get: implement --show-url to display only url/path to remote

dvc/api.py

@@ -10,7 +10,7 @@
 from voluptuous import Schema, Required, Invalid
 
 from dvc.repo import Repo
-from dvc.exceptions import DvcException
+from dvc.exceptions import DvcException, NotDvcRepoError
 from dvc.external_repo import external_repo
 
 
@@ -43,13 +43,27 @@ class SummonError(DvcException):
     pass
 
 
+class UrlNotDvcRepoError(DvcException):
+    """Thrown if given url is not a DVC repository.
+
+    Args:
+        url (str): url to the repository.
+    """
+
+    def __init__(self, url):
+        super().__init__("URL '{}' is not a dvc repository.".format(url))
+
+
 def get_url(path, repo=None, rev=None, remote=None):
     """Returns an url of a resource specified by path in repo"""
-    with _make_repo(repo, rev=rev) as _repo:
-        abspath = os.path.join(_repo.root_dir, path)
-        out, = _repo.find_outs_by_path(abspath)
-        remote_obj = _repo.cloud.get_remote(remote)
-        return str(remote_obj.checksum_to_path_info(out.checksum))
+    try:
+        with _make_repo(repo, rev=rev) as _repo:
+            abspath = os.path.join(_repo.root_dir, path)
+            out, = _repo.find_outs_by_path(abspath)
+            remote_obj = _repo.cloud.get_remote(remote)
+            return str(remote_obj.checksum_to_path_info(out.checksum))
+    except NotDvcRepoError:
+        raise UrlNotDvcRepoError(repo)
 
 
 def open(path, repo=None, rev=None, remote=None, mode="r", encoding=None):

dvc/command/get.py

@@ -10,7 +10,27 @@
 
 
 class CmdGet(CmdBaseNoRepo):
+    def _show_url(self):
+        from dvc.api import get_url
+
+        try:
+            url = get_url(
+                self.args.path, repo=self.args.url, rev=self.args.rev
+            )
+            logger.info(url)
+        except DvcException:
+            logger.exception("failed to show url")
+            return 1
+
+        return 0
+
     def run(self):
+        if self.args.show_url:
+            return self._show_url()
+
+        return self._get_file_from_repo()
+
+    def _get_file_from_repo(self):
         from dvc.repo import Repo
 
         try:
@@ -55,4 +75,9 @@ def add_parser(subparsers, parent_parser):
     get_parser.add_argument(
         "--rev", nargs="?", help="Git revision (e.g. branch, tag, SHA)"
     )
+    get_parser.add_argument(
+        "--show-url",
+        action="store_true",
+        help="Returns path/url to the location in remote for specified path",
+    )
     get_parser.set_defaults(func=CmdGet)

tests/func/test_api.py

@@ -6,7 +6,7 @@
 import pytest
 
 from dvc import api
-from dvc.api import SummonError
+from dvc.api import SummonError, UrlNotDvcRepoError
 from dvc.compat import fspath
 from dvc.exceptions import FileMissingError
 from dvc.main import main
@@ -51,6 +51,16 @@ def test_get_url_external(erepo_dir, remote_url):
     assert api.get_url("foo", repo=repo_url) == expected_url
 
 
+def test_get_url_git_only_repo_throws_exception(tmp_dir, scm):
+    tmp_dir.scm_gen({"foo": "foo"}, commit="initial")
+
+    with pytest.raises(UrlNotDvcRepoError) as exc_info:
+        api.get_url("foo", fspath(tmp_dir))
+
+    # On windows, `exc_info` has path escaped, eg: `C:\\\\Users\\\\travis`
+    assert fspath(tmp_dir) in str(exc_info).replace("\\\\", "\\")
+
+
 @pytest.mark.parametrize("remote_url", all_remote_params, indirect=True)
 def test_open(remote_url, tmp_dir, dvc):
     run_dvc("remote", "add", "-d", "upstream", remote_url)

tests/func/test_get.py

@@ -5,6 +5,7 @@
 
 from dvc.cache import Cache
 from dvc.config import Config
+from dvc.main import main
 from dvc.repo.get import GetDVCFileError, PathMissingError
 from dvc.repo import Repo
 from dvc.system import System
@@ -184,3 +185,30 @@ def test_get_from_non_dvc_master(tmp_dir, erepo_dir, caplog):
 
     assert caplog.text == ""
     assert (tmp_dir / dst).read_text() == "some_contents"
+
+
+def test_get_url_positive(tmp_dir, erepo_dir, caplog):
+    with erepo_dir.chdir():
+        erepo_dir.dvc_gen("foo", "foo")
+
+    caplog.clear()
+    with caplog.at_level(logging.ERROR, logger="dvc"):
+        assert main(["get", fspath(erepo_dir), "foo", "--show-url"]) == 0
+        assert caplog.text == ""
+
+
+def test_get_url_not_existing(tmp_dir, erepo_dir, caplog):
+    with caplog.at_level(logging.ERROR, logger="dvc"):
+        assert (
+            main(["get", fspath(erepo_dir), "not-existing-file", "--show-url"])
+            == 1
+        )
+        assert "failed to show url" in caplog.text
+
+
+def test_get_url_git_only_repo(tmp_dir, scm, caplog):
+    tmp_dir.scm_gen({"foo": "foo"}, commit="initial")
+
+    with caplog.at_level(logging.ERROR):
+        assert main(["get", fspath(tmp_dir), "foo", "--show-url"]) == 1
+        assert "failed to show url" in caplog.text

tests/unit/command/test_get.py

@@ -14,3 +14,18 @@ def test_get(mocker):
     assert cmd.run() == 0
 
     m.assert_called_once_with("repo_url", path="src", out="out", rev="version")
+
+
+def test_get_url(mocker, caplog):
+    cli_args = parse_args(
+        ["get", "repo_url", "src", "--rev", "version", "--show-url"]
+    )
+    assert cli_args.func == CmdGet
+
+    cmd = cli_args.func(cli_args)
+    m = mocker.patch("dvc.api.get_url", return_value="resource_url")
+
+    assert cmd.run() == 0
+    assert "resource_url" in caplog.text
+
+    m.assert_called_once_with("src", repo="repo_url", rev="version")