fs: make _prepare_credentails/fs_args lazy (#9848)

* tests: unit: move s3 tests to dvc-s3

* deps: bump dvc-data to 2.14.0

* tests: unit: move azure tests to dvc-azure

.dvc/.gitignore

@@ -1,12 +1,3 @@
-/state
-/lock
 /config.local
-/updater
-/updater.lock
-/state-journal
-/state-wal
-/cache
-/pkg
-/repos
 /tmp
-/experiments
+/cache

.dvc/config

@@ -1,4 +0,0 @@
-['remote "dvc"']
-url = https://remote.dvc.org/dvc
-[core]
-remote = dvc

.flake8

@@ -1,19 +1,20 @@
 [flake8]
-ignore=
+max_line_length = 88
+extend-ignore=
     # Whitespace before ':'
     E203
-    # Too many leading '#' for block comment
-    E266
     # Line break occurred before a binary operator
     W503
     # unindexed parameters in the str.format, see:
     # https://pypi.org/project/flake8-string-format/
     P1
-    # use `!r` instead of surrounding by quotes
-    B028
-max_line_length = 88
 max-complexity = 10
-select = B,C,E,F,W,T4,B902,T,P,TYP
 show_source = true
 count = true
 min_python_version = 3.8.0
+per-file-ignores =
+    dvc/testing/benchmarks/conftest.py:F401
+    dvc/testing/conftest.py:F401
+    dvc/testing/plugin.py:F401
+    tests/conftest.py:F401
+    tests/func/experiments/executor/test_ssh.py:F401

.github/mergify.yml

@@ -7,3 +7,4 @@ pull_request_rules:
       backport:
         branches:
           - "2.x"
+        title: ({{ destination_branch }}) {{ title }}

.github/workflows/build.yaml

@@ -0,0 +1,47 @@
+name: Build package
+on:
+  push:
+    branches:
+    - main
+  release:
+    types: [released, prereleased]
+  workflow_dispatch:
+
+permissions:
+  contents: read
+  id-token: write
+
+jobs:
+  pip:
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@v3
+      with:
+        fetch-depth: 0
+
+    - uses: actions/setup-python@v4
+      with:
+        python-version: 3.8
+
+    - run: python -m pip install --upgrade pip setuptools_scm build twine
+    - name: Force version for Test PyPI upload
+      if: ${{ !startsWith(github.ref, 'refs/tags') }}
+      run: echo version=$(python -m setuptools_scm | awk -F+ '{print $1}' | tail -1) >> $GITHUB_ENV
+
+    - run: |
+        echo 'PKG = "pip"'>dvc/_build.py
+        python -m build
+      env:
+        SETUPTOOLS_SCM_PRETEND_VERSION: ${{ env.version }}
+    - run: twine check --strict dist/*
+
+    - name: Publish to PyPI
+      if: github.event_name == 'release'
+      uses: pypa/gh-action-pypi-publish@release/v1
+
+    - name: Publish to Test PyPI
+      if: ${{ github.event_name == 'release' || (github.event_name == 'push' && github.ref == 'refs/heads/main') }}
+      uses: pypa/gh-action-pypi-publish@release/v1
+      with:
+        repository-url: https://test.pypi.org/legacy/
+        skip-existing: true

.github/workflows/packages.yaml

@@ -1,111 +0,0 @@
-name: Build packages
-on:
-  push:
-    branches:
-    - main
-  release:
-    types: [released, prereleased]
-  workflow_dispatch:
-
-permissions:  # added using https://github.com/step-security/secure-workflows
-  contents: read
-
-jobs:
-  binary:
-    permissions:
-      contents: write  # for actions/upload-release-asset to upload release asset
-    strategy:
-      matrix:
-        include:
-         - {pkg: "deb",  os: "ubuntu-20.04", asset: "dvc_${{ github.event.release.tag_name }}_amd64.deb"}
-         - {pkg: "rpm", os: "ubuntu-20.04", asset: "dvc-${{ github.event.release.tag_name }}-1.x86_64.rpm"}
-         - {pkg: "osxpkg", os: "macos-11", asset: "dvc-${{ github.event.release.tag_name }}.pkg"}
-         - {pkg: "exe",  os: "windows-2019", asset: "dvc-${{ github.event.release.tag_name }}.exe"}
-
-    name: ${{ matrix.pkg }}
-    runs-on: ${{ matrix.os }}
-    steps:
-    - uses: actions/checkout@v3
-      with:
-          fetch-depth: 0
-
-    - name: Set up Python 3.10
-      uses: actions/setup-python@v4
-      with:
-        python-version: "3.10"
-        cache: pip
-        cache-dependency-path: |
-          pyproject.toml
-          scripts/build-requirements.txt
-
-    - name: Set up Ruby 2.6
-      uses: ruby/setup-ruby@v1
-      if: matrix.pkg != 'exe'
-      with:
-        ruby-version: '2.6'
-
-    - name: Install fpm
-      if: matrix.pkg != 'exe'
-      run: gem install --no-document fpm
-
-    - name: Install deps
-      run: |
-        pip install --upgrade pip wheel setuptools
-        pip install .[all]
-        pip install -r scripts/build-requirements.txt
-
-    - name: Pull images
-      run: dvc pull
-
-    - name: Build ${{ matrix.pkg }}
-      run: python scripts/build.py ${{ matrix.pkg }}
-
-    - name: Publish ${{ matrix.pkg }}
-      if: github.event_name == 'release'
-      uses: actions/[email protected]
-      env:
-        GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-      with:
-        upload_url: ${{ github.event.release.upload_url }}
-        asset_path: scripts/${{ matrix.pkg == 'exe' && 'innosetup' || 'fpm' }}/${{ matrix.asset }}
-        asset_name: ${{ matrix.asset }}
-        asset_content_type: binary/octet-stream
-
-  pip:
-    runs-on: ubuntu-20.04
-    permissions:
-      id-token: write
-    steps:
-    - uses: actions/checkout@v3
-      with:
-        fetch-depth: 0
-
-    - name: Set up Python 3.8
-      uses: actions/setup-python@v4
-      with:
-        python-version: 3.8
-
-    - name: Install
-      run: pip install --upgrade pip wheel
-
-    - name: Force version for Test PyPI uploads
-      if: ${{ !startsWith(github.ref, 'refs/tags') }}
-      run: |
-        pip install setuptools_scm
-        echo version=$(python -m setuptools_scm | awk -F+ '{print $1}' | tail -1) >> $GITHUB_ENV
-
-    - name: Build packages
-      run: ./scripts/build_package.sh
-      env:
-        SETUPTOOLS_SCM_PRETEND_VERSION: ${{ env.version }}
-
-    - name: Publish packages to PyPI
-      if: github.event_name == 'release'
-      uses: pypa/gh-action-pypi-publish@release/v1
-
-    - name: Publish to Test PyPI
-      if: ${{ github.event_name == 'release' || (github.event_name == 'push' && github.ref == 'refs/heads/main') }}
-      uses: pypa/gh-action-pypi-publish@release/v1
-      with:
-        repository_url: https://test.pypi.org/legacy/
-        skip_existing: true

.pre-commit-config.yaml

@@ -23,42 +23,29 @@ repos:
       - id: mixed-line-ending
       - id: sort-simple-yaml
       - id: trailing-whitespace
-  - repo: https://github.com/lovesegfault/beautysh
-    rev: v6.2.1
-    hooks:
-      - id: beautysh
-        args: ["-i", "2"]
-  - repo: https://github.com/asottile/pyupgrade
-    rev: v3.3.1
-    hooks:
-      - id: pyupgrade
-        args:
-          - --py38-plus
   - repo: https://github.com/psf/black
-    rev: 23.3.0
+    rev: 23.7.0
     hooks:
       - id: black
   - repo: https://github.com/PyCQA/isort
     rev: 5.12.0
     hooks:
       - id: isort
-  - repo: https://github.com/charliermarsh/ruff-pre-commit
-    rev: 'v0.0.261'
+  - repo: https://github.com/astral-sh/ruff-pre-commit
+    rev: 'v0.0.282'
     hooks:
       - id: ruff
+        args: [--fix, --exit-non-zero-on-fix]
   - repo: https://github.com/codespell-project/codespell
-    rev: v2.2.4
+    rev: v2.2.5
     hooks:
       - id: codespell
         additional_dependencies: ["tomli"]
   - repo: https://github.com/PyCQA/flake8
-    rev: 6.0.0
+    rev: 6.1.0
     hooks:
       - id: flake8
         additional_dependencies:
-          - flake8-bugbear
-          - flake8-comprehensions
-          - flake8-debugger
           - flake8-string-format
           - flake8-typing-imports
   - repo: https://github.com/PyCQA/bandit

README.rst

@@ -280,8 +280,8 @@ Barrak, A., Eghan, E.E. and Adams, B. `On the Co-evolution of ML Pipelines and S
    :target: https://pypi.org/project/dvc
    :alt: PyPI Downloads
 
-.. |Packages| image:: https://img.shields.io/github/v/release/iterative/dvc?label=deb|pkg|rpm|exe&logo=GitHub
-   :target: https://github.com/iterative/dvc/releases/latest
+.. |Packages| image:: https://img.shields.io/badge/deb|pkg|rpm|exe-blue
+   :target: https://dvc.org/doc/install
    :alt: deb|pkg|rpm|exe
 
 .. |DOI| image:: https://img.shields.io/badge/DOI-10.5281/zenodo.3677553-blue.svg

dvc/__init__.py

@@ -4,6 +4,7 @@ DVC
 Make your data science projects reproducible and shareable.
 """
 import dvc.logger
+from dvc.build import PKG  # noqa: F401
 from dvc.version import __version__, version_tuple  # noqa: F401
 
 dvc.logger.setup()

dvc/__pyinstaller/__init__.py

@@ -0,0 +1,5 @@
+import os
+
+
+def get_hook_dirs():
+    return [os.path.dirname(__file__)]

dvc/__pyinstaller/hook-asyncssh.py

@@ -1 +1,2 @@
+# ruff: noqa: N999
 hiddenimports = ["win32timezone"]

dvc/__pyinstaller/hook-celery.py

@@ -1,3 +1,4 @@
+# ruff: noqa: N999
 # pylint: disable=import-error
 from PyInstaller.utils.hooks import collect_submodules, is_module_or_submodule
 

dvc/__pyinstaller/hook-dvc.py

@@ -1,3 +1,4 @@
+# ruff: noqa: N999
 from PyInstaller.utils.hooks import copy_metadata  # pylint:disable=import-error
 
 # needed for `dvc doctor` to show dep versions
@@ -21,6 +22,7 @@ hiddenimports = [
     "dvc_hdfs",
     "dvc_oss",
     "dvc_s3",
+    "dvc_ssh",
     "dvc_webdav",
     "dvc_webhdfs",
     # https://github.com/pypa/setuptools/issues/1963

dvc/__pyinstaller/hook-dvc.utils.flatten.py

@@ -1,3 +1,4 @@
+# ruff: noqa: N999
 from PyInstaller.utils.hooks import copy_metadata  # pylint:disable=import-error
 
 datas = copy_metadata("flatten-dict")

dvc/__pyinstaller/hook-dvc_task.py

@@ -1,3 +1,4 @@
+# ruff: noqa: N999
 # pylint: disable=import-error
 from PyInstaller.utils.hooks import collect_submodules
 

dvc/__pyinstaller/hook-fsspec.py

@@ -1 +1,2 @@
+# ruff: noqa: N999
 hiddenimports = ["fsspec.implementations.memory"]

dvc/__pyinstaller/hook-pydrive2.py

@@ -1,3 +1,4 @@
+# ruff: noqa: N999
 from PyInstaller.utils.hooks import copy_metadata  # pylint:disable=import-error
 
 datas = copy_metadata("pydrive2")

dvc/_debug.py

@@ -198,33 +198,84 @@ def debugtools(args: Optional["Namespace"] = None, **kwargs):
 def add_debugging_flags(parser):
     from argparse import SUPPRESS
 
-    parser.add_argument("--cprofile", action="store_true", default=False, help=SUPPRESS)
-    parser.add_argument("--yappi", action="store_true", default=False, help=SUPPRESS)
+    # For detailed info see:
+    # https://github.com/iterative/dvc/wiki/Debugging,-Profiling-and-Benchmarking-DVC
+    args, _ = parser.parse_known_args()
+    verbose = args.verbose
+
+    def debug_help(msg):
+        if verbose:
+            return msg
+        return SUPPRESS
+
+    parser = parser.add_argument_group("debug options")
+
+    parser.add_argument(
+        "--cprofile",
+        action="store_true",
+        default=False,
+        help=debug_help("Generate cprofile data for tools like snakeviz / tuna"),
+    )
+    parser.add_argument(
+        "--cprofile-dump", help=debug_help("Location to dump cprofile file")
+    )
+    parser.add_argument(
+        "--yappi",
+        action="store_true",
+        default=False,
+        help=debug_help(
+            "Generate a callgrind file for use with tools like "
+            "kcachegrind / qcachegrind"
+        ),
+    )
     parser.add_argument(
         "--yappi-separate-threads",
         action="store_true",
         default=False,
-        help=SUPPRESS,
+        help=debug_help("Generate one callgrind file per thread"),
     )
     parser.add_argument(
-        "--viztracer", action="store_true", default=False, help=SUPPRESS
-    )
-    parser.add_argument("--viztracer-depth", type=int, help=SUPPRESS)
-    parser.add_argument(
-        "--viztracer-async", action="store_true", default=False, help=SUPPRESS
-    )
-    parser.add_argument("--cprofile-dump", help=SUPPRESS)
-    parser.add_argument("--pdb", action="store_true", default=False, help=SUPPRESS)
-    parser.add_argument(
-        "--instrument", action="store_true", default=False, help=SUPPRESS
+        "--viztracer",
+        action="store_true",
+        default=False,
+        help=debug_help("Generate a viztracer file for use with vizviewer"),
     )
     parser.add_argument(
-        "--instrument-open", action="store_true", default=False, help=SUPPRESS
+        "--viztracer-depth",
+        type=int,
+        help=debug_help("Set viztracer maximum stack depth"),
+    )
+    parser.add_argument(
+        "--viztracer-async",
+        action="store_true",
+        default=False,
+        help=debug_help("Treat async tasks as threads"),
+    )
+    parser.add_argument(
+        "--pdb",
+        action="store_true",
+        default=False,
+        help=debug_help("Drop into the pdb/ipdb debugger on any exception"),
+    )
+    parser.add_argument(
+        "--instrument",
+        action="store_true",
+        default=False,
+        help=debug_help("Use pyinstrument CLI profiler"),
+    )
+    parser.add_argument(
+        "--instrument-open",
+        action="store_true",
+        default=False,
+        help=debug_help("Use pyinstrument web profiler"),
     )
     parser.add_argument(
         "--show-stack",
         "--ss",
         action="store_true",
         default=False,
-        help=SUPPRESS,
+        help=debug_help(
+            r"Use Ctrl+T on macOS or Ctrl+\ on Linux to print the stack "
+            "frame currently executing. Unavailable on Windows."
+        ),
     )

dvc/annotations.py

@@ -17,12 +17,6 @@ class Annotation:
     labels: List[str] = field(default_factory=list)
     meta: Dict[str, Any] = field(default_factory=dict)
 
-    def update(self, **kwargs) -> "Annotation":
-        for attr, value in kwargs.items():
-            if value and hasattr(self, attr):
-                setattr(self, attr, value)
-        return self
-
     def to_dict(self) -> Dict[str, str]:
         return compact(asdict(self))
 

dvc/api/__init__.py

@@ -2,7 +2,7 @@ from dvc.fs.dvc import _DVCFileSystem as DVCFileSystem
 
 from .data import open  # pylint: disable=redefined-builtin
 from .data import get_url, read
-from .experiments import exp_save, exp_show, make_checkpoint
+from .experiments import exp_save, exp_show
 from .scm import all_branches, all_commits, all_tags
 from .show import metrics_show, params_show
 
@@ -13,7 +13,6 @@ __all__ = [
     "exp_save",
     "exp_show",
     "get_url",
-    "make_checkpoint",
     "open",
     "params_show",
     "metrics_show",

dvc/api/data.py

@@ -36,21 +36,22 @@ def get_url(path, repo=None, rev=None, remote=None):
     NOTE: This function does not check for the actual existence of the file or
     directory in the remote storage.
     """
-    with Repo.open(repo, rev=rev, subrepos=True, uninitialized=True) as _repo:
-        with _wrap_exceptions(_repo, path):
-            fs_path = _repo.dvcfs.from_os_path(path)
+    from dvc.config import NoRemoteError
+    from dvc_data.index import StorageKeyError
 
-            with reraise(FileNotFoundError, PathMissingError(path, repo)):
-                info = _repo.dvcfs.info(fs_path)
-
-            dvc_info = info.get("dvc_info")
-            if not dvc_info:
-                raise OutputNotFoundError(path, repo)
-
-            dvc_repo = info["repo"]  # pylint: disable=unsubscriptable-object
-            md5 = dvc_info["md5"]
-
-            return dvc_repo.cloud.get_url_for(remote, checksum=md5)
+    repo_kwargs: Dict[str, Any] = {}
+    if remote:
+        repo_kwargs["config"] = {"core": {"remote": remote}}
+    with Repo.open(
+        repo, rev=rev, subrepos=True, uninitialized=True, **repo_kwargs
+    ) as _repo:
+        index, entry = _repo.get_data_index_entry(path)
+        with reraise(
+            (StorageKeyError, ValueError),
+            NoRemoteError(f"no remote specified in {_repo}"),
+        ):
+            remote_fs, remote_path = index.storage_map.get_remote(entry)
+            return remote_fs.unstrip_protocol(remote_path)
 
 
 class _OpenContextManager(GCM):
@@ -66,13 +67,15 @@ class _OpenContextManager(GCM):
         raise AttributeError("dvc.api.open() should be used in a with statement.")
 
 
-def open(  # noqa, pylint: disable=redefined-builtin
+def open(  # noqa: A001, pylint: disable=redefined-builtin
     path: str,
     repo: Optional[str] = None,
     rev: Optional[str] = None,
     remote: Optional[str] = None,
     mode: str = "r",
     encoding: Optional[str] = None,
+    config: Optional[Dict[str, Any]] = None,
+    remote_config: Optional[Dict[str, Any]] = None,
 ):
     """
     Opens a file tracked in a DVC project.
@@ -114,6 +117,11 @@ def open(  # noqa, pylint: disable=redefined-builtin
             Defaults to None.
             This should only be used in text mode.
             Mirrors the namesake parameter in builtin `open()`_.
+        config(dict, optional): config to be passed to the DVC repository.
+            Defaults to None.
+        remote_config(dict, optional): remote config to be passed to the DVC
+            repository.
+            Defaults to None.
 
     Returns:
         _OpenContextManager: A context manager that generatse a corresponding
@@ -209,14 +217,29 @@ def open(  # noqa, pylint: disable=redefined-builtin
         "rev": rev,
         "mode": mode,
         "encoding": encoding,
+        "config": config,
+        "remote_config": remote_config,
     }
     return _OpenContextManager(_open, args, kwargs)
 
 
-def _open(path, repo=None, rev=None, remote=None, mode="r", encoding=None):
-    repo_kwargs: Dict[str, Any] = {"subrepos": True, "uninitialized": True}
-    if remote:
-        repo_kwargs["config"] = {"core": {"remote": remote}}
+def _open(
+    path,
+    repo=None,
+    rev=None,
+    remote=None,
+    mode="r",
+    encoding=None,
+    config=None,
+    remote_config=None,
+):
+    repo_kwargs: Dict[str, Any] = {
+        "subrepos": True,
+        "uninitialized": True,
+        "remote": remote,
+        "config": config,
+        "remote_config": remote_config,
+    }
 
     with Repo.open(repo, rev=rev, **repo_kwargs) as _repo:
         with _wrap_exceptions(_repo, path):
@@ -251,13 +274,29 @@ def _open(path, repo=None, rev=None, remote=None, mode="r", encoding=None):
                 raise DvcIsADirectoryError(f"'{path}' is a directory") from exc
 
 
-def read(path, repo=None, rev=None, remote=None, mode="r", encoding=None):
+def read(
+    path,
+    repo=None,
+    rev=None,
+    remote=None,
+    mode="r",
+    encoding=None,
+    config=None,
+    remote_config=None,
+):
     """
     Returns the contents of a tracked file (by DVC or Git). For Git repos, HEAD
     is used unless a rev argument is supplied. The default remote is tried
     unless a remote argument is supplied.
     """
     with open(
-        path, repo=repo, rev=rev, remote=remote, mode=mode, encoding=encoding
+        path,
+        repo=repo,
+        rev=rev,
+        remote=remote,
+        mode=mode,
+        encoding=encoding,
+        config=config,
+        remote_config=remote_config,
     ) as fd:
         return fd.read()

dvc/api/experiments.py

@@ -1,40 +1,9 @@
-import builtins
-import os
-from time import sleep
 from typing import Dict, List, Optional, Union
 
 from rich.text import Text
 
-from dvc.env import DVC_CHECKPOINT, DVC_ROOT
 from dvc.repo import Repo
 from dvc.repo.experiments.show import tabulate
-from dvc.stage.monitor import CheckpointTask
-
-
-def make_checkpoint():
-    """
-    Signal DVC to create a checkpoint experiment.
-
-    If the current process is being run from DVC, this function will block
-    until DVC has finished creating the checkpoint. Otherwise, this function
-    will return immediately.
-    """
-    if os.getenv(DVC_CHECKPOINT) is None:
-        return
-
-    root_dir = os.getenv(DVC_ROOT, Repo.find_root())
-    signal_file = os.path.join(
-        root_dir, Repo.DVC_DIR, "tmp", CheckpointTask.SIGNAL_FILE
-    )
-
-    with builtins.open(signal_file, "w", encoding="utf-8") as fobj:
-        # NOTE: force flushing/writing empty file to disk, otherwise when
-        # run in certain contexts (pytest) file may not actually be written
-        fobj.write("")
-        fobj.flush()
-        os.fsync(fobj.fileno())
-    while os.path.exists(signal_file):
-        sleep(0.1)
 
 
 def exp_save(
@@ -96,6 +65,7 @@ def exp_show(
     num: int = 1,
     param_deps: bool = False,
     force: bool = False,
+    config: Optional[Dict] = None,
 ) -> List[Dict]:
     """Get DVC experiments tracked in `repo`.
 
