlNotDvcRepoError
 from dvc.repo.get import GetDVCFileError, PathMissingError
 from dvc.repo import Repo
 from dvc.system import System
@@ -87,9 +86,10 @@ def test_get_repo_rev(tmp_dir, erepo_dir):
 def test_get_from_non_dvc_repo(tmp_dir, erepo_dir):
     erepo_dir.scm.repo.index.remove([erepo_dir.dvc.dvc_dir], r=True)
     erepo_dir.scm.commit("remove dvc")
+    erepo_dir.scm_gen({"some_file": "contents"}, commit="create file")
 
-    with pytest.raises(UrlNotDvcRepoError):
-        Repo.get(fspath(erepo_dir), "some_file.zip")
+    Repo.get(fspath(erepo_dir), "some_file", "file_imported")
+    assert (tmp_dir / "file_imported").read_text() == "contents"
 
 
 def test_get_a_dvc_file(tmp_dir, erepo_dir):
@@ -164,10 +164,6 @@ def test_get_from_non_dvc_master(tmp_dir, erepo_dir, caplog):
         erepo_dir.dvc.scm.repo.index.remove([".dvc"], r=True)
         erepo_dir.dvc.scm.commit("remove .dvc")
 
-    # sanity check
-    with pytest.raises(UrlNotDvcRepoError):
-        Repo.get(fspath(erepo_dir), "some_file")
-
     caplog.clear()
     dst = "file_imported"
     with caplog.at_level(logging.INFO, logger="dvc"):