Split create_stage out of dvcfile to Stage::create() and overwrite

dvc/command/run.py

@@ -47,7 +47,6 @@ def run(self):
                 no_exec=self.args.no_exec,
                 overwrite=overwrite,
                 ignore_build_cache=self.args.ignore_build_cache,
-                remove_outs=self.args.remove_outs,
                 no_commit=self.args.no_commit,
                 outs_persist=self.args.outs_persist,
                 outs_persist_no_cache=self.args.outs_persist_no_cache,
@@ -168,12 +167,6 @@ def add_parser(subparsers, parent_parser):
         help="Run this stage even if it has been already ran with the same "
         "command/dependencies/outputs/etc before.",
     )
-    run_parser.add_argument(
-        "--remove-outs",
-        action="store_true",
-        default=False,
-        help="Deprecated, this is now the default behavior",
-    )
     run_parser.add_argument(
         "--no-commit",
         action="store_true",

dvc/dvcfile.py

@@ -160,43 +160,15 @@ def validate(d, fname=None):
         except MultipleInvalid as exc:
             raise StageFileFormatError(fname, exc)
 
-    @classmethod
-    def create_stage(cls, repo, accompany_outs=False, **kwargs):
-        from dvc.stage import Stage
-
-        stage = Stage.create(repo, accompany_outs=accompany_outs, **kwargs)
-
-        ignore_build_cache = kwargs.get("ignore_build_cache", False)
-        # NOTE: remove outs before we check build cache
-        if kwargs.get("remove_outs", False):
-            logger.warning(
-                "--remove-outs is deprecated."
-                " It is now the default behavior,"
-                " so there's no need to use this option anymore."
-            )
-            stage.remove_outs(ignore_remove=False)
-            logger.warning("Build cache is ignored when using --remove-outs.")
-            ignore_build_cache = True
-
-        dvcfile = Dvcfile(stage.repo, stage.path)
-        if dvcfile.exists():
-            if any(out.persist for out in stage.outs):
-                logger.warning(
-                    "Build cache is ignored when persisting outputs."
-                )
-                ignore_build_cache = True
-
-            if not ignore_build_cache and stage.can_be_skipped:
-                logger.info("Stage is cached, skipping.")
-                return None
-
-            msg = (
-                "'{}' already exists. Do you wish to run the command and "
-                "overwrite it?".format(stage.relpath)
-            )
-            if not (kwargs.get("overwrite", True) or prompt.confirm(msg)):
-                raise StageFileAlreadyExistsError(stage.relpath)
+    def overwrite_with_prompt(self, force=False):
+        if not self.exists():
+            return
 
-            os.unlink(dvcfile.path)
+        msg = (
+            "'{}' already exists. Do you wish to run the command and "
+            "overwrite it?".format(self.path)
+        )
+        if not (force or prompt.confirm(msg)):
+            raise StageFileAlreadyExistsError(self.path)
 
-        return stage
+        os.unlink(self.path)

dvc/repo/add.py

@@ -115,6 +115,8 @@ def _find_all_targets(repo, target, recursive):
 
 
 def _create_stages(repo, targets, fname, pbar=None):
+    from dvc.stage import Stage
+
     stages = []
 
     for out in Tqdm(
@@ -123,9 +125,12 @@ def _create_stages(repo, targets, fname, pbar=None):
         disable=len(targets) < LARGE_DIR_SIZE,
         unit="file",
     ):
-        stage = Dvcfile.create_stage(
+        stage = Stage.create(
             repo, accompany_outs=True, outs=[out], fname=fname
         )
+        if stage:
+            Dvcfile(repo, stage.path).overwrite_with_prompt(force=True)
+
         repo._reset()
 
         if not stage:

dvc/repo/imp_url.py

@@ -7,9 +7,10 @@
 @scm_context
 def imp_url(self, url, out=None, fname=None, erepo=None, locked=True):
     from dvc.dvcfile import Dvcfile
+    from dvc.stage import Stage
 
     out = resolve_output(url, out)
-    stage = Dvcfile.create_stage(
+    stage = Stage.create(
         self,
         cmd=None,
         deps=[url],
@@ -22,12 +23,15 @@ def imp_url(self, url, out=None, fname=None, erepo=None, locked=True):
     if stage is None:
         return None
 
+    dvcfile = Dvcfile(self, stage.path)
+    dvcfile.overwrite_with_prompt(force=True)
+
     self.check_modified_graph([stage])
 
     stage.run()
 
     stage.locked = locked
 
-    Dvcfile(self, stage.path).dump(stage)
+    dvcfile.dump(stage)
 
     return stage

dvc/repo/run.py

@@ -1,22 +1,29 @@
+import logging
+
 from . import locked
 from .scm_context import scm_context
 
+logger = logging.getLogger(__name__)
+
 
 @locked
 @scm_context
 def run(self, no_exec=False, **kwargs):
+    from dvc.stage import Stage
     from dvc.dvcfile import Dvcfile
 
-    stage = Dvcfile.create_stage(self, **kwargs)
-
-    if stage is None:
+    stage = Stage.create(self, **kwargs)
+    if not stage:
         return None
 
+    dvcfile = Dvcfile(self, stage.path)
+    dvcfile.overwrite_with_prompt(force=kwargs.get("overwrite", True))
+
     self.check_modified_graph([stage])
 
     if not no_exec:
         stage.run(no_commit=kwargs.get("no_commit", False))
 
-    Dvcfile(self, stage.path).dump(stage)
+    dvcfile.dump(stage)
 
     return stage

dvc/stage/__init__.py

@@ -381,6 +381,21 @@ def create(repo, accompany_outs=False, **kwargs):
         stage._check_and_set_wdir(wdir, is_wdir=kwargs.get("wdir", False))
         stage._check_and_set_path(fname)
 
+        dvcfile = Dvcfile(stage.repo, stage.path)
+        if dvcfile.exists():
+            has_persist_outs = any(out.persist for out in stage.outs)
+            ignore_build_cache = (
+                kwargs.get("ignore_build_cache", False) or has_persist_outs
+            )
+            if has_persist_outs:
+                logger.warning(
+                    "Build cache is ignored when persisting outputs."
+                )
+
+            if not ignore_build_cache and stage.can_be_skipped:
+                logger.info("Stage is cached, skipping.")
+                return None
+
         return stage
 
     def _fill_stage_outputs(self, **kwargs):