@@ -130,13 +100,15 @@ def exp_show(
             When `force` is specified, DVC will reload all experiment data and
             ignore any previously cached results.
             Defaults to `False`.
+        config (dict, optional): config to be passed through to DVC project.
+            Defaults to `None`.
 
     Returns:
         List[Dict]: Each item in the list will contain a dictionary with
             the info for an individual experiment.
             See Examples below.
     """
-    with Repo.open(repo) as _repo:
+    with Repo.open(repo, config=config) as _repo:
         experiments = _repo.experiments.show(
             revs=revs,
             num=num,

dvc/api/show.py

@@ -41,6 +41,7 @@ def metrics_show(
     *targets: str,
     repo: Optional[str] = None,
     rev: Optional[str] = None,
+    config: Optional[Dict] = None,
 ) -> Dict:
     """Get metrics tracked in `repo`.
 
@@ -69,6 +70,8 @@ def metrics_show(
             hash or a dvc experiment name.
             If `repo` is not a Git repo, this option is ignored.
             If `None`, the current working tree will be used.
+        config (dict, optional): config to be passed through to DVC project.
+            Defaults to `None`.
 
     Returns:
         Dict: See Examples below.
@@ -136,7 +139,7 @@ def metrics_show(
         https://git-scm.com/docs/revisions
     """
 
-    with Repo.open(repo) as _repo:
+    with Repo.open(repo, config=config) as _repo:
         metrics = _repo.metrics.show(
             targets=targets,
             revs=rev if rev is None else [rev],
@@ -157,6 +160,7 @@ def params_show(
     stages: Optional[Union[str, Iterable[str]]] = None,
     rev: Optional[str] = None,
     deps: bool = False,
+    config: Optional[Dict] = None,
 ) -> Dict:
     """Get parameters tracked in `repo`.
 
@@ -195,6 +199,8 @@ def params_show(
         deps (bool, optional): Whether to retrieve only parameters that are
             stage dependencies or not.
             Defaults to `False`.
+        config (dict, optional): config to be passed through to DVC project.
+            Defaults to `None`.
 
     Returns:
         Dict: See Examples below.
@@ -379,7 +385,7 @@ def params_show(
     if isinstance(stages, str):
         stages = [stages]
 
-    with Repo.open(repo) as _repo:
+    with Repo.open(repo, config=config) as _repo:
         params = _repo.params.show(
             revs=rev if rev is None else [rev],
             targets=targets,

dvc/build.py

@@ -0,0 +1,10 @@
+from typing import Optional
+
+try:
+    # file is created during dvc build
+    # pylint:disable=unused-import
+    from . import _build  # type: ignore[attr-defined, import]
+
+    PKG: Optional[str] = _build.PKG  # type: ignore[assignment]
+except ImportError:
+    PKG = None  # type: ignore[assignment]

dvc/cachemgr.py

@@ -1,29 +1,41 @@
 import os
+from typing import TYPE_CHECKING, Optional, Tuple
 
 from dvc.fs import GitFileSystem, Schemes
 from dvc_data.hashfile.db import get_odb
+from dvc_data.hashfile.hash import DEFAULT_ALGORITHM
+
+if TYPE_CHECKING:
+    from dvc.repo import Repo
+
+LEGACY_HASH_NAMES = {"md5-dos2unix", "params"}
 
 
-def _get_odb(repo, settings, fs=None):
+def _get_odb(
+    repo,
+    settings,
+    fs=None,
+    prefix: Optional[Tuple[str, ...]] = None,
+    hash_name: Optional[str] = None,
+    **kwargs,
+):
     from dvc.fs import get_cloud_fs
 
     if not settings:
         return None
 
-    cls, config, fs_path = get_cloud_fs(repo, **settings)
+    cls, config, fs_path = get_cloud_fs(repo.config, **settings)
     fs = fs or cls(**config)
+    if prefix:
+        fs_path = fs.path.join(fs_path, *prefix)
+    if hash_name:
+        config["hash_name"] = hash_name
     return get_odb(fs, fs_path, state=repo.state, **config)
 
 
 class CacheManager:
     CACHE_DIR = "cache"
-    CLOUD_SCHEMES = [
-        Schemes.S3,
-        Schemes.GS,
-        Schemes.SSH,
-        Schemes.HDFS,
-        Schemes.WEBHDFS,
-    ]
+    FILES_DIR = "files"
 
     def __init__(self, repo):
         self._repo = repo
@@ -53,25 +65,74 @@ class CacheManager:
         if not isinstance(repo.fs, GitFileSystem):
             kwargs["fs"] = repo.fs
 
-        odb = _get_odb(repo, settings, **kwargs)
+        odb = _get_odb(
+            repo,
+            settings,
+            prefix=(self.FILES_DIR, DEFAULT_ALGORITHM),
+            **kwargs,
+        )
         self._odb["repo"] = odb
         self._odb[Schemes.LOCAL] = odb
+        legacy_odb = _get_odb(repo, settings, hash_name="md5-dos2unix", **kwargs)
+        self._odb["legacy"] = legacy_odb
 
     def _init_odb(self, schemes):
         for scheme in schemes:
             remote = self.config.get(scheme)
             settings = {"name": remote} if remote else None
-            self._odb[scheme] = _get_odb(self._repo, settings)
+            self._odb[scheme] = _get_odb(
+                self._repo,
+                settings,
+                prefix=(self.FILES_DIR, DEFAULT_ALGORITHM),
+            )
 
     def __getattr__(self, name):
-        if name not in self._odb and name in self.CLOUD_SCHEMES:
-            self._init_odb([name])
-
         try:
             return self._odb[name]
         except KeyError as exc:
             raise AttributeError from exc
 
     def by_scheme(self):
-        self._init_odb(self.CLOUD_SCHEMES)
         yield from self._odb.items()
+
+    @property
+    def local_cache_dir(self) -> str:
+        """Return base local cache directory without any prefixes.
+
+        (i.e. `dvc cache dir`).
+        """
+        return self.legacy.path
+
+
+def migrate_2_to_3(repo: "Repo", dry: bool = False):
+    """Migrate legacy 2.x objects to 3.x cache.
+
+    Legacy 'md5-dos2unix' objects will be re-hashed with 'md5', added to 3.x cache,
+    and then a link from the legacy 2.x location to the 3.x location will be created.
+    """
+    from dvc.fs.callbacks import TqdmCallback
+    from dvc.ui import ui
+    from dvc_data.hashfile.db.migrate import migrate, prepare
+
+    src = repo.cache.legacy
+    dest = repo.cache.local
+    if dry:
+        oids = list(src._list_oids())  # pylint: disable=protected-access
+        ui.write(
+            f"{len(oids)} files will be re-hashed and migrated to the DVC 3.0 cache "
+            "location."
+        )
+        return
+
+    with TqdmCallback(
+        desc="Computing DVC 3.0 hashes",
+        unit="files",
+    ) as cb:
+        migration = prepare(src, dest, callback=cb)
+
+    with TqdmCallback(
+        desc="Migrating to DVC 3.0 cache",
+        unit="files",
+    ) as cb:
+        count = migrate(migration, callback=cb)
+    ui.write(f"Migrated {count} files to DVC 3.0 cache location.")

dvc/cli/__init__.py

@@ -75,7 +75,7 @@ def _log_exceptions(exc: Exception) -> Optional[int]:
     from dvc.fs import AuthError, ConfigError, RemoteMissingDepsError
 
     if isinstance(exc, RemoteMissingDepsError):
-        from dvc.utils.pkg import PKG
+        from dvc import PKG
 
         proto = exc.protocol
         by_pkg = {
@@ -83,9 +83,9 @@ def _log_exceptions(exc: Exception) -> Optional[int]:
             "conda": f"conda install -c conda-forge dvc-{proto}",
         }
 
-        cmd = by_pkg.get(PKG)
-        if cmd:
+        if PKG in by_pkg:
             link = format_link("https://dvc.org/doc/install")
+            cmd = by_pkg.get(PKG)
             hint = (
                 "To install dvc with those dependencies, run:\n"
                 "\n"
@@ -190,10 +190,9 @@ def main(argv=None):  # noqa: C901, PLR0912, PLR0915
         if level and level <= logging.DEBUG:
             from platform import platform, python_implementation, python_version
 
-            from dvc import __version__
-            from dvc.utils.pkg import PKG
+            from dvc import PKG, __version__
 
-            pyv = " ".join([python_implementation(), python_version()])
+            pyv = f"{python_implementation()} {python_version()}"
             pkg = f" ({PKG})" if PKG else ""
             logger.debug("v%s%s, %s on %s", __version__, pkg, pyv, platform())
             logger.debug("command: %s", " ".join(argv or sys.argv))
@@ -231,7 +230,7 @@ def main(argv=None):  # noqa: C901, PLR0912, PLR0915
         logger.exception("")
     except DvcParserError:
         ret = 254
-    except Exception as exc:  # noqa, pylint: disable=broad-except
+    except Exception as exc:  # noqa: BLE001, pylint: disable=broad-except
         # pylint: disable=no-member
         ret = _log_exceptions(exc) or 255
 

dvc/cli/actions.py

@@ -1,20 +0,0 @@
-from argparse import _AppendAction
-
-
-class CommaSeparatedArgs(_AppendAction):  # pylint: disable=protected-access
-    def __call__(self, parser, namespace, values, option_string=None):  # noqa: ARG002
-        from funcy import ldistinct
-
-        items = getattr(namespace, self.dest) or []
-        items.extend(map(str.strip, values.split(",")))
-        setattr(namespace, self.dest, ldistinct(items))
-
-
-class KeyValueArgs(_AppendAction):
-    def __call__(self, parser, namespace, values, option_string=None):  # noqa: ARG002
-        items = getattr(namespace, self.dest) or {}
-        for value in filter(bool, values):
-            key, _, value = value.partition("=")
-            items[key.strip()] = value
-
-        setattr(namespace, self.dest, items)

dvc/cli/parser.py

@@ -31,7 +31,6 @@ from dvc.commands import (
     install,
     ls,
     ls_url,
-    machine,
     metrics,
     move,
     params,
@@ -41,7 +40,6 @@ from dvc.commands import (
     remove,
     repro,
     root,
-    run,
     stage,
     unprotect,
     update,
@@ -62,7 +60,6 @@ COMMANDS = [
     remove,
     move,
     unprotect,
-    run,
     repro,
     data_sync,
     gc,
@@ -91,7 +88,6 @@ COMMANDS = [
     stage,
     experiments,
     check_ignore,
-    machine,
     data,
 ]
 
@@ -139,7 +135,6 @@ def get_parent_parser():
     from dvc._debug import add_debugging_flags
 
     parent_parser = argparse.ArgumentParser(add_help=False)
-    add_debugging_flags(parent_parser)
     log_level_group = parent_parser.add_mutually_exclusive_group()
     log_level_group.add_argument(
         "-q", "--quiet", action="count", default=0, help="Be quiet."
@@ -147,6 +142,7 @@ def get_parent_parser():
     log_level_group.add_argument(
         "-v", "--verbose", action="count", default=0, help="Be verbose."
     )
+    add_debugging_flags(parent_parser)
 
     return parent_parser
 

dvc/cli/utils.py

@@ -1,3 +1,27 @@
+import argparse
+
+
+class DictAction(argparse.Action):
+    def __call__(self, parser, args, values, option_string=None):  # noqa: ARG002
+        d = getattr(args, self.dest) or {}
+
+        if isinstance(values, list):
+            kvs = values
+        else:
+            kvs = [values]
+
+        for kv in kvs:
+            key, value = kv.split("=")
+            if not value:
+                raise argparse.ArgumentError(
+                    self,
+                    f'Could not parse argument "{values}" as k1=v1 k2=v2 ... format',
+                )
+            d[key] = value
+
+        setattr(args, self.dest, d)
+
+
 def fix_subparsers(subparsers):
     """Workaround for bug in Python 3. See more info at:
     https://bugs.python.org/issue16308
@@ -16,9 +40,7 @@ def append_doc_link(help_message, path):
     if not path:
         return help_message
     doc_base = "https://man.dvc.org/"
-    return "{message}\nDocumentation: {link}".format(
-        message=help_message, link=format_link(doc_base + path)
-    )
+    return f"{help_message}\nDocumentation: {format_link(doc_base + path)}"
 
 
 def hide_subparsers_from_help(subparsers):

dvc/commands/add.py

@@ -2,66 +2,56 @@ import argparse
 import logging
 
 from dvc.cli import completion
-from dvc.cli.actions import KeyValueArgs
 from dvc.cli.command import CmdBase
 from dvc.cli.utils import append_doc_link
 
 logger = logging.getLogger(__name__)
 
 
-def _add_annotating_args(parser: argparse.ArgumentParser) -> None:
-    parser.add_argument(
-        "--desc",
-        type=str,
-        metavar="<text>",
-        help="User description of the data.",
-    )
-    parser.add_argument(
-        "--meta",
-        metavar="key=value",
-        nargs=1,
-        action=KeyValueArgs,
-        help="Custom metadata to add to the data.",
-    )
-    parser.add_argument(
-        "--label",
-        dest="labels",
-        type=str,
-        action="append",
-        metavar="<str>",
-        help="Label for the data.",
-    )
-    parser.add_argument(
-        "--type",
-        type=str,
-        metavar="<str>",
-        help="Type of the data.",
-    )
-
-
 class CmdAdd(CmdBase):
+    def validate_args(self) -> None:
+        from dvc.exceptions import InvalidArgumentError
+
+        args = self.args
+        invalid_opt = None
+
+        if args.to_remote or args.out:
+            message = "{option} can't be used with "
+            message += "--to-remote" if args.to_remote else "--out"
+            if len(args.targets) != 1:
+                invalid_opt = "multiple targets"
+            elif args.glob:
+                invalid_opt = "--glob option"
+            elif args.no_commit:
+                invalid_opt = "--no-commit option"
+        else:
+            message = "{option} can't be used without --to-remote"
+            if args.remote:
+                invalid_opt = "--remote"
+            elif args.remote_jobs:
+                invalid_opt = "--remote-jobs"
+
+        if invalid_opt is not None:
+            raise InvalidArgumentError(message.format(option=invalid_opt))
+
     def run(self):
-        from dvc.exceptions import DvcException, RecursiveAddingWhileUsingFilename
+        from dvc.exceptions import DvcException, InvalidArgumentError
 
         try:
-            if len(self.args.targets) > 1 and self.args.file:
-                raise RecursiveAddingWhileUsingFilename()
+            self.validate_args()
+        except InvalidArgumentError:
+            logger.exception("")
+            return 1
 
+        try:
             self.repo.add(
                 self.args.targets,
-                recursive=self.args.recursive,
                 no_commit=self.args.no_commit,
-                fname=self.args.file,
-                external=self.args.external,
                 glob=self.args.glob,
-                desc=self.args.desc,
                 out=self.args.out,
-                type=self.args.type,
-                labels=self.args.labels,
-                meta=self.args.meta,
                 remote=self.args.remote,
                 to_remote=self.args.to_remote,
-                jobs=self.args.jobs,
+                remote_jobs=self.args.remote_jobs,
                 force=self.args.force,
             )
         except FileNotFoundError:
@@ -83,36 +73,18 @@ def add_parser(subparsers, parent_parser):
         help=ADD_HELP,
         formatter_class=argparse.RawDescriptionHelpFormatter,
     )
-    parser.add_argument(
-        "-R",
-        "--recursive",
-        action="store_true",
-        default=False,
-        help="Recursively add files under directory targets.",
-    )
     parser.add_argument(
         "--no-commit",
         action="store_true",
         default=False,
         help="Don't put files/directories into cache.",
     )
-    parser.add_argument(
-        "--external",
-        action="store_true",
-        default=False,
-        help="Allow targets that are outside of the DVC repository.",
-    )
     parser.add_argument(
         "--glob",
         action="store_true",
         default=False,
         help="Allows targets containing shell-style wildcards.",
     )
-    parser.add_argument(
-        "--file",
-        help="Specify name of the .dvc file this command will generate.",
-        metavar="<filename>",
-    )
     parser.add_argument(
         "-o",
         "--out",
@@ -132,8 +104,7 @@ def add_parser(subparsers, parent_parser):
         metavar="<name>",
     )
     parser.add_argument(
-        "-j",
-        "--jobs",
+        "--remote-jobs",
         type=int,
         help=(
             "Only used along with '--to-remote'. "
@@ -150,8 +121,6 @@ def add_parser(subparsers, parent_parser):
         default=False,
         help="Override local file or folder if exists.",
     )
-
-    _add_annotating_args(parser)
     parser.add_argument(
         "targets", nargs="+", help="Input files/directories to add."
     ).complete = completion.FILE

dvc/commands/cache.py

@@ -2,6 +2,7 @@ import argparse
 import os
 
 from dvc.cli import completion
+from dvc.cli.command import CmdBase
 from dvc.cli.utils import append_doc_link, fix_subparsers
 from dvc.commands.config import CmdConfig
 from dvc.ui import ui
@@ -36,6 +37,14 @@ class CmdCacheDir(CmdConfig):
         return 0
 
 
+class CmdCacheMigrate(CmdBase):
+    def run(self):
+        from dvc.cachemgr import migrate_2_to_3
+
+        migrate_2_to_3(self.repo, dry=self.args.dry)
+        return 0
+
+
 def add_parser(subparsers, parent_parser):
     from dvc.commands.config import parent_config_parser
 
@@ -54,8 +63,6 @@ def add_parser(subparsers, parent_parser):
         help="Use `dvc cache CMD --help` for command-specific help.",
     )
 
-    fix_subparsers(cache_subparsers)
-
     parent_cache_config_parser = argparse.ArgumentParser(
         add_help=False, parents=[parent_config_parser]
     )
@@ -86,3 +93,23 @@ def add_parser(subparsers, parent_parser):
         nargs="?",
     ).complete = completion.DIR
     cache_dir_parser.set_defaults(func=CmdCacheDir)
+
+    CACHE_MIGRATE_HELP = "Migrate cached files to the DVC 3.0 cache location."
+    cache_migrate_parser = cache_subparsers.add_parser(
+        "migrate",
+        parents=[parent_parser],
+        description=append_doc_link(CACHE_HELP, "cache/migrate"),
+        help=CACHE_MIGRATE_HELP,
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+    )
+    cache_migrate_parser.add_argument(
+        "--dry",
+        help=(
+            "Only print actions which would be taken without actually migrating "
+            "any data."
+        ),
+        action="store_true",
+    )
+    cache_migrate_parser.set_defaults(func=CmdCacheMigrate)
+
+    fix_subparsers(cache_subparsers)

dvc/commands/commit.py

@@ -47,7 +47,10 @@ def add_parser(subparsers, parent_parser):
         "--force",
         action="store_true",
         default=False,
-        help="Commit even if hash value for dependencies/outputs changed.",
+        help=(
+            "Commit data even if hash values for dependencies or "
+            "outputs did not change."
+        ),
     )
     commit_parser.add_argument(
         "-d",

dvc/commands/completion.py

@@ -9,6 +9,9 @@ from dvc.ui import ui
 logger = logging.getLogger(__name__)
 
 
+SUPPORTED_SHELLS = ["bash", "zsh"]
+
+
 class CmdCompletion(CmdBaseNoRepo):
     def run(self):
         import shtab
@@ -35,6 +38,6 @@ def add_parser(subparsers, parent_parser):
         "--shell",
         help="Shell syntax for completions.",
         default="bash",
-        choices=["bash", "zsh"],
+        choices=SUPPORTED_SHELLS,
     )
     completion_parser.set_defaults(func=CmdCompletion)

dvc/commands/dag.py

@@ -176,7 +176,6 @@ def add_parser(subparsers, parent_parser):
     )
     dag_parser.add_argument(
         "--md",
-        "--show-md",
         action="store_true",
         default=False,
         dest="markdown",

dvc/commands/data.py

@@ -1,14 +1,11 @@
 import argparse
 import logging
-from operator import itemgetter
-from typing import TYPE_CHECKING, Any, Dict, Iterable, Set
+from typing import TYPE_CHECKING
 
 from funcy import chunks, compact, log_durations
 
-from dvc.cli import completion
-from dvc.cli.actions import CommaSeparatedArgs
 from dvc.cli.command import CmdBase
-from dvc.cli.utils import append_doc_link, fix_subparsers, hide_subparsers_from_help
+from dvc.cli.utils import append_doc_link, fix_subparsers
 from dvc.ui import ui
 from dvc.utils import colorize
 
