From 5a711e8b9ec28c080d590753d3e6a0d68440701d Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Pawe=C5=82=20Redzy=C5=84ski?= <pawelredzynski@gmail.com>
Date: Wed, 19 Feb 2020 17:04:31 +0100
Subject: [PATCH] Stage: get checksum from repo if no checksum in stage file

---
 dvc/output/base.py               | 21 ++++++++++----
 tests/func/test_stage.py         |  1 +
 tests/unit/output/test_output.py | 50 +++++++++++++++++++++++++++++++-
 3 files changed, 66 insertions(+), 6 deletions(-)

diff --git a/dvc/output/base.py b/dvc/output/base.py
index d59276da7a..d07010ad6a 100644
--- a/dvc/output/base.py
+++ b/dvc/output/base.py
@@ -409,12 +409,23 @@ def get_used_cache(self, **kwargs):
             cache.external[dep.repo_pair].add(dep.def_path)
             return cache
 
-        if not self.info:
-            logger.warning(
-                "Output '{}'({}) is missing version info. Cache for it will "
-                "not be collected. Use `dvc repro` to get your pipeline up to "
-                "date.".format(self, self.stage)
+        if not self.checksum:
+            msg = (
+                "Output '{}'({}) is missing version info. "
+                "Cache for it will not be collected. "
+                "Use `dvc repro` to get your pipeline up to date.".format(
+                    self, self.stage
+                )
             )
+            if self.exists:
+                msg += (
+                    "\n"
+                    "You can also use `dvc commit {stage}` to associate "
+                    "existing '{out}' with '{stage}'.".format(
+                        out=self, stage=self.stage.relpath
+                    )
+                )
+            logger.warning(msg)
             return NamedCache()
 
         ret = NamedCache.make(self.scheme, self.checksum, str(self))
diff --git a/tests/func/test_stage.py b/tests/func/test_stage.py
index ce4991fd80..949e8b5671 100644
--- a/tests/func/test_stage.py
+++ b/tests/func/test_stage.py
@@ -1,6 +1,7 @@
 import os
 import tempfile
 
+
 from dvc.main import main
 from dvc.output.local import OutputLOCAL
 from dvc.remote.local import RemoteLOCAL
diff --git a/tests/unit/output/test_output.py b/tests/unit/output/test_output.py
index 2cf30e9680..fb9325471b 100644
--- a/tests/unit/output/test_output.py
+++ b/tests/unit/output/test_output.py
@@ -1,8 +1,12 @@
+import logging
+
 import pytest
+from funcy import first
 
 from voluptuous import Schema, MultipleInvalid
 
-from dvc.output import CHECKSUM_SCHEMA
+from dvc.cache import NamedCache
+from dvc.output import CHECKSUM_SCHEMA, OutputBase
 
 
 @pytest.mark.parametrize(
@@ -40,3 +44,47 @@ def test_checksum_schema(value, expected):
 def test_checksum_schema_fail(value):
     with pytest.raises(MultipleInvalid):
         Schema(CHECKSUM_SCHEMA)(value)["md5"]
+
+
+@pytest.mark.parametrize(
+    "exists, expected_message",
+    [
+        (
+            False,
+            (
+                "Output 'path'(Stage stage.dvc) is missing version info. "
+                "Cache for it will not be collected. "
+                "Use `dvc repro` to get your pipeline up to date."
+            ),
+        ),
+        (
+            True,
+            (
+                "Output 'path'(Stage stage.dvc) is missing version info. "
+                "Cache for it will not be collected. "
+                "Use `dvc repro` to get your pipeline up to date.\n"
+                "You can also use `dvc commit stage.dvc` to associate "
+                "existing 'path' with 'stage.dvc'."
+            ),
+        ),
+    ],
+)
+def test_get_used_cache(exists, expected_message, mocker, caplog):
+    stage = mocker.MagicMock()
+    mocker.patch.object(stage, "__str__", return_value="Stage stage.dvc")
+    mocker.patch.object(stage, "relpath", "stage.dvc")
+
+    output = OutputBase(stage, "path")
+
+    mocker.patch.object(output, "use_cache", True)
+    mocker.patch.object(stage, "is_repo_import", False)
+    mocker.patch.object(
+        OutputBase, "checksum", new_callable=mocker.PropertyMock
+    ).return_value = None
+    mocker.patch.object(
+        OutputBase, "exists", new_callable=mocker.PropertyMock
+    ).return_value = exists
+
+    with caplog.at_level(logging.WARNING, logger="dvc"):
+        assert isinstance(output.get_used_cache(), NamedCache)
+    assert first(caplog.messages) == expected_message