@@ -93,9 +90,7 @@ class CmdDataStatus(CmdBase):
                     ui.write(f"  ({hint})")
 
             if isinstance(stage_status, dict):
-                items = [
-                    ": ".join([state, file]) for file, state in stage_status.items()
-                ]
+                items = [f"{state}: {file}" for file, state in stage_status.items()]
             else:
                 items = stage_status
 
@@ -132,63 +127,6 @@ class CmdDataStatus(CmdBase):
         return self._show_status(status)
 
 
-class CmdDataLs(CmdBase):
-    @staticmethod
-    def _show_table(
-        d: Iterable[Dict[str, Any]],
-        filter_types: Set[str],
-        filter_labels: Set[str],
-        markdown: bool = False,
-    ) -> None:
-        from rich.style import Style
-
-        from dvc.compare import TabularData
-
-        td = TabularData(
-            columns=["Path", "Type", "Labels", "Description"], fill_value="-"
-        )
-        for entry in sorted(d, key=itemgetter("path")):
-            typ = entry.get("type", "")
-            desc = entry.get("desc", "-")
-            labels = entry.get("labels", [])
-
-            if filter_types and typ not in filter_types:
-                continue
-            if filter_labels and filter_labels.isdisjoint(labels):
-                continue
-
-            rich_label = ui.rich_text()
-            for index, label in enumerate(labels):
-                if index:
-                    rich_label.append(",")
-                style = Style(bold=label in filter_labels, color="green")
-                rich_label.append(label, style=style)
-
-            if markdown and desc:
-                desc = desc.partition("\n")[0]
-
-            path = ui.rich_text(entry["path"], style="cyan")
-            type_style = Style(bold=typ in filter_types, color="yellow")
-            typ = ui.rich_text(entry.get("type", ""), style=type_style)
-            td.append([path, typ or "-", rich_label or "-", desc])
-
-        td.render(markdown=markdown, rich_table=True)
-
-    def run(self):
-        from dvc.repo.data import ls
-
-        filter_labels = set(self.args.labels)
-        filter_types = set(self.args.type)
-        d = ls(self.repo, targets=self.args.targets, recursive=self.args.recursive)
-        self._show_table(
-            d,
-            filter_labels=filter_labels,
-            filter_types=filter_types,
-            markdown=self.args.markdown,
-        )
-        return 0
-
-
 def add_parser(subparsers, parent_parser):
     data_parser = subparsers.add_parser(
         "data",
@@ -217,13 +155,6 @@ def add_parser(subparsers, parent_parser):
         default=False,
         help="Show output in JSON format.",
     )
-    data_status_parser.add_argument(
-        "--show-json",
-        action="store_true",
-        default=False,
-        dest="json",
-        help=argparse.SUPPRESS,
-    )
     data_status_parser.add_argument(
         "--granular",
         action="store_true",
@@ -257,51 +188,3 @@ def add_parser(subparsers, parent_parser):
         help="Use cached remote index (don't check remote).",
     )
     data_status_parser.set_defaults(func=CmdDataStatus)
-
-    DATA_LS_HELP = "List data tracked by DVC with its metadata."
-    data_ls_parser = data_subparsers.add_parser(
-        "ls",
-        aliases=["list"],
-        parents=[parent_parser],
-        description=append_doc_link(DATA_LS_HELP, "data/ls"),
-        formatter_class=argparse.RawDescriptionHelpFormatter,
-        add_help=False,
-    )
-    data_ls_parser.add_argument(
-        "--md",
-        "--show-md",
-        dest="markdown",
-        action="store_true",
-        default=False,
-        help="Show tabulated output in the Markdown format (GFM).",
-    )
-    data_ls_parser.add_argument(
-        "--type",
-        action=CommaSeparatedArgs,
-        default=[],
-        help="Comma-separated list of type to filter.",
-    )
-    data_ls_parser.add_argument(
-        "--labels",
-        action=CommaSeparatedArgs,
-        default=[],
-        help="Comma-separated list of labels to filter.",
-    )
-    data_ls_parser.add_argument(
-        "-R",
-        "--recursive",
-        action="store_true",
-        default=False,
-        help="Recursively list from the specified directories.",
-    )
-    data_ls_parser.add_argument(
-        "targets",
-        default=None,
-        nargs="*",
-        help=(
-            "Limit command scope to these tracked files/directories, "
-            ".dvc files, or stage names."
-        ),
-    ).complete = completion.DVCFILES_AND_STAGE
-    data_ls_parser.set_defaults(func=CmdDataLs)
-    hide_subparsers_from_help(data_subparsers)

dvc/commands/data_sync.py

@@ -91,6 +91,8 @@ class CmdDataFetch(CmdDataBase):
                 with_deps=self.args.with_deps,
                 recursive=self.args.recursive,
                 run_cache=self.args.run_cache,
+                max_size=self.args.max_size,
+                types=self.args.types,
             )
             self.log_summary({"fetched": processed_files_count})
         except DvcException:
@@ -322,6 +324,22 @@ def add_parser(subparsers, _parent_parser):
         default=False,
         help="Fetch run history for all stages.",
     )
+    fetch_parser.add_argument(
+        "--max-size",
+        type=int,
+        help="Fetch data files/directories that are each below specified size (bytes).",
+    )
+    fetch_parser.add_argument(
+        "--type",
+        dest="types",
+        action="append",
+        default=[],
+        help=(
+            "Only fetch data files/directories that are of a particular "
+            "type (metrics, plots)."
+        ),
+        choices=["metrics", "plots"],
+    )
     fetch_parser.set_defaults(func=CmdDataFetch)
 
     # Status
@@ -403,7 +421,6 @@ def add_parser(subparsers, _parent_parser):
     )
     status_parser.add_argument(
         "--json",
-        "--show-json",
         action="store_true",
         default=False,
         help="Show status in JSON format.",

dvc/commands/destroy.py

@@ -21,7 +21,7 @@ class CmdDestroy(CmdBase):
             )
 
             if not self.args.force and not ui.confirm(statement):
-                raise DvcException(
+                raise DvcException(  # noqa: TRY301
                     "cannot destroy without a confirmation from the user."
                     " Use `-f` to force."
                 )

dvc/commands/diff.py

@@ -184,7 +184,6 @@ def add_parser(subparsers, parent_parser):
     )
     diff_parser.add_argument(
         "--json",
-        "--show-json",
         help="Format the output into a JSON",
         action="store_true",
         default=False,
@@ -197,7 +196,6 @@ def add_parser(subparsers, parent_parser):
     )
     diff_parser.add_argument(
         "--md",
-        "--show-md",
         help="Show tabulated output in the Markdown format (GFM).",
         action="store_true",
         dest="markdown",

dvc/commands/experiments/__init__.py

@@ -7,8 +7,6 @@ from dvc.commands.experiments import (
     clean,
     diff,
     exec_run,
-    gc,
-    init,
     ls,
     pull,
     push,
@@ -25,8 +23,6 @@ SUB_COMMANDS = [
     clean,
     diff,
     exec_run,
-    gc,
-    init,
     ls,
     pull,
     push,

dvc/commands/experiments/diff.py

@@ -83,14 +83,12 @@ def add_parser(experiments_subparsers, parent_parser):
     )
     experiments_diff_parser.add_argument(
         "--json",
-        "--show-json",
         action="store_true",
         default=False,
         help="Show output in JSON format.",
     )
     experiments_diff_parser.add_argument(
         "--md",
-        "--show-md",
         action="store_true",
         default=False,
         dest="markdown",

dvc/commands/experiments/gc.py

@@ -1,146 +0,0 @@
-import argparse
-import logging
-
-from dvc.cli.command import CmdBase
-from dvc.cli.utils import append_doc_link
-from dvc.exceptions import InvalidArgumentError
-from dvc.ui import ui
-
-logger = logging.getLogger(__name__)
-
-
-def _raise_error_if_all_disabled(**kwargs):
-    if not any(kwargs.values()):
-        raise InvalidArgumentError(
-            "Either of `-w|--workspace`, `-a|--all-branches`, `-T|--all-tags`,"
-            " `--all-commits` or `--date` needs to be set."
-        )
-
-
-class CmdExperimentsGC(CmdBase):
-    def run(self):
-        _raise_error_if_all_disabled(
-            all_branches=self.args.all_branches,
-            all_tags=self.args.all_tags,
-            all_commits=self.args.all_commits,
-            commit_date=self.args.commit_date,
-            workspace=self.args.workspace,
-        )
-
-        msg = "This will remove all experiments except those derived from "
-
-        msg += "the workspace"
-        if self.args.all_commits:
-            msg += " and all git commits"
-        else:
-            if self.args.all_branches and self.args.all_tags:
-                msg += " and all git branches and tags"
-            elif self.args.all_branches:
-                msg += " and all git branches"
-            elif self.args.all_tags:
-                msg += " and all git tags"
-
-            if self.args.commit_date:
-                msg += f" and all git commits before date {self.args.commit_date}"
-        msg += " of the current repo."
-
-        if self.args.queued:
-            msg += " Run queued experiments will be preserved."
-        else:
-            msg += " Run queued experiments will be removed."
-
-        logger.warning(msg)
-
-        msg = "Are you sure you want to proceed?"
-        if not self.args.force and not ui.confirm(msg):
-            return 1
-
-        removed = self.repo.experiments.gc(
-            all_branches=self.args.all_branches,
-            all_tags=self.args.all_tags,
-            all_commits=self.args.all_commits,
-            workspace=self.args.workspace,
-            commit_date=self.args.commit_date,
-            queued=self.args.queued,
-        )
-
-        if removed:
-            ui.write(
-                f"Removed {removed} experiments.",
-                "To remove unused cache files",
-                "use 'dvc gc'.",
-            )
-        else:
-            ui.write("No experiments to remove.")
-        return 0
-
-
-def add_parser(experiments_subparsers, parent_parser):
-    EXPERIMENTS_GC_HELP = "Garbage collect unneeded experiments."
-    EXPERIMENTS_GC_DESCRIPTION = (
-        "Removes all experiments which are not derived"
-        " from the specified Git revisions."
-    )
-    experiments_gc_parser = experiments_subparsers.add_parser(
-        "gc",
-        parents=[parent_parser],
-        description=append_doc_link(EXPERIMENTS_GC_DESCRIPTION, "exp/gc"),
-        help=EXPERIMENTS_GC_HELP,
-        formatter_class=argparse.RawDescriptionHelpFormatter,
-    )
-    experiments_gc_parser.add_argument(
-        "-w",
-        "--workspace",
-        action="store_true",
-        default=False,
-        help="Keep experiments derived from the current workspace.",
-    )
-    experiments_gc_parser.add_argument(
-        "-a",
-        "--all-branches",
-        action="store_true",
-        default=False,
-        help="Keep experiments derived from the tips of all Git branches.",
-    )
-    experiments_gc_parser.add_argument(
-        "-T",
-        "--all-tags",
-        action="store_true",
-        default=False,
-        help="Keep experiments derived from all Git tags.",
-    )
-    experiments_gc_parser.add_argument(
-        "-A",
-        "--all-commits",
-        action="store_true",
-        default=False,
-        help="Keep experiments derived from all Git commits.",
-    )
-    experiments_gc_parser.add_argument(
-        "--date",
-        type=str,
-        dest="commit_date",
-        metavar="<YYYY-MM-DD>",
-        default=None,
-        help=(
-            "Keep experiments from the commits after (inclusive) a certain "
-            "date. Date must match the extended ISO 8601 format (YYYY-MM-DD)."
-        ),
-    )
-    experiments_gc_parser.add_argument(
-        "--queued",
-        action="store_true",
-        default=False,
-        help=(
-            "Keep queued experiments (experiments run queue will be cleared "
-            "by default)."
-        ),
-    )
-    experiments_gc_parser.add_argument(
-        "-f",
-        "--force",
-        action="store_true",
-        default=False,
-        help="Force garbage collection - automatically agree to all prompts.",
-    )
-    experiments_gc_parser.set_defaults(func=CmdExperimentsGC)

dvc/commands/experiments/init.py

@@ -1,219 +0,0 @@
-import argparse
-import logging
-from typing import TYPE_CHECKING, List
-
-from funcy import compact
-
-from dvc.cli.command import CmdBase
-from dvc.cli.utils import append_doc_link
-from dvc.exceptions import InvalidArgumentError
-from dvc.ui import ui
-
-if TYPE_CHECKING:
-    from dvc.dependency import Dependency
-    from dvc.stage import PipelineStage
-
-
-logger = logging.getLogger(__name__)
-
-
-class CmdExperimentsInit(CmdBase):
-    DEFAULT_NAME = "train"
-    CODE = "src"
-    DATA = "data"
-    MODELS = "models"
-    DEFAULT_METRICS = "metrics.json"
-    DEFAULT_PARAMS = "params.yaml"
-    PLOTS = "plots"
-    DEFAULTS = {
-        "code": CODE,
-        "data": DATA,
-        "models": MODELS,
-        "metrics": DEFAULT_METRICS,
-        "params": DEFAULT_PARAMS,
-        "plots": PLOTS,
-    }
-
-    def run(self):
-        from dvc.commands.stage import parse_cmd
-
-        cmd = parse_cmd(self.args.command)
-        if not self.args.interactive and not cmd:
-            raise InvalidArgumentError("command is not specified")
-
-        from dvc.repo.experiments.init import init
-
-        defaults = {}
-        if not self.args.explicit:
-            config = self.repo.config["exp"]
-            defaults.update({**self.DEFAULTS, **config})
-
-        cli_args = compact(
-            {
-                "cmd": cmd,
-                "code": self.args.code,
-                "data": self.args.data,
-                "models": self.args.models,
-                "metrics": self.args.metrics,
-                "params": self.args.params,
-                "plots": self.args.plots,
-                "live": self.args.live,
-            }
-        )
-
-        initialized_stage, initialized_deps, initialized_out_dirs = init(
-            self.repo,
-            name=self.args.name,
-            type=self.args.type,
-            defaults=defaults,
-            overrides=cli_args,
-            interactive=self.args.interactive,
-            force=self.args.force,
-        )
-        self._post_init_display(
-            initialized_stage, initialized_deps, initialized_out_dirs
-        )
-        if self.args.run:
-            self.repo.experiments.run(targets=[initialized_stage.addressing])
-        return 0
-
-    def _post_init_display(
-        self,
-        stage: "PipelineStage",
-        new_deps: List["Dependency"],
-        new_out_dirs: List[str],
-    ) -> None:
-        from dvc.utils import humanize
-
-        path_fmt = "[green]{}[/green]".format
-        if new_deps:
-            deps_paths = humanize.join(map(path_fmt, new_deps))
-            ui.write(f"Creating dependencies: {deps_paths}", styled=True)
-
-        if new_out_dirs:
-            out_dirs_paths = humanize.join(map(path_fmt, new_out_dirs))
-            ui.write(f"Creating output directories: {out_dirs_paths}", styled=True)
-
-        ui.write(
-            f"Creating [b]{self.args.name}[/b] stage in [green]dvc.yaml[/]",
-            styled=True,
-        )
-        if stage.outs or not self.args.run:
-            # separate the above status-like messages with help/tips section
-            ui.write(styled=True)
-
-        if stage.outs:
-            outs_paths = humanize.join(map(path_fmt, stage.outs))
-            tips = f"Ensure your experiment command creates {outs_paths}."
-            ui.write(tips, styled=True)
-
-        if not self.args.run:
-            ui.write(
-                'You can now run your experiment using [b]"dvc exp run"[/].',
-                styled=True,
-            )
-        else:
-            # separate between `exp.run` output and `dvc exp init` output
-            ui.write(styled=True)
-
-
-def add_parser(experiments_subparsers, parent_parser):
-    EXPERIMENTS_INIT_HELP = "Quickly setup any project to use experiments."
-    experiments_init_parser = experiments_subparsers.add_parser(
-        "init",
-        parents=[parent_parser],
-        description=append_doc_link(EXPERIMENTS_INIT_HELP, "exp/init"),
-        formatter_class=argparse.RawDescriptionHelpFormatter,
-        help=EXPERIMENTS_INIT_HELP,
-    )
-    experiments_init_parser.add_argument(
-        "command",
-        nargs=argparse.REMAINDER,
-        help="Command to execute.",
-        metavar="command",
-    )
-    experiments_init_parser.add_argument(
-        "--run",
-        action="store_true",
-        help="Run the experiment after initializing it",
-    )
-    experiments_init_parser.add_argument(
-        "--interactive",
-        "-i",
-        action="store_true",
-        help="Prompt for values that are not provided",
-    )
-    experiments_init_parser.add_argument(
-        "-f",
-        "--force",
-        action="store_true",
-        default=False,
-        help="Overwrite existing stage",
-    )
-    experiments_init_parser.add_argument(
-        "--explicit",
-        action="store_true",
-        default=False,
-        help="Only use the path values explicitly provided",
-    )
-    experiments_init_parser.add_argument(
-        "--name",
-        "-n",
-        help="Name of the stage to create (default: %(default)s)",
-        default=CmdExperimentsInit.DEFAULT_NAME,
-    )
-    experiments_init_parser.add_argument(
-        "--code",
-        help=(
-            "Path to the source file or directory "
-            "which your experiments depend"
-            f" (default: {CmdExperimentsInit.CODE})"
-        ),
-    )
-    experiments_init_parser.add_argument(
-        "--data",
-        help=(
-            "Path to the data file or directory "
-            "which your experiments depend"
-            f" (default: {CmdExperimentsInit.DATA})"
-        ),
-    )
-    experiments_init_parser.add_argument(
-        "--models",
-        help=(
-            "Path to the model file or directory for your experiments"
-            f" (default: {CmdExperimentsInit.MODELS})"
-        ),
-    )
-    experiments_init_parser.add_argument(
-        "--params",
-        help=(
-            "Path to the parameters file for your experiments"
-            f" (default: {CmdExperimentsInit.DEFAULT_PARAMS})"
-        ),
-    )
-    experiments_init_parser.add_argument(
-        "--metrics",
-        help=(
-            "Path to the metrics file for your experiments"
-            f" (default: {CmdExperimentsInit.DEFAULT_METRICS})"
-        ),
-    )
-    experiments_init_parser.add_argument(
-        "--plots",
-        help=(
-            "Path to the plots file or directory for your experiments"
-            f" (default: {CmdExperimentsInit.PLOTS})"
-        ),
-    )
-    experiments_init_parser.add_argument(
-        "--live",
-        help="Path to log dvclive outputs for your experiments",
-    )
-    experiments_init_parser.add_argument(
-        "--type",
-        choices=["default", "checkpoint"],
-        default="default",
-        help="Select type of stage to create (default: %(default)s)",
-    )
-    experiments_init_parser.set_defaults(func=CmdExperimentsInit)

dvc/commands/experiments/ls.py

@@ -3,6 +3,7 @@ import logging
 
 from dvc.cli.command import CmdBase
 from dvc.cli.utils import append_doc_link
+from dvc.exceptions import InvalidArgumentError
 from dvc.ui import ui
 
 logger = logging.getLogger(__name__)
@@ -11,19 +12,43 @@ logger = logging.getLogger(__name__)
 class CmdExperimentsList(CmdBase):
     def run(self):
         name_only = self.args.name_only
+        sha_only = self.args.sha_only
+        git_remote = self.args.git_remote
+        if sha_only and git_remote:
+            raise InvalidArgumentError("--sha-only not supported with git_remote.")
         exps = self.repo.experiments.ls(
             all_commits=self.args.all_commits,
             rev=self.args.rev,
             num=self.args.num,
-            git_remote=self.args.git_remote,
+            git_remote=git_remote,
         )
 
+        from dvc.repo.experiments.utils import describe
+        from dvc.scm import Git
+
+        if name_only or sha_only:
+            names = {}
+        else:
+            assert isinstance(self.repo.scm, Git)
+            names = describe(
+                self.repo.scm,
+                (baseline for baseline in exps),
+                logger=logger,
+            )
+
         for baseline in exps:
-            if not name_only:
-                ui.write(f"{baseline}:")
-            for exp_name in exps[baseline]:
-                indent = "" if name_only else "\t"
-                ui.write(f"{indent}{exp_name}")
+            if not (name_only or sha_only):
+                name = names.get(baseline) or baseline[:7]
+                ui.write(f"{name}:")
+            for exp_name, rev in exps[baseline]:
+                if name_only:
+                    ui.write(exp_name)
+                elif sha_only:
+                    ui.write(rev)
+                elif rev:
+                    ui.write(f"\t{rev[:7]} [{exp_name}]")
+                else:
+                    ui.write(f"\t{exp_name}")
 
         return 0
 
@@ -40,11 +65,18 @@ def add_parser(experiments_subparsers, parent_parser):
         formatter_class=argparse.RawDescriptionHelpFormatter,
     )
     add_rev_selection_flags(experiments_list_parser, "List")
-    experiments_list_parser.add_argument(
+    display_group = experiments_list_parser.add_mutually_exclusive_group()
+    display_group.add_argument(
         "--name-only",
         "--names-only",
         action="store_true",
-        help="Only output experiment names (without parent commits).",
+        help="Only output experiment names (without SHAs or parent commits).",
+    )
+    display_group.add_argument(
+        "--sha-only",
+        "--shas-only",
+        action="store_true",
+        help="Only output experiment commit SHAs (without names or parent commits).",
     )
     experiments_list_parser.add_argument(
         "git_remote",

dvc/commands/experiments/pull.py

@@ -3,23 +3,13 @@ import logging
 
 from dvc.cli.command import CmdBase
 from dvc.cli.utils import append_doc_link
-from dvc.exceptions import InvalidArgumentError
 from dvc.ui import ui
 
 logger = logging.getLogger(__name__)
 
 
 class CmdExperimentsPull(CmdBase):
-    def raise_error_if_all_disabled(self):
-        if not any([self.args.experiment, self.args.all_commits, self.args.rev]):
-            raise InvalidArgumentError(
-                "Either provide an `experiment` argument, or use the "
-                "`--rev` or `--all-commits` flag."
-            )
-
     def run(self):
-        self.raise_error_if_all_disabled()
-
         pulled_exps = self.repo.experiments.pull(
             self.args.git_remote,
             self.args.experiment,
@@ -61,7 +51,7 @@ def add_parser(experiments_subparsers, parent_parser):
         help=EXPERIMENTS_PULL_HELP,
         formatter_class=argparse.RawDescriptionHelpFormatter,
     )
-    add_rev_selection_flags(experiments_pull_parser, "Pull", False)
+    add_rev_selection_flags(experiments_pull_parser, "Pull", True)
     experiments_pull_parser.add_argument(
         "-f",
         "--force",

dvc/commands/experiments/push.py

@@ -5,20 +5,12 @@ from typing import Any, Dict
 from dvc.cli import completion
 from dvc.cli.command import CmdBase
 from dvc.cli.utils import append_doc_link
-from dvc.exceptions import InvalidArgumentError
 from dvc.ui import ui
 
 logger = logging.getLogger(__name__)
 
 
 class CmdExperimentsPush(CmdBase):
-    def raise_error_if_all_disabled(self):
-        if not any([self.args.experiment, self.args.all_commits, self.args.rev]):
-            raise InvalidArgumentError(
-                "Either provide an `experiment` argument, or use the "
-                "`--rev` or `--all-commits` flag."
-            )
-
     @staticmethod
     def log_result(result: Dict[str, Any], remote: str):
         from dvc.utils import humanize
@@ -59,8 +51,6 @@ class CmdExperimentsPush(CmdBase):
     def run(self):
         from dvc.repo.experiments.push import UploadError
 
-        self.raise_error_if_all_disabled()
-
         try:
             result = self.repo.experiments.push(
                 self.args.git_remote,
@@ -102,7 +92,7 @@ def add_parser(experiments_subparsers, parent_parser):
         help=EXPERIMENTS_PUSH_HELP,
         formatter_class=argparse.RawDescriptionHelpFormatter,
     )
-    add_rev_selection_flags(experiments_push_parser, "Push", False)
+    add_rev_selection_flags(experiments_push_parser, "Push", True)
     experiments_push_parser.add_argument(
         "-f",
         "--force",

dvc/commands/experiments/run.py

@@ -1,51 +1,27 @@
 import argparse
 import logging
 
-from dvc.cli import completion
 from dvc.cli.utils import append_doc_link
 from dvc.commands.repro import CmdRepro
 from dvc.commands.repro import add_arguments as add_repro_arguments
-from dvc.exceptions import InvalidArgumentError
-from dvc.ui import ui
 
 logger = logging.getLogger(__name__)
 
 
 class CmdExperimentsRun(CmdRepro):
     def run(self):
-        from dvc.compare import show_metrics
-
-        if self.args.checkpoint_resume:
-            if self.args.reset:
-                raise InvalidArgumentError("--reset and --rev are mutually exclusive.")
-            if not (self.args.queue or self.args.tmp_dir):
-                raise InvalidArgumentError(
-                    "--rev can only be used in conjunction with --queue or --temp."
-                )
-
-        if self.args.reset:
-            ui.write("Any existing checkpoints will be reset and re-run.")
-
-        results = self.repo.experiments.run(
+        self.repo.experiments.run(
             name=self.args.name,
             queue=self.args.queue,
             run_all=self.args.run_all,
             jobs=self.args.jobs,
             params=self.args.set_param,
-            checkpoint_resume=self.args.checkpoint_resume,
-            reset=self.args.reset,
             tmp_dir=self.args.tmp_dir,
-            machine=self.args.machine,
             copy_paths=self.args.copy_paths,
             message=self.args.message,
             **self._common_kwargs,
         )
 
-        if self.args.metrics and results:
-            metrics = self.repo.metrics.show(revs=list(results))
-            metrics.pop("workspace", None)
-            show_metrics(metrics)
-
         return 0
 
 
@@ -59,22 +35,6 @@ def add_parser(experiments_subparsers, parent_parser):
         formatter_class=argparse.RawDescriptionHelpFormatter,
     )
     _add_run_common(experiments_run_parser)
-    experiments_run_parser.add_argument(
-        "-r",
-        "--rev",
-        type=str,
-        dest="checkpoint_resume",
-        help=(
-            "Continue the specified checkpoint experiment. Can only be used "
-            "in conjunction with --queue or --temp."
-        ),
-        metavar="<experiment_rev>",
-    ).complete = completion.EXPERIMENT
-    experiments_run_parser.add_argument(
-        "--reset",
-        action="store_true",
-        help="Reset existing checkpoints and restart the experiment.",
-    )
     experiments_run_parser.set_defaults(func=CmdExperimentsRun)
 
 
@@ -129,15 +89,6 @@ def _add_run_common(parser):
             "your workspace."
         ),
     )
-    parser.add_argument(
-        "--machine",
-        default=None,
-        help=argparse.SUPPRESS,
-        # help=(
-        #     "Run this experiment on the specified 'dvc machine' instance."
-        # )
-        # metavar="<name>",
-    )
     parser.add_argument(
         "-C",
         "--copy-paths",
@@ -149,7 +100,7 @@ def _add_run_common(parser):
         ),
     )
     parser.add_argument(
-        "-M",
+        "-m",
         "--message",
         type=str,
         default=None,

dvc/commands/experiments/save.py

@@ -53,7 +53,6 @@ def add_parser(experiments_subparsers, parent_parser):
     )
     save_parser.add_argument(
         "--json",
-        "--show-json",
         action="store_true",
         default=False,
         help="Show output in JSON format.",

dvc/commands/experiments/show.py

@@ -1,13 +1,11 @@
 import argparse
 import logging
-import os
 import re
 from datetime import date, datetime
 from typing import TYPE_CHECKING, Dict, Iterable
 
 from funcy import lmap
 
-from dvc.cli import completion
 from dvc.cli.command import CmdBase
 from dvc.cli.utils import append_doc_link
 from dvc.commands.metrics import DEFAULT_PRECISION
@@ -27,9 +25,6 @@ logger = logging.getLogger(__name__)
 
 
 experiment_types = {
-    "checkpoint_tip": "│ ╓",
-    "checkpoint_commit": "│ ╟",
-    "checkpoint_base": "├─╨",
     "branch_commit": "├──",
     "branch_base": "└──",
     "baseline": "",
@@ -59,7 +54,7 @@ def baseline_styler(typ):
     return {"style": "bold"} if typ == "baseline" else {}
 
 
-def show_experiments(  # noqa: C901, PLR0912
+def show_experiments(  # noqa: C901
     td: "TabularData",
     headers: Dict[str, Iterable[str]],
     keep=None,
@@ -67,7 +62,6 @@ def show_experiments(  # noqa: C901, PLR0912
     pager=True,
     csv=False,
     markdown=False,
-    pcp=False,
     **kwargs,
 ):
     if keep:
@@ -110,51 +104,23 @@ def show_experiments(  # noqa: C901, PLR0912
         }
     )
 
-    if kwargs.get("only_changed", False) or pcp:
+    if kwargs.get("only_changed", False):
         td.drop_duplicates("cols", ignore_empty=False)
 
     cols_to_drop = set()
     if drop is not None:
         cols_to_drop = {col for col in td.keys() if re.match(drop, col)}  # noqa: SIM118
-    if pcp:
-        cols_to_drop.add("Created")
     td.drop(*cols_to_drop)
 
-    if pcp:
-        subset = {x for x in td.keys() if x != "Experiment"}  # noqa: SIM118
-        td.dropna(
-            "rows",
-            how="all",
-            subset=subset,
-        )
-        td.drop_duplicates("rows", subset=subset)
-        if "Experiment" in td:
-            td.column("Experiment")[:] = [
-                # remove tree characters
-                str(x).encode("ascii", "ignore").strip().decode()
-                for x in td.column("Experiment")
-            ]
-        out = kwargs.get("out") or "dvc_plots"
-        output_file = os.path.join(out, "index.html")
-        ui.write(
-            td.to_parallel_coordinates(
-                output_path=os.path.abspath(output_file),
-                color_by=kwargs.get("sort_by") or "Experiment",
-            )
-        )
-        if kwargs.get("open"):
-            return ui.open_browser(output_file)
-
-    else:
-        td.render(
-            pager=pager,
-            borders="horizontals",
-            rich_table=True,
-            header_styles=styles,
-            row_styles=row_styles,
-            csv=csv,
-            markdown=markdown,
-        )
+    td.render(
+        pager=pager,
+        borders="horizontals",
+        rich_table=True,
+        header_styles=styles,
+        row_styles=row_styles,
+        csv=csv,
+        markdown=markdown,
+    )
 
 
 def _normalize_headers(names, count):
@@ -221,9 +187,6 @@ class CmdExperimentsShow(CmdBase):
                 csv=self.args.csv,
                 markdown=self.args.markdown,
                 only_changed=self.args.only_changed,
-                pcp=self.args.pcp,
-                out=self.args.out,
-                open=self.args.open,
             )
         return 0
 
@@ -316,21 +279,18 @@ def add_parser(experiments_subparsers, parent_parser):
     )
     experiments_show_parser.add_argument(
         "--json",
-        "--show-json",
         action="store_true",
         default=False,
         help="Print output in JSON format instead of a human-readable table.",
     )
     experiments_show_parser.add_argument(
         "--csv",
-        "--show-csv",
         action="store_true",
         default=False,
         help="Print output in csv format instead of a human-readable table.",
     )
     experiments_show_parser.add_argument(
         "--md",
-        "--show-md",
         action="store_true",
         default=False,
         dest="markdown",
@@ -345,27 +305,6 @@ def add_parser(experiments_subparsers, parent_parser):
         ),
         metavar="<n>",
     )
-    experiments_show_parser.add_argument(
-        "--parallel-coordinates-plot",
-        "--pcp",
-        dest="pcp",
-        action="store_true",
-        default=False,
-        help="Generate a Parallel Coordinates Plot from the tabulated output.",
-    )
-    experiments_show_parser.add_argument(
-        "-o",
-        "--out",
-        default=None,
-        help="Destination folder to save the Parallel Coordinates Plot to",
-        metavar="<path>",
-    ).complete = completion.DIR
-    experiments_show_parser.add_argument(
-        "--open",
-        action="store_true",
-        default=False,
-        help="Open the Parallel Coordinates Plot directly in the browser.",
-    )
     experiments_show_parser.add_argument(
         "--no-fetch",
         dest="fetch_running",

dvc/commands/get.py

@@ -3,7 +3,7 @@ import logging
 
 from dvc.cli import completion
 from dvc.cli.command import CmdBaseNoRepo
-from dvc.cli.utils import append_doc_link
+from dvc.cli.utils import DictAction, append_doc_link
 from dvc.exceptions import DvcException
 
 logger = logging.getLogger(__name__)
@@ -37,6 +37,9 @@ class CmdGet(CmdBaseNoRepo):
                 rev=self.args.rev,
                 jobs=self.args.jobs,
                 force=self.args.force,
+                config=self.args.config,
+                remote=self.args.remote,
+                remote_config=self.args.remote_config,
             )
             return 0
         except CloneError:
@@ -102,4 +105,27 @@ def add_parser(subparsers, parent_parser):
         default=False,
         help="Override local file or folder if exists.",
     )
+    get_parser.add_argument(
+        "--config",
+        type=str,
+        help=(
+            "Path to a config file that will be merged with the config "
+            "in the target repository."
+        ),
+    )
+    get_parser.add_argument(
+        "--remote",
+        type=str,
+        help="Remote name to set as a default in the target repository.",
+    )
+    get_parser.add_argument(
+        "--remote-config",
+        type=str,
+        nargs="*",
+        action=DictAction,
+        help=(
+            "Remote config options to merge with a remote's config (default or one "
+            "specified by '--remote') in the target repository."
+        ),
+    )
     get_parser.set_defaults(func=CmdGet)

dvc/commands/imp.py

@@ -3,7 +3,7 @@ import logging
 
 from dvc.cli import completion
 from dvc.cli.command import CmdBase
-from dvc.cli.utils import append_doc_link
+from dvc.cli.utils import DictAction, append_doc_link
 from dvc.exceptions import DvcException
 
 logger = logging.getLogger(__name__)
@@ -18,15 +18,13 @@ class CmdImport(CmdBase):
                 self.args.url,
                 self.args.path,
                 out=self.args.out,
-                fname=self.args.file,
                 rev=self.args.rev,
                 no_exec=self.args.no_exec,
                 no_download=self.args.no_download,
-                desc=self.args.desc,
-                type=self.args.type,
-                labels=self.args.labels,
-                meta=self.args.meta,
                 jobs=self.args.jobs,
+                config=self.args.config,
+                remote=self.args.remote,
+                remote_config=self.args.remote_config,
             )
         except CloneError:
             logger.exception("failed to import '%s'", self.args.path)
@@ -42,8 +40,6 @@ class CmdImport(CmdBase):
 
 
 def add_parser(subparsers, parent_parser):
-    from .add import _add_annotating_args
-
     IMPORT_HELP = (
         "Download file or directory tracked by DVC or by Git "
         "into the workspace, and track it."
@@ -75,11 +71,6 @@ def add_parser(subparsers, parent_parser):
         help="Git revision (e.g. SHA, branch, tag)",
         metavar="<commit>",
     )
-    import_parser.add_argument(
-        "--file",
-        help="Specify name of the .dvc file this command will generate.",
-        metavar="<filename>",
-    )
     no_download_exec_group = import_parser.add_mutually_exclusive_group()
     no_download_exec_group.add_argument(
         "--no-exec",
@@ -106,6 +97,27 @@ def add_parser(subparsers, parent_parser):
         ),
         metavar="<number>",
     )
-
-    _add_annotating_args(import_parser)
+    import_parser.add_argument(
+        "--config",
+        type=str,
+        help=(
+            "Path to a config file that will be merged with the config "
+            "in the target repository."
+        ),
+    )
+    import_parser.add_argument(
+        "--remote",
+        type=str,
+        help="Remote name to set as a default in the target repository.",
+    )
+    import_parser.add_argument(
+        "--remote-config",
+        type=str,
+        nargs="*",
+        action=DictAction,
+        help=(
+            "Remote config options to merge with a remote's config (default or one "
+            "specified by '--remote') in the target repository."
+        ),
+    )
     import_parser.set_defaults(func=CmdImport)

dvc/commands/imp_url.py

@@ -15,15 +15,10 @@ class CmdImportUrl(CmdBase):
             self.repo.imp_url(
                 self.args.url,
                 out=self.args.out,
-                fname=self.args.file,
                 no_exec=self.args.no_exec,
                 no_download=self.args.no_download,
                 remote=self.args.remote,
                 to_remote=self.args.to_remote,
-                desc=self.args.desc,
-                type=self.args.type,
-                labels=self.args.labels,
-                meta=self.args.meta,
                 jobs=self.args.jobs,
                 force=self.args.force,
                 version_aware=self.args.version_aware,
@@ -41,8 +36,6 @@ class CmdImportUrl(CmdBase):
 
 
 def add_parser(subparsers, parent_parser):
-    from .add import _add_annotating_args
-
     IMPORT_HELP = "Download or copy file from URL and take it under DVC control."
 
     import_parser = subparsers.add_parser(
@@ -70,11 +63,6 @@ def add_parser(subparsers, parent_parser):
     import_parser.add_argument(
         "out", nargs="?", help="Destination path to put files to."
     ).complete = completion.DIR
-    import_parser.add_argument(
-        "--file",
-        help="Specify name of the .dvc file this command will generate.",
-        metavar="<filename>",
-    ).complete = completion.DIR
     import_parser.add_argument(
         "--to-remote",
         action="store_true",
@@ -126,5 +114,4 @@ def add_parser(subparsers, parent_parser):
         default=False,
         help="Import using cloud versioning. Implied if the URL contains a version ID.",
     )
-    _add_annotating_args(import_parser)
     import_parser.set_defaults(func=CmdImportUrl)

dvc/commands/ls/__init__.py

@@ -3,7 +3,7 @@ import logging
 
 from dvc.cli import completion
 from dvc.cli.command import CmdBaseNoRepo
-from dvc.cli.utils import append_doc_link
+from dvc.cli.utils import DictAction, append_doc_link
 from dvc.commands.ls.ls_colors import LsColors
 from dvc.exceptions import DvcException
 from dvc.ui import ui
@@ -34,6 +34,9 @@ class CmdList(CmdBaseNoRepo):
                 rev=self.args.rev,
                 recursive=self.args.recursive,
                 dvc_only=self.args.dvc_only,
+                config=self.args.config,
+                remote=self.args.remote,
+                remote_config=self.args.remote_config,
             )
             if self.args.json:
                 ui.write_json(entries)
@@ -71,7 +74,6 @@ def add_parser(subparsers, parent_parser):
     )
     list_parser.add_argument(
         "--json",
-        "--show-json",
         action="store_true",
         help="Show output in JSON format.",
     )
@@ -81,6 +83,29 @@ def add_parser(subparsers, parent_parser):
         help="Git revision (e.g. SHA, branch, tag)",
         metavar="<commit>",
     )
+    list_parser.add_argument(
+        "--config",
+        type=str,
+        help=(
+            "Path to a config file that will be merged with the config "
+            "in the target repository."
+        ),
+    )
+    list_parser.add_argument(
+        "--remote",
+        type=str,
+        help="Remote name to set as a default in the target repository.",
+    )
+    list_parser.add_argument(
+        "--remote-config",
+        type=str,
+        nargs="*",
+        action=DictAction,
+        help=(
+            "Remote config options to merge with a remote's config (default or one "
+            "specified by '--remote') in the target repository."
+        ),
+    )
     list_parser.add_argument(
         "path",
         nargs="?",

dvc/commands/machine.py

@@ -1,490 +0,0 @@
-import argparse
-from typing import Dict, List
-
-from dvc.cli.command import CmdBase
-from dvc.cli.utils import append_doc_link, fix_subparsers
-from dvc.commands.config import CmdConfig
-from dvc.compare import TabularData
-from dvc.config import ConfigError
-from dvc.exceptions import DvcException
-from dvc.ui import ui
-from dvc.utils import format_link
-
-
-class MachineDisabledError(ConfigError):
-    def __init__(self):
-        super().__init__("Machine feature is disabled")
-
-
-class CmdMachineConfig(CmdConfig):
-    def __init__(self, args):
-        super().__init__(args)
-        if not self.config["feature"].get("machine", False):
-            raise MachineDisabledError
-
-        if getattr(self.args, "name", None):
-            self.args.name = self.args.name.lower()
-
-    def _check_exists(self, conf):
-        if self.args.name not in conf["machine"]:
-            raise ConfigError(f"machine '{self.args.name}' doesn't exist.")
-
-
-class CmdMachineAdd(CmdMachineConfig):
-    def run(self):
-        from dvc.machine import validate_name
-
-        validate_name(self.args.name)
-
-        if self.args.default:
-            ui.write(f"Setting '{self.args.name}' as a default machine.")
-
-        with self.config.edit(self.args.level) as conf:
-            if self.args.name in conf["machine"] and not self.args.force:
-                raise ConfigError(
-                    "machine '{}' already exists. Use `-f|--force` to "
-                    "overwrite it.".format(self.args.name)
-                )
-
-            conf["machine"][self.args.name] = {"cloud": self.args.cloud}
-            if self.args.default:
-                conf["core"]["machine"] = self.args.name
-
-        return 0
-
-
-class CmdMachineRemove(CmdMachineConfig):
-    def run(self):
-        with self.config.edit(self.args.level) as conf:
-            self._check_exists(conf)
-            del conf["machine"][self.args.name]
-
-        up_to_level = self.args.level or "repo"
-        # Remove core.machine refs to this machine in any shadowing configs
-        for level in reversed(self.config.LEVELS):
-            with self.config.edit(level) as conf:
-                if conf["core"].get("machine") == self.args.name:
-                    del conf["core"]["machine"]
-
-            if level == up_to_level:
-                break
-
-        return 0
-
-
-class CmdMachineList(CmdMachineConfig):
-    TABLE_COLUMNS = [
-        "name",
-        "cloud",
-        "region",
-        "image",
-        "spot",
-        "spot_price",
-        "instance_hdd_size",
-        "instance_type",
-        "ssh_private",
-        "startup_script",
-    ]
-
-    PRIVATE_COLUMNS = ["ssh_private", "startup_script"]
-
-    def _hide_private(self, conf):
-        for machine in conf:
-            for column in self.PRIVATE_COLUMNS:
-                if column in conf[machine]:
-                    conf[machine][column] = "***"
-
-    def _show_origin(self):
-        levels = [self.args.level] if self.args.level else self.config.LEVELS
-        for level in levels:
-            conf = self.config.read(level)["machine"]
-            if self.args.name:
-                conf = conf.get(self.args.name, {})
-            self._hide_private(conf)
-            prefix = self._config_file_prefix(True, self.config, level)
-            configs = list(self._format_config(conf, prefix))
-            if configs:
-                ui.write("\n".join(configs))
-
-    def _show_table(self):
-        td = TabularData(self.TABLE_COLUMNS, fill_value="-")
-        conf = self.config.read()["machine"]
-        if self.args.name:
-            conf = {self.args.name: conf.get(self.args.name, {})}
-        self._hide_private(conf)
-        for machine, machine_config in conf.items():
-            machine_config["name"] = machine
-            td.row_from_dict(machine_config)
-        td.dropna("cols", "all")
-        td.render()
-
-    def run(self):
-        if self.args.show_origin:
-            self._show_origin()
-        else:
-            self._show_table()
-        return 0
-
-
-class CmdMachineModify(CmdMachineConfig):
-    def run(self):
-        from dvc.config import merge
-
-        with self.config.edit(self.args.level) as conf:
-            merged = self.config.load_config_to_level(self.args.level)
-            merge(merged, conf)
-            self._check_exists(merged)
-
-            if self.args.name not in conf["machine"]:
-                conf["machine"][self.args.name] = {}
-            section = conf["machine"][self.args.name]
-            if self.args.unset:
-                section.pop(self.args.option, None)
-            else:
-                section[self.args.option] = self.args.value
-        return 0
-
-
-class CmdMachineRename(CmdBase):
-    def _check_exists(self, conf):
-        if self.args.name not in conf["machine"]:
-            raise ConfigError(f"machine '{self.args.name}' doesn't exist.")
-
-    def _rename_default(self, conf):
-        if conf["core"].get("machine") == self.args.name:
-            conf["core"]["machine"] = self.args.new
-
-    def _check_before_rename(self):
-        from dvc.machine import validate_name
-
-        validate_name(self.args.new)
-
-        all_config = self.config.load_config_to_level(None)
-        if self.args.new in all_config.get("machine", {}):
-            raise ConfigError(
-                f"Rename failed. Machine '{self.args.new}' already exists."
-            )
-        ui.write(f"Rename machine '{self.args.name}' to '{self.args.new}'.")
-
-    def run(self):
-        self._check_before_rename()
-
-        with self.config.edit(self.args.level) as conf:
-            self._check_exists(conf)
-            conf["machine"][self.args.new] = conf["machine"][self.args.name]
-            try:
-                assert self.repo.machine
-                self.repo.machine.rename(self.args.name, self.args.new)
-            except DvcException as error:
-                del conf["machine"][self.args.new]
-                raise ConfigError("terraform rename failed") from error
-            del conf["machine"][self.args.name]
-            self._rename_default(conf)
-
-        up_to_level = self.args.level or "repo"
-        for level in reversed(self.config.LEVELS):
-            if level == up_to_level:
-                break
-            with self.config.edit(level) as level_conf:
-                self._rename_default(level_conf)
-
-        return 0
-
-
-class CmdMachineDefault(CmdMachineConfig):
-    def run(self):
-        if self.args.name is None and not self.args.unset:
-            conf = self.config.read(self.args.level)
-            try:
-                ui.write(conf["core"]["machine"])
-            except KeyError:
-                ui.write("No default machine set")
-                return 1
-        else:
-            with self.config.edit(self.args.level) as conf:
-                if self.args.unset:
-                    conf["core"].pop("machine", None)
-                else:
-                    merged_conf = self.config.load_config_to_level(self.args.level)
-                    if (
-                        self.args.name in conf["machine"]
-                        or self.args.name in merged_conf["machine"]
-                    ):
-                        conf["core"]["machine"] = self.args.name
-                    else:
-                        raise ConfigError(
-                            "default machine must be present in machine list."
-                        )
-        return 0
-
-
-class CmdMachineCreate(CmdBase):
-    def run(self):
-        if self.repo.machine is None:
-            raise MachineDisabledError
-
-        self.repo.machine.create(self.args.name)
-        return 0
-
-
-class CmdMachineStatus(CmdBase):
-    INSTANCE_FIELD = ["name", "instance", "status"]
-    SHOWN_FIELD = [
-        "cloud",
-        "instance_ip",
-        "instance_type",
-        "instance_hdd_size",
-        "instance_gpu",
-    ]
-    FILL_VALUE = "-"
-
-    def _add_row(
-        self,
-        name: str,
-        all_status: List[Dict],
-        td: TabularData,
-    ):
-        if not all_status:
-            row = [
-                name,
-                self.FILL_VALUE,
-                "offline",
-            ]  # back to `None` after #7167
-            td.append(row)
-        for i, status in enumerate(all_status, start=1):
-            row = [name, f"num_{i}", "running" if status else "offline"]
-            for field in self.SHOWN_FIELD:
-                value = str(status.get(field, ""))
-                row.append(value)
-            td.append(row)
-
-    def run(self):
-        if self.repo.machine is None:
-            raise MachineDisabledError
-
-        td = TabularData(
-            self.INSTANCE_FIELD + self.SHOWN_FIELD, fill_value=self.FILL_VALUE
-        )
-
-        if self.args.name:
-            all_status = list(self.repo.machine.status(self.args.name))
-            self._add_row(self.args.name, all_status, td)
-        else:
-            name_set = set()
-            for level in self.repo.config.LEVELS:
-                conf = self.repo.config.read(level)["machine"]
-                name_set.update(conf.keys())
-            name_list = list(name_set)
-            for name in sorted(name_list):
-                all_status = list(self.repo.machine.status(name))
-                self._add_row(name, all_status, td)
-
-        td.dropna("cols", "all")
-        td.render()
-        return 0
-
-
-class CmdMachineDestroy(CmdBase):
-    def run(self):
-        if self.repo.machine is None:
-            raise MachineDisabledError
-
-        self.repo.machine.destroy(self.args.name)
-        return 0
-
-
-class CmdMachineSsh(CmdBase):
-    def run(self):
-        if self.repo.machine is None:
-            raise MachineDisabledError
-
-        self.repo.machine.run_shell(self.args.name)
-        return 0
-
-
-def add_parser(subparsers, parent_parser):  # noqa: PLR0915
-    from dvc.commands.config import parent_config_parser
-
-    machine_HELP = "Set up and manage cloud machines."
-    machine_parser = subparsers.add_parser(
-        "machine",
-        parents=[parent_parser],
-        description=append_doc_link(machine_HELP, "machine"),
-        # NOTE: suppress help during development to hide command
-        # help=machine_HELP,
-        formatter_class=argparse.RawDescriptionHelpFormatter,
-    )
-
-    machine_subparsers = machine_parser.add_subparsers(
-        dest="cmd",
-        help="Use `dvc machine CMD --help` for command-specific help.",
-    )
-
-    fix_subparsers(machine_subparsers)
-
-    machine_ADD_HELP = "Add a new data machine."
-    machine_add_parser = machine_subparsers.add_parser(
-        "add",
-        parents=[parent_config_parser, parent_parser],
-        description=append_doc_link(machine_ADD_HELP, "machine/add"),
-        help=machine_ADD_HELP,
-        formatter_class=argparse.RawDescriptionHelpFormatter,
-    )
-    machine_add_parser.add_argument("name", help="Name of the machine")
-    machine_add_parser.add_argument(
-        "cloud",
-        help="Machine cloud. See full list of supported clouds at {}".format(
-            format_link(
-                "https://github.com/iterative/terraform-provider-iterative#machine"
-            )
-        ),
-    )
-    machine_add_parser.add_argument(
-        "-d",
-        "--default",
-        action="store_true",
-        default=False,
-        help="Set as default machine.",
-    )
-    machine_add_parser.add_argument(
-        "-f",
-        "--force",
-        action="store_true",
-        default=False,
-        help="Force overwriting existing configs",
-    )
-    machine_add_parser.set_defaults(func=CmdMachineAdd)
-
-    machine_DEFAULT_HELP = "Set/unset the default machine."
-    machine_default_parser = machine_subparsers.add_parser(
-        "default",
-        parents=[parent_config_parser, parent_parser],
-        description=append_doc_link(machine_DEFAULT_HELP, "machine/default"),
-        help=machine_DEFAULT_HELP,
-        formatter_class=argparse.RawDescriptionHelpFormatter,
-    )
-    machine_default_parser.add_argument("name", nargs="?", help="Name of the machine")
-    machine_default_parser.add_argument(
-        "-u",
-        "--unset",
-        action="store_true",
-        default=False,
-        help="Unset default machine.",
-    )
-    machine_default_parser.set_defaults(func=CmdMachineDefault)
-
-    machine_LIST_HELP = "List the configuration of one/all machines."
-    machine_list_parser = machine_subparsers.add_parser(
-        "list",
-        parents=[parent_config_parser, parent_parser],
-        description=append_doc_link(machine_LIST_HELP, "machine/list"),
-        help=machine_LIST_HELP,
-        formatter_class=argparse.RawDescriptionHelpFormatter,
-    )
-    machine_list_parser.add_argument(
-        "--show-origin",
-        default=False,
-        action="store_true",
-        help="Show the source file containing each config value.",
-    )
-    machine_list_parser.add_argument(
-        "name",
-        nargs="?",
-        type=str,
-        help="name of machine to specify",
-    )
-    machine_list_parser.set_defaults(func=CmdMachineList)
-    machine_MODIFY_HELP = "Modify the configuration of an machine."
-    machine_modify_parser = machine_subparsers.add_parser(
-        "modify",
-        parents=[parent_config_parser, parent_parser],
-        description=append_doc_link(machine_MODIFY_HELP, "machine/modify"),
-        help=machine_MODIFY_HELP,
-        formatter_class=argparse.RawDescriptionHelpFormatter,
-    )
-    machine_modify_parser.add_argument("name", help="Name of the machine")
-    machine_modify_parser.add_argument("option", help="Name of the option to modify.")
-    machine_modify_parser.add_argument(
-        "value", nargs="?", help="(optional) Value of the option."
-    )
-    machine_modify_parser.add_argument(
-        "-u",
-        "--unset",
-        default=False,
-        action="store_true",
-        help="Unset option.",
-    )
-    machine_modify_parser.set_defaults(func=CmdMachineModify)
-
-    machine_RENAME_HELP = "Rename a machine "
-    machine_rename_parser = machine_subparsers.add_parser(
-        "rename",
-        parents=[parent_config_parser, parent_parser],
-        description=append_doc_link(machine_RENAME_HELP, "remote/rename"),
-        help=machine_RENAME_HELP,
-        formatter_class=argparse.RawDescriptionHelpFormatter,
-    )
-    machine_rename_parser.add_argument("name", help="Machine to be renamed")
-    machine_rename_parser.add_argument("new", help="New name of the machine")
-    machine_rename_parser.set_defaults(func=CmdMachineRename)
-
-    machine_REMOVE_HELP = "Remove an machine."
-    machine_remove_parser = machine_subparsers.add_parser(
-        "remove",
-        parents=[parent_config_parser, parent_parser],
-        description=append_doc_link(machine_REMOVE_HELP, "machine/remove"),
-        help=machine_REMOVE_HELP,
-        formatter_class=argparse.RawDescriptionHelpFormatter,
-    )
-    machine_remove_parser.add_argument("name", help="Name of the machine to remove.")
-    machine_remove_parser.set_defaults(func=CmdMachineRemove)
-
-    machine_CREATE_HELP = "Create and start a machine instance."
-    machine_create_parser = machine_subparsers.add_parser(
-        "create",
-        parents=[parent_parser],
-        description=append_doc_link(machine_CREATE_HELP, "machine/create"),
-        help=machine_CREATE_HELP,
-        formatter_class=argparse.RawDescriptionHelpFormatter,
-    )
-    machine_create_parser.add_argument("name", help="Name of the machine to create.")
-    machine_create_parser.set_defaults(func=CmdMachineCreate)
-
-    machine_STATUS_HELP = "List the status of running instances for one/all machines."
-    machine_status_parser = machine_subparsers.add_parser(
-        "status",
-        parents=[parent_parser],
-        description=append_doc_link(machine_STATUS_HELP, "machine/status"),
-        help=machine_STATUS_HELP,
-        formatter_class=argparse.RawDescriptionHelpFormatter,
-    )
-    machine_status_parser.add_argument(
-        "name", nargs="?", help="(optional) Name of the machine."
-    )
-    machine_status_parser.set_defaults(func=CmdMachineStatus)
-
-    machine_DESTROY_HELP = "Destroy an machine instance."
-    machine_destroy_parser = machine_subparsers.add_parser(
-        "destroy",
-        parents=[parent_parser],
-        description=append_doc_link(machine_DESTROY_HELP, "machine/destroy"),
-        help=machine_DESTROY_HELP,
-        formatter_class=argparse.RawDescriptionHelpFormatter,
-    )
-    machine_destroy_parser.add_argument(
-        "name", help="Name of the machine instance to destroy."
-    )
-    machine_destroy_parser.set_defaults(func=CmdMachineDestroy)
-
-    machine_SSH_HELP = "Connect to a machine via SSH."
-    machine_ssh_parser = machine_subparsers.add_parser(
-        "ssh",
-        parents=[parent_parser],
-        description=append_doc_link(machine_SSH_HELP, "machine/ssh"),
-        help=machine_SSH_HELP,
-        formatter_class=argparse.RawDescriptionHelpFormatter,
-    )
-    machine_ssh_parser.add_argument(
-        "name", help="Name of the machine instance to connect to."
-    )
-    machine_ssh_parser.set_defaults(func=CmdMachineSsh)

dvc/commands/metrics.py

@@ -140,14 +140,12 @@ def add_parser(subparsers, parent_parser):
     )
     metrics_show_parser.add_argument(
         "--json",
-        "--show-json",
         action="store_true",
         default=False,
         help="Show output in JSON format.",
     )
     metrics_show_parser.add_argument(
         "--md",
-        "--show-md",
         action="store_true",
         default=False,
         dest="markdown",
@@ -223,14 +221,12 @@ def add_parser(subparsers, parent_parser):
     )
     metrics_diff_parser.add_argument(
         "--json",
-        "--show-json",
         action="store_true",
         default=False,
         help="Show output in JSON format.",
     )
     metrics_diff_parser.add_argument(
         "--md",
-        "--show-md",
         action="store_true",
         default=False,
         dest="markdown",

dvc/commands/params.py

@@ -105,14 +105,12 @@ def add_parser(subparsers, parent_parser):
     )
     params_diff_parser.add_argument(
         "--json",
-        "--show-json",
         action="store_true",
         default=False,
         help="Show output in JSON format.",
     )
     params_diff_parser.add_argument(
         "--md",
-        "--show-md",
         action="store_true",
         default=False,
         dest="markdown",

dvc/commands/plots.py

@@ -430,7 +430,6 @@ def _add_ui_arguments(parser):
     )
     parser.add_argument(
         "--json",
-        "--show-json",
         action="store_true",
         default=False,
         help=argparse.SUPPRESS,

dvc/commands/remote.py

@@ -127,9 +127,7 @@ class CmdRemoteRename(CmdRemote):
         all_config = self.config.load_config_to_level(None)
         if self.args.new in all_config.get("remote", {}):
             raise ConfigError(
-                "Rename failed. Remote name '{}' already exists.".format(
-                    {self.args.new}
-                )
+                f"Rename failed. Remote name {self.args.new!r} already exists."
             )
 
         with self.config.edit(self.args.level) as conf:

dvc/commands/repro.py

@@ -16,12 +16,6 @@ class CmdRepro(CmdBase):
         else:
             ui.write("Use `dvc push` to send your updates to remote storage.")
 
-        if self.args.metrics:
-            from dvc.compare import show_metrics
-
-            metrics = self.repo.metrics.show()
-            show_metrics(metrics)
-
         return 0
 
     @property
@@ -39,6 +33,7 @@ class CmdRepro(CmdBase):
             "force_downstream": self.args.force_downstream,
             "pull": self.args.pull,
             "allow_missing": self.args.allow_missing,
+            "on_error": self.args.on_error,
         }
 
     @property
@@ -105,13 +100,6 @@ and then the stage name name.
         default=False,
         help="Reproduce all stages in the specified directory.",
     )
-    repro_parser.add_argument(
-        "-m",
-        "--metrics",
-        action="store_true",
-        default=False,
-        help="Show metrics after reproduction.",
-    )
     repro_parser.add_argument(
         "--downstream",
         action="store_true",
@@ -150,6 +138,26 @@ and then the stage name name.
             "Only print the commands that would be executed without actually executing."
         ),
     )
+    repro_parser.add_argument(
+        "-k",
+        "--keep-going",
+        action="store_const",
+        default="fail",
+        const="keep-going",
+        dest="on_error",
+        help=(
+            "Continue executing, skipping stages having dependencies "
+            "on the failed stages"
+        ),
+    )
+    repro_parser.add_argument(
+        "--ignore-errors",
+        action="store_const",
+        default="fail",
+        const="ignore",
+        dest="on_error",
+        help="Ignore errors from stages.",
+    )
 
 
 def add_parser(subparsers, parent_parser):

dvc/commands/run.py

@@ -1,96 +0,0 @@
-import argparse
-import logging
-
-from dvc.cli.command import CmdBase
-from dvc.cli.utils import append_doc_link
-from dvc.commands.stage import parse_cmd
-from dvc.exceptions import DvcException
-
-logger = logging.getLogger(__name__)
-
-
-class CmdRun(CmdBase):
-    def run(self):
-        if not any(
-            [
-                self.args.deps,
-                self.args.outs,
-                self.args.outs_no_cache,
-                self.args.metrics,
-                self.args.metrics_no_cache,
-                self.args.plots,
-                self.args.plots_no_cache,
-                self.args.outs_persist,
-                self.args.outs_persist_no_cache,
-                self.args.checkpoints,
-                self.args.params,
-                self.args.command,
-            ]
-        ):  # pragma: no cover
-            logger.error(
-                "too few arguments. Specify at least one: `-d`, `-o`, `-O`, "
-                "`-m`, `-M`, `-p`, `--plots`, `--plots-no-cache`, "
-                "`--outs-persist`, `--outs-persist-no-cache`, `command`."
-            )
-            return 1
-
-        kwargs = vars(self.args)
-        kwargs.update(
-            {
-                "cmd": parse_cmd(self.args.command),
-                "fname": kwargs.pop("file"),
-                "no_exec": self.args.no_exec or bool(self.args.checkpoints),
-                "run_cache": not kwargs.pop("no_run_cache"),
-            }
-        )
-        try:
-            self.repo.run(**kwargs)
-        except DvcException:
-            logger.exception("")
-            return 1
-
-        return 0
-
-
-def add_parser(subparsers, parent_parser):
-    from dvc.commands.stage import _add_common_args
-
-    RUN_HELP = "Generate a dvc.yaml file from a command and execute the command."
-    run_parser = subparsers.add_parser(
-        "run",
-        parents=[parent_parser],
-        description=append_doc_link(RUN_HELP, "run"),
-        help=RUN_HELP,
-        formatter_class=argparse.RawDescriptionHelpFormatter,
-    )
-    run_parser.add_argument("-n", "--name", help="Stage name.")
-    run_parser.add_argument("--file", metavar="<filename>", help=argparse.SUPPRESS)
-    run_parser.add_argument(
-        "--single-stage",
-        action="store_true",
-        default=False,
-        help=argparse.SUPPRESS,
-    )
-    _add_common_args(run_parser)
-    run_parser.add_argument(
-        "--no-exec",
-        action="store_true",
-        default=False,
-        help="Only create dvc.yaml without actually running it.",
-    )
-    run_parser.add_argument(
-        "--no-commit",
-        action="store_true",
-        default=False,
-        help="Don't put files/directories into cache.",
-    )
-    run_parser.add_argument(
-        "--no-run-cache",
-        action="store_true",
-        default=False,
-        help=(
-            "Execute the command even if this stage has already been run "
-            "with the same command/dependencies/outputs/etc before."
-        ),
-    )
-    run_parser.set_defaults(func=CmdRun)

dvc/commands/stage.py

@@ -1,5 +1,6 @@
 import argparse
 import logging
+from contextlib import contextmanager
 from itertools import chain, filterfalse
 from typing import TYPE_CHECKING, Dict, Iterable, List
 
@@ -119,8 +120,22 @@ def parse_cmd(commands: List[str]) -> str:
     return " ".join(map(quote_argument, commands))
 
 
+@contextmanager
+def _disable_logging(highest_level=logging.CRITICAL):
+    previous_level = logging.root.manager.disable
+
+    logging.disable(highest_level)
+
+    try:
+        yield
+    finally:
+        logging.disable(previous_level)
+
+
 class CmdStageAdd(CmdBase):
     def run(self):
+        from dvc.repo import lock_repo
+
         kwargs = vars(self.args)
         kwargs.update(
             {
@@ -128,7 +143,15 @@ class CmdStageAdd(CmdBase):
                 "params": parse_params(self.args.params),
             }
         )
-        self.repo.stage.add(**kwargs)
+
+        with self.repo.scm_context, lock_repo(self.repo):
+            with _disable_logging(logging.INFO):
+                stage = self.repo.stage.add(**kwargs)
+            logger.info("Added stage %r in %r", stage.addressing, stage.relpath)
+            if self.args.run:
+                stage.run()
+                stage.dump(update_pipeline=False)
+
         return 0
 
 
@@ -172,23 +195,6 @@ def _add_common_args(parser):
         help="Declare output file or directory (do not put into DVC cache).",
         metavar="<filename>",
     ).complete = completion.FILE
-    parser.add_argument(
-        "-c",
-        "--checkpoints",
-        action="append",
-        default=[],
-        help=(
-            "Declare checkpoint output file or directory for 'dvc exp run'. "
-            "Not compatible with 'dvc repro'."
-        ),
-        metavar="<filename>",
-    ).complete = completion.FILE
-    parser.add_argument(
-        "--external",
-        action="store_true",
-        default=False,
-        help="Allow outputs that are outside of the DVC repository.",
-    )
     parser.add_argument(
         "--outs-persist",
         action="append",
@@ -257,6 +263,12 @@ def _add_common_args(parser):
             "This doesn't affect any DVC operations."
         ),
     )
+    parser.add_argument(
+        "--run",
+        action="store_true",
+        default=False,
+        help="Execute the stage after generating it.",
+    )
     parser.add_argument(
         "command",
         nargs=argparse.REMAINDER,

dvc/compare.py

@@ -22,7 +22,6 @@ from typing import (
 from funcy import reraise
 
 if TYPE_CHECKING:
-    from dvc.types import StrPath
     from dvc.ui.table import CellT
 
 
@@ -58,7 +57,7 @@ class TabularData(MutableSequence[Sequence["CellT"]]):
         return self._columns[name]
 
     def items(self) -> ItemsView[str, Column]:
-        projection = {k: self.column(k) for k in self.keys()}  # noqa: SIM118
+        projection = {k: self.column(k) for k in self.keys()}
         return projection.items()
 
     def keys(self) -> List[str]:
@@ -161,24 +160,6 @@ class TabularData(MutableSequence[Sequence["CellT"]]):
             writer.writerow(row)
         return buff.getvalue()
 
-    def to_parallel_coordinates(
-        self, output_path: "StrPath", color_by: Optional[str] = None
-    ) -> "StrPath":
-        from dvc_render.html import render_html
-        from dvc_render.plotly import ParallelCoordinatesRenderer
-
-        render_html(
-            renderers=[
-                ParallelCoordinatesRenderer(
-                    self.as_dict(),
-                    color_by=color_by,
-                    fill_value=self._fill_value,
-                )
-            ],
-            output_file=output_path,
-        )
-        return output_path
-
     def add_column(self, name: str) -> None:
         self._columns[name] = Column([self._fill_value] * len(self))
         self._keys.append(name)
@@ -190,8 +171,7 @@ class TabularData(MutableSequence[Sequence["CellT"]]):
                 self.add_column(key)
 
         row: List["CellT"] = [
-            with_value(d.get(key), self._fill_value)
-            for key in self.keys()  # noqa: SIM118
+            with_value(d.get(key), self._fill_value) for key in self.keys()
         ]
         self.append(row)
 
@@ -246,7 +226,7 @@ class TabularData(MutableSequence[Sequence["CellT"]]):
             to_drop -= match_line
 
         if axis == "rows":
-            for name in self.keys():  # noqa: SIM118
+            for name in self.keys():
                 self._columns[name] = Column(
                     [x for n, x in enumerate(self._columns[name]) if n not in to_drop]
                 )
@@ -294,7 +274,7 @@ class TabularData(MutableSequence[Sequence["CellT"]]):
                 else:
                     unique_rows.append(tuple_row)
 
-            for name in self.keys():  # noqa: SIM118
+            for name in self.keys():
                 self._columns[name] = Column(
                     [
                         x

dvc/config.py

@@ -90,9 +90,12 @@ class Config(dict):
     def __init__(
         self,
         dvc_dir: Optional["StrPath"] = None,
+        local_dvc_dir: Optional["StrPath"] = None,
         validate: bool = True,
         fs: Optional["FileSystem"] = None,
         config: Optional["DictStrAny"] = None,
+        remote: Optional[str] = None,
+        remote_config: Optional["DictStrAny"] = None,
     ):  # pylint: disable=super-init-not-called
         from dvc.fs import LocalFileSystem
 
@@ -101,9 +104,15 @@ class Config(dict):
         self.fs = fs or self.wfs
 
         if dvc_dir:
-            self.dvc_dir = self.fs.path.abspath(self.fs.path.realpath(dvc_dir))
+            self.dvc_dir = self.fs.path.abspath(dvc_dir)
 
-        self.load(validate=validate, config=config)
+        self.local_dvc_dir = local_dvc_dir
+        if not fs and not local_dvc_dir:
+            self.local_dvc_dir = dvc_dir
+
+        self.load(
+            validate=validate, config=config, remote=remote, remote_config=remote_config
+        )
 
     @classmethod
     def from_cwd(cls, fs: Optional["FileSystem"] = None, **kwargs):
@@ -136,7 +145,9 @@ class Config(dict):
 
         if self.dvc_dir is not None:
             files["repo"] = self.fs.path.join(self.dvc_dir, self.CONFIG)
-            files["local"] = self.fs.path.join(self.dvc_dir, self.CONFIG_LOCAL)
+
+        if self.local_dvc_dir is not None:
+            files["local"] = self.wfs.path.join(self.local_dvc_dir, self.CONFIG_LOCAL)
 
         return files
 
@@ -154,7 +165,16 @@ class Config(dict):
         with open(config_file, "w+", encoding="utf-8"):
             return Config(dvc_dir)
 
-    def load(self, validate: bool = True, config: Optional["DictStrAny"] = None):
+    def merge(self, config):
+        merge(self, config)
+
+    def load(
+        self,
+        validate: bool = True,
+        config: Optional["DictStrAny"] = None,
+        remote: Optional[str] = None,
+        remote_config: Optional["DictStrAny"] = None,
+    ):
         """Loads config from all the config files.
 
         Raises:
@@ -169,6 +189,17 @@ class Config(dict):
             conf = self.validate(conf)
 
         self.clear()
+
+        if remote:
+            conf["core"]["remote"] = remote
+
+        if remote_config:
+            remote = remote or conf["core"].get("remote")
+            if not remote:
+                raise ValueError("Missing remote name")
+
+            merge(conf, {"remote": {remote: remote_config}})
+
         self.update(conf)
 
     def _get_fs(self, level):
@@ -176,23 +207,32 @@ class Config(dict):
         # the repo.
         return self.fs if level == "repo" else self.wfs
 
-    def _load_config(self, level):
+    @staticmethod
+    def load_file(path, fs=None) -> dict:
         from configobj import ConfigObj, ConfigObjError
 
+        from dvc.fs import localfs
+
+        fs = fs or localfs
+
+        with fs.open(path) as fobj:
+            try:
+                conf_obj = ConfigObj(fobj)
+            except UnicodeDecodeError as exc:
+                raise ConfigError(str(exc)) from exc
+            except ConfigObjError as exc:
+                raise ConfigError(str(exc)) from exc
+
+        return _parse_named(_lower_keys(conf_obj.dict()))
+
+    def _load_config(self, level):
         filename = self.files[level]
         fs = self._get_fs(level)
 
-        if fs.exists(filename):
-            with fs.open(filename) as fobj:
-                try:
-                    conf_obj = ConfigObj(fobj)
-                except UnicodeDecodeError as exc:
-                    raise ConfigError(str(exc)) from exc
-                except ConfigObjError as exc:
-                    raise ConfigError(str(exc)) from exc
-        else:
-            conf_obj = ConfigObj()
-        return _parse_named(_lower_keys(conf_obj.dict()))
+        try:
+            return self.load_file(filename, fs=fs)
+        except FileNotFoundError:
+            return {}
 
     def _save_config(self, level, conf_dict):
         from configobj import ConfigObj
@@ -220,21 +260,30 @@ class Config(dict):
         return conf
 
     @staticmethod
-    def _load_paths(conf, filename):
-        abs_conf_dir = os.path.abspath(os.path.dirname(filename))
+    def _resolve(conf_dir, path):
+        from .config_schema import ExpPath, RelPath
 
-        def resolve(path):
-            from .config_schema import RelPath
+        if re.match(r"\w+://", path):
+            return path
 
-            if os.path.isabs(path) or re.match(r"\w+://", path):
-                return path
+        if os.path.isabs(path):
+            return path
 
-            # on windows convert slashes to backslashes
-            # to have path compatible with abs_conf_dir
-            if os.path.sep == "\\" and "/" in path:
-                path = path.replace("/", "\\")
+        # on windows convert slashes to backslashes
+        # to have path compatible with abs_conf_dir
+        if os.path.sep == "\\" and "/" in path:
+            path = path.replace("/", "\\")
 
-            return RelPath(os.path.join(abs_conf_dir, path))
+        expanded = os.path.expanduser(path)
+        if os.path.isabs(expanded):
+            return ExpPath(expanded, path)
+
+        return RelPath(os.path.abspath(os.path.join(conf_dir, path)))
+
+    @classmethod
+    def _load_paths(cls, conf, filename):
+        conf_dir = os.path.abspath(os.path.dirname(filename))
+        resolve = partial(cls._resolve, conf_dir)
 
         return Config._map_dirs(conf, resolve)
 
@@ -243,11 +292,17 @@ class Config(dict):
         from dvc.fs import localfs
         from dvc.utils import relpath
 
-        from .config_schema import RelPath
+        from .config_schema import ExpPath, RelPath
 
         if re.match(r"\w+://", path):
             return path
 
+        if isinstance(path, ExpPath):
+            return path.def_path
+
+        if os.path.expanduser(path) != path:
+            return localfs.path.as_posix(path)
+
         if isinstance(path, RelPath) or not os.path.isabs(path):
             path = relpath(path, conf_dir)
             return localfs.path.as_posix(path)

dvc/config_schema.py

@@ -8,6 +8,7 @@ from voluptuous import (
     All,
     Any,
     Coerce,
+    Exclusive,
     Invalid,
     Lower,
     Optional,
@@ -74,8 +75,19 @@ def ByUrl(mapping):  # noqa: N802
     return validate
 
 
+class ExpPath(str):
+    __slots__ = ("def_path",)
+
+    def_path: str
+
+    def __new__(cls, string, def_path):
+        ret = super().__new__(cls, string)
+        ret.def_path = def_path
+        return ret
+
+
 class RelPath(str):
-    pass
+    __slots__ = ()
 
 
 class FeatureSchema(Schema):
@@ -154,13 +166,13 @@ SCHEMA = {
         "machine": Lower,
     },
     "cache": {
-        "local": str,
-        "s3": str,
-        "gs": str,
-        "hdfs": str,
-        "webhdfs": str,
-        "ssh": str,
-        "azure": str,
+        "local": str,  # obsoleted
+        "s3": str,  # obsoleted
+        "gs": str,  # obsoleted
+        "hdfs": str,  # obsoleted
+        "webhdfs": str,  # obsoleted
+        "ssh": str,  # obsoleted
+        "azure": str,  # obsoleted
         # This is for default local cache
         "dir": str,
         **LOCAL_COMMON,
@@ -324,7 +336,8 @@ SCHEMA = {
     },
     "hydra": {
         Optional("enabled", default=False): Bool,
-        "config_dir": str,
+        Exclusive("config_dir", "config_source"): str,
+        Exclusive("config_module", "config_source"): str,
         "config_name": str,
     },
     "studio": {

dvc/daemon.py

@@ -28,7 +28,9 @@ def _popen(cmd, **kwargs) -> Popen:
             os.path.abspath(os.path.dirname(__file__)), "__main__.py"
         )
         prefix += [main_entrypoint]
-    return Popen(prefix + cmd, close_fds=True, shell=False, **kwargs)  # nosec B603
+    return Popen(
+        prefix + cmd, close_fds=True, shell=False, **kwargs  # nosec B603  # noqa: S603
+    )
 
 
 def _spawn_windows(cmd, env):

dvc/dagascii.py

@@ -209,6 +209,30 @@ def draw(vertices, edges):
     Args:
         vertices (list): list of graph vertices.
         edges (list): list of graph edges.
+
+    Returns:
+        str: ASCII representation
+
+    Example:
+        >>> from dvc.dagascii import draw
+        >>> vertices = [1, 2, 3, 4]
+        >>> edges = [(1, 2), (2, 3), (2, 4), (1, 4)]
+        >>> print(draw(vertices, edges))
+        +---+     +---+
+        | 3 |     | 4 |
+        +---+    *+---+
+          *    **   *
+          *  **     *
+          * *       *
+        +---+       *
+        | 2 |      *
+        +---+     *
+             *    *
+              *  *
+               **
+             +---+
+             | 1 |
+             +---+
     """
     # pylint: disable=too-many-locals
     # NOTE: coordinates might me negative, so we need to shift

dvc/data_cloud.py

@@ -1,18 +1,19 @@
 """Manages dvc remotes that user can use with push/pull/status commands."""
 
 import logging
-from typing import TYPE_CHECKING, Iterable, Optional
+from typing import TYPE_CHECKING, Iterable, Optional, Set, Tuple
 
 from dvc.config import NoRemoteError, RemoteConfigError
+from dvc.fs.callbacks import Callback
 from dvc.utils.objects import cached_property
 from dvc_data.hashfile.db import get_index
+from dvc_data.hashfile.transfer import TransferResult
 
 if TYPE_CHECKING:
     from dvc.fs import FileSystem
     from dvc_data.hashfile.db import HashFileDB
     from dvc_data.hashfile.hash_info import HashInfo
     from dvc_data.hashfile.status import CompareStatusResult
-    from dvc_data.hashfile.transfer import TransferResult
 
 logger = logging.getLogger(__name__)
 
@@ -29,13 +30,40 @@ class Remote:
 
     @cached_property
     def odb(self) -> "HashFileDB":
+        from dvc.cachemgr import CacheManager
         from dvc_data.hashfile.db import get_odb
+        from dvc_data.hashfile.hash import DEFAULT_ALGORITHM
 
         path = self.path
         if self.worktree:
-            path = self.fs.path.join(path, ".dvc", "cache")
+            path = self.fs.path.join(
+                path, ".dvc", CacheManager.FILES_DIR, DEFAULT_ALGORITHM
+            )
+        else:
+            path = self.fs.path.join(path, CacheManager.FILES_DIR, DEFAULT_ALGORITHM)
+        return get_odb(self.fs, path, hash_name=DEFAULT_ALGORITHM, **self.config)
 
-        return get_odb(self.fs, path, hash_name="md5", **self.config)
+    @cached_property
+    def legacy_odb(self) -> "HashFileDB":
+        from dvc_data.hashfile.db import get_odb
+
+        path = self.path
+        return get_odb(self.fs, path, hash_name="md5-dos2unix", **self.config)
+
+
+def _split_legacy_hash_infos(
+    hash_infos: Iterable["HashInfo"],
+) -> Tuple[Set["HashInfo"], Set["HashInfo"]]:
+    from dvc.cachemgr import LEGACY_HASH_NAMES
+
+    legacy = set()
+    default = set()
+    for hi in hash_infos:
+        if hi.name in LEGACY_HASH_NAMES:
+            legacy.add(hi)
+        else:
+            default.add(hi)
+    return legacy, default
 
 
 class DataCloud:
@@ -63,7 +91,7 @@ class DataCloud:
         if name:
             from dvc.fs import get_cloud_fs
 
-            cls, config, fs_path = get_cloud_fs(self.repo, name=name)
+            cls, config, fs_path = get_cloud_fs(self.repo.config, name=name)
 
             if config.get("worktree"):
                 version_aware = config.get("version_aware")
@@ -101,12 +129,17 @@ class DataCloud:
         self,
         name: Optional[str] = None,
         command: str = "<command>",
+        hash_name: str = "md5",
     ) -> "HashFileDB":
+        from dvc.cachemgr import LEGACY_HASH_NAMES
+
         remote = self.get_remote(name=name, command=command)
         if remote.fs.version_aware or remote.worktree:
             raise NoRemoteError(
                 f"'{command}' is unsupported for cloud versioned remotes"
             )
+        if hash_name in LEGACY_HASH_NAMES:
+            return remote.legacy_odb
         return remote.odb
 
     def _log_missing(self, status: "CompareStatusResult"):
@@ -150,16 +183,47 @@ class DataCloud:
                 By default remote from core.remote config option is used.
             odb: optional ODB to push to. Overrides remote.
         """
-        odb = odb or self.get_remote_odb(remote, "push")
-        return self.transfer(
-            self.repo.cache.local,
-            odb,
-            objs,
-            jobs=jobs,
-            dest_index=get_index(odb),
-            cache_odb=self.repo.cache.local,
-            validate_status=self._log_missing,
-        )
+        if odb is not None:
+            return self._push(objs, jobs=jobs, odb=odb)
+        legacy_objs, default_objs = _split_legacy_hash_infos(objs)
+        result = TransferResult(set(), set())
+        if legacy_objs:
+            odb = self.get_remote_odb(remote, "push", hash_name="md5-dos2unix")
+            t, f = self._push(legacy_objs, jobs=jobs, odb=odb)
+            result.transferred.update(t)
+            result.failed.update(f)
+        if default_objs:
+            odb = self.get_remote_odb(remote, "push")
+            t, f = self._push(default_objs, jobs=jobs, odb=odb)
+            result.transferred.update(t)
+            result.failed.update(f)
+        return result
+
+    def _push(
+        self,
+        objs: Iterable["HashInfo"],
+        *,
+        jobs: Optional[int] = None,
+        odb: "HashFileDB",
+    ) -> "TransferResult":
+        if odb.hash_name == "md5-dos2unix":
+            cache = self.repo.cache.legacy
+        else:
+            cache = self.repo.cache.local
+        with Callback.as_tqdm_callback(
+            desc=f"Pushing to {odb.fs.unstrip_protocol(odb.path)}",
+            unit="file",
+        ) as cb:
+            return self.transfer(
+                cache,
+                odb,
+                objs,
+                jobs=jobs,
+                dest_index=get_index(odb),
+                cache_odb=cache,
+                validate_status=self._log_missing,
+                callback=cb,
+            )
 
     def pull(
         self,
@@ -177,17 +241,49 @@ class DataCloud:
                 By default remote from core.remote config option is used.
             odb: optional ODB to pull from. Overrides remote.
         """
-        odb = odb or self.get_remote_odb(remote, "pull")
-        return self.transfer(
-            odb,
-            self.repo.cache.local,
-            objs,
-            jobs=jobs,
-            src_index=get_index(odb),
-            cache_odb=self.repo.cache.local,
-            verify=odb.verify,
-            validate_status=self._log_missing,
-        )
+        if odb is not None:
+            return self._pull(objs, jobs=jobs, odb=odb)
+        legacy_objs, default_objs = _split_legacy_hash_infos(objs)
+        result = TransferResult(set(), set())
+        if legacy_objs:
+            odb = self.get_remote_odb(remote, "pull", hash_name="md5-dos2unix")
+            assert odb.hash_name == "md5-dos2unix"
+            t, f = self._pull(legacy_objs, jobs=jobs, odb=odb)
+            result.transferred.update(t)
+            result.failed.update(f)
+        if default_objs:
+            odb = self.get_remote_odb(remote, "pull")
+            t, f = self._pull(default_objs, jobs=jobs, odb=odb)
+            result.transferred.update(t)
+            result.failed.update(f)
+        return result
+
+    def _pull(
+        self,
+        objs: Iterable["HashInfo"],
+        *,
+        jobs: Optional[int] = None,
+        odb: "HashFileDB",
+    ) -> "TransferResult":
+        if odb.hash_name == "md5-dos2unix":
+            cache = self.repo.cache.legacy
+        else:
+            cache = self.repo.cache.local
+        with Callback.as_tqdm_callback(
+            desc=f"Fetching from {odb.fs.unstrip_protocol(odb.path)}",
+            unit="file",
+        ) as cb:
+            return self.transfer(
+                odb,
+                cache,
+                objs,
+                jobs=jobs,
+                src_index=get_index(odb),
+                cache_odb=cache,
+                verify=odb.verify,
+                validate_status=self._log_missing,
+                callback=cb,
+            )
 
     def status(
         self,
@@ -206,17 +302,49 @@ class DataCloud:
                 is used.
             odb: optional ODB to check status from. Overrides remote.
         """
+        from dvc_data.hashfile.status import CompareStatusResult
+
+        if odb is not None:
+            return self._status(objs, jobs=jobs, odb=odb)
+        result = CompareStatusResult(set(), set(), set(), set())
+        legacy_objs, default_objs = _split_legacy_hash_infos(objs)
+        if legacy_objs:
+            odb = self.get_remote_odb(remote, "status", hash_name="md5-dos2unix")
+            assert odb.hash_name == "md5-dos2unix"
+            o, m, n, d = self._status(legacy_objs, jobs=jobs, odb=odb)
+            result.ok.update(o)
+            result.missing.update(m)
+            result.new.update(n)
+            result.deleted.update(d)
+        if default_objs:
+            odb = self.get_remote_odb(remote, "status")
+            o, m, n, d = self._status(default_objs, jobs=jobs, odb=odb)
+            result.ok.update(o)
+            result.missing.update(m)
+            result.new.update(n)
+            result.deleted.update(d)
+        return result
+
+    def _status(
+        self,
+        objs: Iterable["HashInfo"],
+        *,
+        jobs: Optional[int] = None,
+        odb: "HashFileDB",
+    ):
         from dvc_data.hashfile.status import compare_status
 
-        if not odb:
-            odb = self.get_remote_odb(remote, "status")
+        if odb.hash_name == "md5-dos2unix":
+            cache = self.repo.cache.legacy
+        else:
+            cache = self.repo.cache.local
         return compare_status(
-            self.repo.cache.local,
+            cache,
             odb,
             objs,
             jobs=jobs,
             dest_index=get_index(odb),
-            cache_odb=self.repo.cache.local,
+            cache_odb=cache,
         )
 
     def get_url_for(self, remote, checksum):

dvc/dependency/__init__.py

@@ -14,7 +14,6 @@ from .repo import RepoDependency
 SCHEMA: Mapping[str, Any] = {
     **ARTIFACT_SCHEMA,
     **RepoDependency.REPO_SCHEMA,
-    **ParamsDependency.PARAM_SCHEMA,
     Output.PARAM_FILES: [DIR_FILES_SCHEMA],
 }
 
@@ -36,7 +35,8 @@ def loadd_from(stage, d_list):
     for d in d_list:
         p = d.pop(Output.PARAM_PATH, None)
         files = d.pop(Output.PARAM_FILES, None)
-        ret.append(_get(stage, p, d, files=files))
+        hash_name = d.pop(Output.PARAM_HASH, None)
+        ret.append(_get(stage, p, d, files=files, hash_name=hash_name))
     return ret
 
 

dvc/dependency/base.py

@@ -50,7 +50,7 @@ class Dependency(Output):
             self.fs_path = self.fs.path.version_path(self.fs_path, self.meta.version_id)
 
     def download(self, to, jobs=None):
-        fs_download(self.fs, self.fs_path, to, jobs=jobs)
+        fs_download(self.fs, self.fs_path, to.fs_path, jobs=jobs)
 
     def save(self):
         super().save()

dvc/dependency/param.py

@@ -2,10 +2,9 @@ import logging
 import os
 import typing
 from collections import defaultdict
-from typing import Dict
+from typing import Any, Dict
 
 import dpath
-from voluptuous import Any
 
 from dvc.exceptions import DvcException
 from dvc.utils.serialize import ParseError, load_path
@@ -34,7 +33,6 @@ class BadParamFileError(DvcException):
 
 class ParamsDependency(Dependency):
     PARAM_PARAMS = "params"
-    PARAM_SCHEMA = {PARAM_PARAMS: Any(dict, list, None)}
     DEFAULT_PARAMS_FILE = "params.yaml"
 
     def __init__(self, stage, path, params=None, repo=None):
@@ -48,6 +46,7 @@ class ParamsDependency(Dependency):
         repo = repo or stage.repo
         path = path or os.path.join(repo.root_dir, self.DEFAULT_PARAMS_FILE)
         super().__init__(stage, path, repo=repo)
+        self.hash_name = self.PARAM_PARAMS
         self.hash_info = hash_info
 
     def dumpd(self, **kwargs):

dvc/dependency/repo.py

@@ -1,6 +1,6 @@
 from typing import TYPE_CHECKING, Dict, Optional, Union
 
-from voluptuous import Required
+from voluptuous import Any, Required
 
 from dvc.utils import as_posix
 
@@ -16,12 +16,16 @@ class RepoDependency(Dependency):
     PARAM_URL = "url"
     PARAM_REV = "rev"
     PARAM_REV_LOCK = "rev_lock"
+    PARAM_CONFIG = "config"
+    PARAM_REMOTE = "remote"
 
     REPO_SCHEMA = {
         PARAM_REPO: {
             Required(PARAM_URL): str,
             PARAM_REV: str,
             PARAM_REV_LOCK: str,
+            PARAM_CONFIG: str,
+            PARAM_REMOTE: Any(str, dict),
         }
     }
 
@@ -60,7 +64,28 @@ class RepoDependency(Dependency):
             self.def_repo[self.PARAM_REV_LOCK] = rev
 
     def dumpd(self, **kwargs) -> Dict[str, Union[str, Dict[str, str]]]:
-        return {self.PARAM_PATH: self.def_path, self.PARAM_REPO: self.def_repo}
+        repo = {self.PARAM_URL: self.def_repo[self.PARAM_URL]}
+
+        rev = self.def_repo.get(self.PARAM_REV)
+        if rev:
+            repo[self.PARAM_REV] = self.def_repo[self.PARAM_REV]
+
+        rev_lock = self.def_repo.get(self.PARAM_REV_LOCK)
+        if rev_lock:
+            repo[self.PARAM_REV_LOCK] = rev_lock
+
+        config = self.def_repo.get(self.PARAM_CONFIG)
+        if config:
+            repo[self.PARAM_CONFIG] = config
+
+        remote = self.def_repo.get(self.PARAM_REMOTE)
+        if remote:
+            repo[self.PARAM_REMOTE] = remote
+
+        return {
+            self.PARAM_PATH: self.def_path,
+            self.PARAM_REPO: repo,
+        }
 
     def update(self, rev: Optional[str] = None):
         if rev:
@@ -77,16 +102,33 @@ class RepoDependency(Dependency):
     def _make_fs(
         self, rev: Optional[str] = None, locked: bool = True
     ) -> "DVCFileSystem":
+        from dvc.config import Config
         from dvc.fs import DVCFileSystem
 
-        config = {"cache": self.repo.config["cache"]}
-        config["cache"]["dir"] = self.repo.cache.local.path
+        rem = self.def_repo.get("remote")
+        if isinstance(rem, dict):
+            remote = None  # type: ignore[unreachable]
+            remote_config = rem
+        else:
+            remote = rem
+            remote_config = None
+
+        conf = self.def_repo.get("config")
+        if conf:
+            config = Config.load_file(conf)
+        else:
+            config = {}
+
+        config["cache"] = self.repo.config["cache"]
+        config["cache"]["dir"] = self.repo.cache.local_cache_dir
 
         return DVCFileSystem(
             url=self.def_repo[self.PARAM_URL],
             rev=rev or self._get_rev(locked=locked),
             subrepos=True,
             config=config,
+            remote=remote,
+            remote_config=remote_config,
         )
 
     def _get_rev(self, locked: bool = True):

dvc/dvcfile.py

@@ -14,7 +14,6 @@ from typing import (
 )
 
 from dvc.exceptions import DvcException
-from dvc.parsing.versions import LOCKFILE_VERSION, SCHEMA_KWD
 from dvc.stage import serialize
 from dvc.stage.exceptions import (
     StageFileBadNameError,
@@ -36,7 +35,6 @@ if TYPE_CHECKING:
 logger = logging.getLogger(__name__)
 _T = TypeVar("_T")
 
-DVC_FILE = "Dvcfile"
 DVC_FILE_SUFFIX = ".dvc"
 PROJECT_FILE = "dvc.yaml"
 LOCK_FILE = "dvc.lock"
@@ -56,10 +54,7 @@ class ParametrizedDumpError(DvcException):
 
 
 def is_valid_filename(path):
-    return path.endswith(DVC_FILE_SUFFIX) or os.path.basename(path) in [
-        DVC_FILE,
-        PROJECT_FILE,
-    ]
+    return path.endswith(DVC_FILE_SUFFIX) or os.path.basename(path) == PROJECT_FILE
 
 
 def is_dvc_file(path):
@@ -102,9 +97,7 @@ class FileMixin:
         self.verify = verify
 
     def __repr__(self):
-        return "{}: {}".format(
-            self.__class__.__name__, relpath(self.path, self.repo.root_dir)
-        )
+        return f"{self.__class__.__name__}: {relpath(self.path, self.repo.root_dir)}"
 
     def __hash__(self):
         return hash(self.path)
@@ -358,36 +351,8 @@ class ProjectFile(FileMixin):
         raise NotImplementedError
 
 
-def get_lockfile_schema(d):
-    from dvc.schema import COMPILED_LOCKFILE_V1_SCHEMA, COMPILED_LOCKFILE_V2_SCHEMA
-
-    schema = {
-        LOCKFILE_VERSION.V1: COMPILED_LOCKFILE_V1_SCHEMA,
-        LOCKFILE_VERSION.V2: COMPILED_LOCKFILE_V2_SCHEMA,
-    }
-
-    version = LOCKFILE_VERSION.from_dict(d)
-    return schema[version]
-
-
-def migrate_lock_v1_to_v2(d, version_info):
-    stages = dict(d)
-
-    for key in stages:
-        d.pop(key)
-
-    # forcing order, meta should always be at the top
-    d.update(version_info)
-    d["stages"] = stages
-
-
-def lockfile_schema(data: _T) -> _T:
-    schema = get_lockfile_schema(data)
-    return schema(data)
-
-
 class Lockfile(FileMixin):
-    SCHEMA = staticmethod(lockfile_schema)  # type: ignore[assignment]
+    from dvc.schema import COMPILED_LOCKFILE_SCHEMA as SCHEMA
 
     def _verify_filename(self):
         pass  # lockfile path is hardcoded, so no need to verify here
@@ -401,21 +366,12 @@ class Lockfile(FileMixin):
             self._check_gitignored()
             return {}, ""
 
-    @property
-    def latest_version_info(self):
-        version = LOCKFILE_VERSION.V2.value  # pylint:disable=no-member
-        return {SCHEMA_KWD: version}
-
     def dump(self, stage, **kwargs):
         stage_data = serialize.to_lockfile(stage, **kwargs)
 
         with modify_yaml(self.path, fs=self.repo.fs) as data:
-            version = LOCKFILE_VERSION.from_dict(data)
-            if version == LOCKFILE_VERSION.V1:
-                logger.info("Migrating lock file '%s' from v1 to v2", self.relpath)
-                migrate_lock_v1_to_v2(data, self.latest_version_info)
-            elif not data:
-                data.update(self.latest_version_info)
+            if not data:
+                data.update({"schema": "2.0"})
                 # order is important, meta should always be at the top
                 logger.info("Generating lock file '%s'", self.relpath)
 
@@ -436,8 +392,7 @@ class Lockfile(FileMixin):
             return
 
         d, _ = self._load_yaml(round_trip=True)
-        version = LOCKFILE_VERSION.from_dict(d)
-        data = d if version == LOCKFILE_VERSION.V1 else d.get("stages", {})
+        data = d.get("stages", {})
         if stage.name not in data:
             return
 

dvc/env.py

@@ -1,5 +1,3 @@
-DVCLIVE_RESUME = "DVCLIVE_RESUME"
-DVC_CHECKPOINT = "DVC_CHECKPOINT"
 DVC_DAEMON = "DVC_DAEMON"
 DVC_EXP_AUTO_PUSH = "DVC_EXP_AUTO_PUSH"
 DVC_EXP_BASELINE_REV = "DVC_EXP_BASELINE_REV"

dvc/exceptions.py

@@ -17,6 +17,10 @@ class DvcException(Exception):
 class InvalidArgumentError(ValueError, DvcException):
     """Thrown if arguments are invalid."""
 
+    def __init__(self, msg, *args):
+        self.msg = msg
+        super().__init__(msg, *args)
+
 
 class OutputDuplicationError(DvcException):
     """Thrown if a file/directory is specified as an output in more than one
@@ -63,12 +67,14 @@ class OutputNotFoundError(DvcException):
         self.output = output
         self.repo = repo
         super().__init__(
-            "Unable to find DVC file with output '{path}'".format(
-                path=relpath(self.output)
-            )
+            f"Unable to find DVC file with output {relpath(self.output)!r}"
         )
 
 
+class StageNotFoundError(DvcException):
+    pass
+
+
 class StagePathAsOutputError(DvcException):
     """Thrown if directory that stage is going to be saved in is specified as
     an output of another stage.
@@ -80,11 +86,7 @@ class StagePathAsOutputError(DvcException):
 
     def __init__(self, stage, output):
         assert isinstance(output, str)
-        super().__init__(
-            "{stage} is within an output '{output}' of another stage".format(
-                stage=stage, output=output
-            )
-        )
+        super().__init__(f"{stage} is within an output {output!r} of another stage")
 
 
 class CircularDependencyError(DvcException):
@@ -156,9 +158,7 @@ class CyclicGraphError(DvcException):
 class ConfirmRemoveError(DvcException):
     def __init__(self, path):
         super().__init__(
-            "unable to remove '{}' without a confirmation. Use `-f` to force.".format(
-                path
-            )
+            f"unable to remove {path!r} without a confirmation. Use `-f` to force."
         )
 
 
@@ -167,9 +167,7 @@ class InitError(DvcException):
 
 
 class ReproductionError(DvcException):
-    def __init__(self, name):
-        self.name = name
-        super().__init__(f"failed to reproduce '{name}'")
+    pass
 
 
 class BadMetricError(DvcException):
@@ -182,11 +180,6 @@ class BadMetricError(DvcException):
         )
 
 
-class RecursiveAddingWhileUsingFilename(DvcException):
-    def __init__(self):
-        super().__init__("cannot use `fname` with multiple targets or `-R|--recursive`")
-
-
 class OverlappingOutputPathsError(DvcException):
     def __init__(self, parent, overlapping_out, message):
         self.parent = parent
@@ -304,22 +297,7 @@ class URLMissingError(DvcException):
         super().__init__(f"The path '{url}' does not exist")
 
 
-class RemoteCacheRequiredError(DvcException):
-    def __init__(self, scheme, fs_path):
-        super().__init__(
-            (
-                "Current operation was unsuccessful because '{}' requires "
-                "existing cache on '{}' remote. See {} for information on how "
-                "to set up remote cache."
-            ).format(
-                fs_path,
-                scheme,
-                format_link("https://man.dvc.org/config#cache"),
-            )
-        )
-
-
-class IsADirectoryError(DvcException):  # noqa,pylint:disable=redefined-builtin
+class IsADirectoryError(DvcException):  # noqa: A001,pylint:disable=redefined-builtin
     """Raised when a file operation is requested on a directory."""
 
 

dvc/fs/__init__.py

@@ -1,3 +1,4 @@
+from typing import Optional
 from urllib.parse import urlparse
 
 from dvc_http import HTTPFileSystem, HTTPSFileSystem  # noqa: F401
@@ -18,7 +19,7 @@ from dvc_objects.fs import (  # noqa: F401
     system,
     utils,
 )
-from dvc_objects.fs.base import AnyFSPath, FileSystem  # noqa: F401
+from dvc_objects.fs.base import AnyFSPath, FileSystem  # noqa: F401, TCH001
 from dvc_objects.fs.errors import (  # noqa: F401
     AuthError,
     ConfigError,
@@ -48,18 +49,45 @@ known_implementations.update(
 # pylint: enable=unused-import
 
 
-def download(fs, fs_path, to, jobs=None):
+def download(fs: "FileSystem", fs_path: str, to: str, jobs: Optional[int] = None):
     with Callback.as_tqdm_callback(
         desc=f"Downloading {fs.path.name(fs_path)}",
         unit="files",
     ) as cb:
-        fs.get(fs_path, to.fs_path, batch_size=jobs, callback=cb)
+        # NOTE: We use dvc-objects generic.copy over fs.get since it makes file
+        # download atomic and avoids fsspec glob/regex path expansion.
+        if fs.isdir(fs_path):
+            from_infos = [
+                path
+                for path in fs.find(fs_path)
+                if not path.endswith(fs.path.flavour.sep)
+            ]
+            if not from_infos:
+                return localfs.makedirs(to, exist_ok=True)
+            to_infos = [
+                localfs.path.join(to, *fs.path.relparts(info, fs_path))
+                for info in from_infos
+            ]
+        else:
+            from_infos = [fs_path]
+            to_infos = [to]
+
+        cb.set_size(len(from_infos))
+        jobs = jobs or fs.jobs
+        generic.copy(
+            fs,
+            from_infos,
+            localfs,
+            to_infos,
+            callback=cb,
+            batch_size=jobs,
+        )
 
 
 def parse_external_url(url, config=None):
     remote_config = dict(config) if config else {}
     remote_config["url"] = url
-    fs_cls, fs_config, fs_path = get_cloud_fs(None, **remote_config)
+    fs_cls, fs_config, fs_path = get_cloud_fs({}, **remote_config)
     fs = fs_cls(**fs_config)
     return fs, fs_path
 
@@ -105,18 +133,14 @@ def _resolve_remote_refs(config, remote_conf):
         return remote_conf
 
     base = get_fs_config(config, name=parsed.netloc)
-    cls, _, _ = _get_cloud_fs(config, **base)
+    cls, _, _ = get_cloud_fs(config, **base)
     relpath = parsed.path.lstrip("/").replace("/", cls.sep)
     url = cls.sep.join((base["url"], relpath))
     return {**base, **remote_conf, "url": url}
 
 
-def get_cloud_fs(repo, **kwargs):
-    repo_config = repo.config if repo else {}
-    return _get_cloud_fs(repo_config, **kwargs)
-
-
-def _get_cloud_fs(repo_config, **kwargs):
+def get_cloud_fs(repo_config, **kwargs):
+    repo_config = repo_config or {}
     core_config = repo_config.get("core", {})
 
     remote_conf = get_fs_config(repo_config, **kwargs)

dvc/fs/callbacks.py

@@ -2,12 +2,9 @@
 from contextlib import ExitStack
 from typing import TYPE_CHECKING, Any, BinaryIO, Dict, Optional, Union
 
+from dvc.progress import Tqdm
 from dvc.utils.objects import cached_property
-from dvc_objects.fs.callbacks import (  # noqa: F401
-    DEFAULT_CALLBACK,
-    Callback,
-    TqdmCallback,
-)
+from dvc_objects.fs.callbacks import DEFAULT_CALLBACK, Callback  # noqa: F401
 
 if TYPE_CHECKING:
     from rich.progress import TaskID
@@ -15,6 +12,59 @@ if TYPE_CHECKING:
     from dvc.ui._rich_progress import RichTransferProgress
 
 
+# NOTE: this is very similar to dvc-objects.fs.callbacks.TqdmCallback,
+# but it works with our own Tqdm instance.
+class TqdmCallback(Callback):
+    def __init__(
+        self,
+        size: Optional[int] = None,
+        value: int = 0,
+        progress_bar: Optional["Tqdm"] = None,
+        **tqdm_kwargs,
+    ):
+        tqdm_kwargs["total"] = size or -1
+        self._tqdm_kwargs = tqdm_kwargs
+        self._progress_bar = progress_bar
+        self._stack = ExitStack()
+        super().__init__(size=size, value=value)
+
+    @cached_property
+    def progress_bar(self):
+        progress_bar = (
+            self._progress_bar
+            if self._progress_bar is not None
+            else Tqdm(**self._tqdm_kwargs)
+        )
+        return self._stack.enter_context(progress_bar)
+
+    def __enter__(self):
+        return self
+
+    def close(self):
+        self._stack.close()
+
+    def set_size(self, size):
+        # Tqdm tries to be smart when to refresh,
+        # so we try to force it to re-render.
+        super().set_size(size)
+        self.progress_bar.refresh()
+
+    def call(self, hook_name=None, **kwargs):  # noqa: ARG002
+        self.progress_bar.update_to(self.value, total=self.size)
+
+    def branch(
+        self,
+        path_1: "Union[str, BinaryIO]",
+        path_2: str,
+        kwargs: Dict[str, Any],
+        child: Optional[Callback] = None,
+    ):
+        child = child or TqdmCallback(
+            bytes=True, desc=path_1 if isinstance(path_1, str) else path_2
+        )
+        return super().branch(path_1, path_2, kwargs, child=child)
+
+
 class RichCallback(Callback):
     def __init__(
         self,

dvc/fs/dvc.py

@@ -5,7 +5,7 @@ import ntpath
 import os
 import posixpath
 import threading
-from contextlib import suppress
+from contextlib import ExitStack, suppress
 from typing import TYPE_CHECKING, Any, Callable, Dict, Optional, Tuple, Type, Union
 
 from fsspec.spec import AbstractFileSystem
@@ -38,17 +38,24 @@ def _is_dvc_file(fname):
     return is_valid_filename(fname) or fname == DvcIgnore.DVCIGNORE_FILE
 
 
-def _merge_info(repo, fs_info, dvc_info):
+def _merge_info(repo, key, fs_info, dvc_info):
     from . import utils
 
     ret = {"repo": repo}
 
     if dvc_info:
+        dvc_info["isout"] = any(
+            (len(out_key) <= len(key) and key[: len(out_key)] == out_key)
+            for out_key in repo.index.data_keys["repo"]
+        )
+        dvc_info["isdvc"] = dvc_info["isout"]
         ret["dvc_info"] = dvc_info
         ret["type"] = dvc_info["type"]
         ret["size"] = dvc_info["size"]
         if not fs_info and "md5" in dvc_info:
             ret["md5"] = dvc_info["md5"]
+        if not fs_info and "md5-dos2unix" in dvc_info:
+            ret["md5-dos2unix"] = dvc_info["md5-dos2unix"]
 
     if fs_info:
         ret["type"] = fs_info["type"]
@@ -116,11 +123,13 @@ class _DVCFileSystem(AbstractFileSystem):  # pylint:disable=abstract-method
         from pygtrie import Trie
 
         super().__init__()
+        self._repo_stack = ExitStack()
         if repo is None:
             repo = self._make_repo(url=url, rev=rev, subrepos=subrepos, **repo_kwargs)
             assert repo is not None
             # pylint: disable=protected-access
             repo_factory = repo._fs_conf["repo_factory"]
+            self._repo_stack.enter_context(repo)
 
         if not repo_factory:
             from dvc.repo import Repo
@@ -209,6 +218,7 @@ class _DVCFileSystem(AbstractFileSystem):  # pylint:disable=abstract-method
                     scm=self.repo.scm,
                     repo_factory=self.repo_factory,
                 )
+                self._repo_stack.enter_context(repo)
                 self._datafss[key] = DataFileSystem(index=repo.index.data["repo"])
             self._subrepos_trie[key] = repo
 
@@ -263,10 +273,10 @@ class _DVCFileSystem(AbstractFileSystem):  # pylint:disable=abstract-method
 
     def isdvc(self, path, **kwargs) -> bool:
         """Is this entry dvc-tracked?"""
-        key = self._get_key_from_relative(path)
-        _, dvc_fs, subkey = self._get_subrepo_info(key)
-        dvc_path = _get_dvc_path(dvc_fs, subkey)
-        return dvc_fs is not None and dvc_fs.isdvc(dvc_path, **kwargs)
+        try:
+            return self.info(path).get("dvc_info", {}).get("isout", False)
+        except FileNotFoundError:
+            return False
 
     def ls(  # pylint: disable=arguments-differ # noqa: C901
         self, path, detail=True, dvc_only=False, **kwargs
@@ -278,7 +288,7 @@ class _DVCFileSystem(AbstractFileSystem):  # pylint:disable=abstract-method
         dvc_infos = {}
         if dvc_fs:
             dvc_path = _get_dvc_path(dvc_fs, subkey)
-            with suppress(FileNotFoundError):
+            with suppress(FileNotFoundError, NotADirectoryError):
                 for info in dvc_fs.ls(dvc_path, detail=True):
                     dvc_infos[dvc_fs.path.name(info["name"])] = info
             dvc_exists = bool(dvc_infos) or dvc_fs.exists(dvc_path)
@@ -314,7 +324,9 @@ class _DVCFileSystem(AbstractFileSystem):  # pylint:disable=abstract-method
                 continue
 
             entry_path = self.path.join(path, name)
-            info = _merge_info(repo, fs_infos.get(name), dvc_infos.get(name))
+            info = _merge_info(
+                repo, (*subkey, name), fs_infos.get(name), dvc_infos.get(name)
+            )
             info["name"] = entry_path
             infos.append(info)
             paths.append(entry_path)
@@ -340,7 +352,7 @@ class _DVCFileSystem(AbstractFileSystem):  # pylint:disable=abstract-method
                 dvc_info = dvc_fs.fs.index.info(subkey)
                 dvc_path = _get_dvc_path(dvc_fs, subkey)
                 dvc_info["name"] = dvc_path
-            except FileNotFoundError:
+            except KeyError:
                 pass
 
         fs_info = None
@@ -370,7 +382,7 @@ class _DVCFileSystem(AbstractFileSystem):  # pylint:disable=abstract-method
         if not dvc_info and not fs_info:
             raise FileNotFoundError
 
-        info = _merge_info(repo, fs_info, dvc_info)
+        info = _merge_info(repo, subkey, fs_info, dvc_info)
         info["name"] = path
         return info
 
@@ -387,6 +399,9 @@ class _DVCFileSystem(AbstractFileSystem):  # pylint:disable=abstract-method
         dvc_path = _get_dvc_path(dvc_fs, subkey)
         return dvc_fs.get_file(dvc_path, lpath, **kwargs)
 
+    def close(self):
+        self._repo_stack.close()
+
 
 class DVCFileSystem(FileSystem):
     protocol = "local"
@@ -397,7 +412,7 @@ class DVCFileSystem(FileSystem):
 
     @functools.cached_property
     # pylint: disable-next=invalid-overridden-method
-    def fs(self) -> "DVCFileSystem":
+    def fs(self) -> "_DVCFileSystem":
         return _DVCFileSystem(**self.fs_args)
 
     def isdvc(self, path, **kwargs) -> bool:
@@ -420,3 +435,7 @@ class DVCFileSystem(FileSystem):
             path = os.path.relpath(path, self.repo.root_dir)
 
         return as_posix(path)
+
+    def close(self):
+        if "fs" in self.__dict__:
+            self.fs.close()

dvc/info.py

@@ -6,13 +6,12 @@ import platform
 
 import psutil
 
-from dvc import __version__
+from dvc import PKG, __version__
 from dvc.exceptions import NotDvcRepoError
 from dvc.fs import Schemes, generic, get_fs_cls, get_fs_config, registry
 from dvc.repo import Repo
 from dvc.scm import SCMError
 from dvc.utils import error_link
-from dvc.utils.pkg import PKG
 
 SUBPROJECTS = (
     "dvc_data",
@@ -68,7 +67,7 @@ def _get_caches(cache):
     caches = (
         cache_type
         for cache_type, cache_instance in cache.by_scheme()
-        if cache_instance and cache_type != "repo"
+        if cache_instance and cache_type not in ("repo", "legacy")
     )
 
     # Caches will be always non-empty including the local cache

dvc/logger.py

@@ -56,7 +56,7 @@ def add_logging_level(level_name, level_num, method_name=None):
 
 class LoggingException(Exception):
     def __init__(self, record):
-        msg = f"failed to log {str(record)}"
+        msg = f"failed to log {record!s}"
         super().__init__(msg)
 
 
@@ -160,7 +160,7 @@ class LoggerHandler(logging.StreamHandler):
             self.flush()
         except (BrokenPipeError, RecursionError):
             raise
-        except Exception:  # noqa, pylint: disable=broad-except
+        except Exception:  # noqa: BLE001, pylint: disable=broad-except
             self.handleError(record)
 
 

dvc/machine/__init__.py

@@ -1,227 +0,0 @@
-import logging
-import os
-from typing import (
-    TYPE_CHECKING,
-    Any,
-    Dict,
-    Iterable,
-    Iterator,
-    Mapping,
-    Optional,
-    Tuple,
-    Type,
-)
-
-from dvc.exceptions import DvcException
-
-if TYPE_CHECKING:
-    from dvc.repo import Repo
-    from dvc.types import StrPath
-
-    from .backend.base import BaseMachineBackend
-
-    BackendCls = Type[BaseMachineBackend]
-
-logger = logging.getLogger(__name__)
-
-
-RESERVED_NAMES = {"local", "localhost"}
-
-DEFAULT_STARTUP_SCRIPT = """#!/bin/bash
-sudo apt-get update
-sudo apt-get install --yes python3 python3-pip
-python3 -m pip install --upgrade pip
-
-pushd /etc/apt/sources.list.d
-sudo wget https://dvc.org/deb/dvc.list
-sudo apt-get update
-sudo apt-get install --yes dvc
-popd
-
-sudo echo "OK" > /var/log/dvc-machine-init.log
-"""
-
-
-def validate_name(name: str):
-    from dvc.exceptions import InvalidArgumentError
-
-    name = name.lower()
-    if name in RESERVED_NAMES:
-        raise InvalidArgumentError(
-            f"Machine name '{name}' is reserved for internal DVC use."
-        )
-
-
-class MachineBackends(Mapping):
-    try:
-        from .backend.terraform import TerraformBackend
-    except ImportError:
-        TerraformBackend = None  # type: ignore[assignment, misc]
-
-    DEFAULT: Dict[str, Optional["BackendCls"]] = {
-        "terraform": TerraformBackend,
-    }
-
-    def __getitem__(self, key: str) -> "BaseMachineBackend":
-        """Lazily initialize backends and cache it afterwards"""
-        initialized = self.initialized.get(key)
-        if not initialized:
-            backend = self.backends[key]
-            initialized = backend(os.path.join(self.tmp_dir, key), **self.kwargs)
-            self.initialized[key] = initialized
-        return initialized
-
-    def __init__(
-        self,
-        selected: Optional[Iterable[str]],
-        tmp_dir: "StrPath",
-        **kwargs,
-    ) -> None:
-        selected = selected or list(self.DEFAULT)
-        self.backends: Dict[str, "BackendCls"] = {}
-        for key in selected:
-            cls = self.DEFAULT.get(key)
-            if cls is None:
-                raise DvcException(
-                    f"'dvc machine' backend '{key}' is missing required "
-                    "dependencies. Install them with:\n"
-                    f"\tpip install dvc[{key}]"
-                )
-            self.backends[key] = cls
-
-        self.initialized: Dict[str, "BaseMachineBackend"] = {}
-
-        self.tmp_dir = tmp_dir
-        self.kwargs = kwargs
-
-    def __iter__(self):
-        return iter(self.backends)
-
-    def __len__(self) -> int:
-        return len(self.backends)
-
-    def close_initialized(self) -> None:
-        for backend in self.initialized.values():
-            backend.close()
-
-
-class MachineManager:
-    """Class that manages dvc cloud machines.
-
-    Args:
-        repo (dvc.repo.Repo): repo instance that belongs to the repo that
-            we are working on.
-
-    Raises:
-        config.ConfigError: thrown when config has invalid format.
-    """
-
-    CLOUD_BACKENDS = {
-        "aws": "terraform",
-        "azure": "terraform",
-    }
-
-    def __init__(
-        self, repo: "Repo", backends: Optional[Iterable[str]] = None, **kwargs
-    ):
-        self.repo = repo
-        assert self.repo.tmp_dir
-        tmp_dir = os.path.join(self.repo.tmp_dir, "machine")
-        self.backends = MachineBackends(backends, tmp_dir=tmp_dir, **kwargs)
-
-    def get_config_and_backend(
-        self,
-        name: Optional[str] = None,
-    ) -> Tuple[dict, "BaseMachineBackend"]:
-        from dvc.config import NoMachineError
-
-        if not name:
-            name = self.repo.config["core"].get("machine")
-
-        if name:
-            config = self._get_config(name=name)
-            backend = self._get_backend(config["cloud"])
-            return config, backend
-
-        if bool(self.repo.config["machine"]):
-            error_msg = (
-                "no machine specified. Setup default machine with\n"
-                "    dvc machine default <name>\n"
-            )
-        else:
-            error_msg = (
-                "no machine specified. Create a default machine with\n"
-                "    dvc machine add -d <name> <cloud>"
-            )
-
-        raise NoMachineError(error_msg)
-
-    def _get_config(self, **kwargs):
-        config = self.repo.config
-        name = kwargs.get("name")
-        if name:
-            try:
-                conf = config["machine"][name.lower()]
-                conf["name"] = name
-            except KeyError:
-                from dvc.config import MachineNotFoundError
-
-                raise MachineNotFoundError(  # noqa: B904
-                    f"machine '{name}' doesn't exist"
-                )
-        else:
-            conf = kwargs
-        return conf
-
-    def _get_backend(self, cloud: str) -> "BaseMachineBackend":
-        from dvc.config import NoMachineError
-
-        try:
-            backend = self.CLOUD_BACKENDS[cloud]
-            return self.backends[backend]
-        except KeyError:
-            raise NoMachineError(  # noqa: B904
-                f"Machine platform '{cloud}' unsupported"
-            )
-
-    def create(self, name: Optional[str]):
-        """Create and start the specified machine instance."""
-        config, backend = self.get_config_and_backend(name)
-        if "startup_script" in config:
-            with open(config["startup_script"], encoding="utf-8") as fobj:
-                startup_script = fobj.read()
-        else:
-            startup_script = DEFAULT_STARTUP_SCRIPT
-        config["startup_script"] = startup_script
-        config.pop("setup_script", None)
-        return backend.create(**config)
-
-    def destroy(self, name: Optional[str]):
-        """Destroy the specified machine instance."""
-        config, backend = self.get_config_and_backend(name)
-        return backend.destroy(**config)
-
-    def get_sshfs(self, name: Optional[str]):
-        config, backend = self.get_config_and_backend(name)
-        return backend.get_sshfs(**config)
-
-    def run_shell(self, name: Optional[str]):
-        config, backend = self.get_config_and_backend(name)
-        return backend.run_shell(**config)
-
-    def status(self, name: str) -> Iterator[Dict[Any, Any]]:
-        config, backend = self.get_config_and_backend(name)
-        return backend.instances(**config)
-
-    def rename(self, name: str, new: str):
-        """move configuration to a new location if the machine is running."""
-        config, backend = self.get_config_and_backend(name)
-        return backend.rename(new=new, **config)
-
-    def get_executor_kwargs(self, name: Optional[str]):
-        config, backend = self.get_config_and_backend(name)
-        return backend.get_executor_kwargs(**config)
-
-    def get_setup_script(self, name: Optional[str]) -> Optional[str]:
-        config, _backend = self.get_config_and_backend(name)
-        return config.get("setup_script")

dvc/machine/backend/base.py

@@ -1,50 +0,0 @@
-from abc import ABC, abstractmethod
-from contextlib import contextmanager
-from typing import TYPE_CHECKING, Iterator, Optional
-
-if TYPE_CHECKING:
-    from dvc_ssh import SSHFileSystem
-
-    from dvc.types import StrPath
-
-
-class BaseMachineBackend(ABC):
-    def __init__(self, tmp_dir: "StrPath", **kwargs):
-        self.tmp_dir = tmp_dir
-
-    @abstractmethod
-    def create(self, name: Optional[str] = None, **config):
-        """Create and start an instance of the specified machine."""
-
-    @abstractmethod
-    def destroy(self, name: Optional[str] = None, **config):
-        """Stop and destroy all instances of the specified machine."""
-
-    @abstractmethod
-    def instances(self, name: Optional[str] = None, **config) -> Iterator[dict]:
-        """Iterate over status of all instances of the specified machine."""
-
-    def close(self):  # noqa: B027
-        pass
-
-    @abstractmethod
-    def run_shell(self, name: Optional[str] = None, **config):
-        """Spawn an interactive SSH shell for the specified machine."""
-
-    @abstractmethod
-    def get_executor_kwargs(self, name: str, **config) -> dict:
-        """Return SSHExecutor kwargs which can be used for DVC
-        experiment/pipeline execution on the specified machine.
-        """
-
-    @abstractmethod
-    @contextmanager
-    def get_sshfs(
-        self, name: Optional[str] = None, **config
-    ) -> Iterator["SSHFileSystem"]:
-        """Return an sshfs instance for the default directory on the
-        specified machine."""
-
-    @abstractmethod
-    def rename(self, name: str, new: str, **config):
-        """Rename a machine instance."""

dvc/machine/backend/terraform.py

@@ -1,99 +0,0 @@
-import os
-from contextlib import contextmanager
-from functools import partial, partialmethod
-from typing import TYPE_CHECKING, Iterator, Optional
-
-from dvc_ssh import DEFAULT_PORT, SSHFileSystem
-
-from dvc.exceptions import DvcException
-
-from .base import BaseMachineBackend
-
-if TYPE_CHECKING:
-    from dvc.types import StrPath
-
-
-@contextmanager
-def _sshfs(resource: dict):
-    from tpi import TerraformProviderIterative
-
-    with TerraformProviderIterative.pemfile(resource) as pem:
-        fs = SSHFileSystem(
-            host=resource["instance_ip"],
-            user="ubuntu",
-            keyfile=pem,
-        )
-        yield fs
-
-
-class TerraformBackend(BaseMachineBackend):
-    def __init__(self, tmp_dir: "StrPath", **kwargs):
-        super().__init__(tmp_dir, **kwargs)
-        os.makedirs(tmp_dir, exist_ok=True)
-
-    @contextmanager
-    def make_tpi(self, name: str):
-        from tpi import TPIError
-        from tpi.terraform import TerraformBackend as TPIBackend
-
-        try:
-            working_dir = os.path.join(self.tmp_dir, name)
-            os.makedirs(working_dir, exist_ok=True)
-            yield TPIBackend(working_dir=working_dir)
-        except TPIError as exc:
-            raise DvcException("TPI operation failed") from exc
-
-    def _tpi_func(self, fname, name: Optional[str] = None, **config):
-        from tpi import TPIError
-
-        assert name
-        with self.make_tpi(name) as tpi:
-            func = getattr(tpi, fname)
-            try:
-                return func(name=name, **config)
-            except TPIError as exc:
-                raise DvcException(f"TPI {fname} failed") from exc
-
-    create = partialmethod(_tpi_func, "create")  # type: ignore[assignment]
-    destroy = partialmethod(_tpi_func, "destroy")  # type: ignore[assignment]
-    instances = partialmethod(_tpi_func, "instances")  # type: ignore[assignment]
-    run_shell = partialmethod(_tpi_func, "run_shell")  # type: ignore[assignment]
-
-    def get_executor_kwargs(self, name: str, **config) -> dict:
-        with self.make_tpi(name) as tpi:
-            resource = tpi.default_resource(name)
-        return {
-            "host": resource["instance_ip"],
-            "port": DEFAULT_PORT,
-            "username": "ubuntu",
-            "fs_factory": partial(_sshfs, dict(resource)),
-        }
-
-    @contextmanager
-    def get_sshfs(  # pylint: disable=unused-argument
-        self, name: Optional[str] = None, **config
-    ) -> Iterator["SSHFileSystem"]:
-        assert name
-        with self.make_tpi(name) as tpi:
-            resource = tpi.default_resource(name)
-        with _sshfs(resource) as fs:
-            yield fs
-
-    def rename(self, name: str, new: str, **config):
-        """rename a dvc machine instance to a new name"""
-        import shutil
-
-        mtype = "iterative_machine"
-
-        new_dir = os.path.join(self.tmp_dir, new)
-        old_dir = os.path.join(self.tmp_dir, name)
-        if os.path.exists(new_dir):
-            raise DvcException(f"rename failed: path {new_dir} already exists")
-
-        if not os.path.exists(old_dir):
-            return
-
-        with self.make_tpi(name) as tpi:
-            tpi.state_mv(f"{mtype}.{name}", f"{mtype}.{new}", **config)
-
-        shutil.move(old_dir, new_dir)

dvc/main.py

@@ -1,2 +0,0 @@
-# TODO: remove in 3.0, kept for backward compatibility
-from .cli import main  # noqa: F401, pylint: disable=unused-import

dvc/output.py

@@ -19,14 +19,15 @@ from dvc.exceptions import (
     ConfirmRemoveError,
     DvcException,
     MergeError,
-    RemoteCacheRequiredError,
 )
+from dvc.utils import format_link
 from dvc.utils.objects import cached_property
 from dvc_data.hashfile import check as ocheck
 from dvc_data.hashfile import load as oload
 from dvc_data.hashfile.build import build
 from dvc_data.hashfile.checkout import checkout
 from dvc_data.hashfile.db import HashFileDB, add_update_tree
+from dvc_data.hashfile.hash import DEFAULT_ALGORITHM
 from dvc_data.hashfile.hash_info import HashInfo
 from dvc_data.hashfile.istextfile import istextfile
 from dvc_data.hashfile.meta import Meta
@@ -64,7 +65,7 @@ CASE_SENSITIVE_CHECKSUM_SCHEMA = Any(
 
 # NOTE: currently there are only 3 possible checksum names:
 #
-#    1) md5 (LOCAL, SSH);
+#    1) md5 (LOCAL, SSH) (actually DVC 2.x md5-dos2unix)
 #    2) etag (S3, GS, OSS, AZURE, HTTP);
 #    3) checksum (HDFS);
 #
@@ -73,7 +74,7 @@ CASE_SENSITIVE_CHECKSUM_SCHEMA = Any(
 HDFS_PARAM_CHECKSUM = "checksum"
 S3_PARAM_CHECKSUM = "etag"
 CHECKSUMS_SCHEMA = {
-    LocalFileSystem.PARAM_CHECKSUM: CHECKSUM_SCHEMA,
+    "md5": CHECKSUM_SCHEMA,  # DVC 2.x md5-dos2unix
     HDFS_PARAM_CHECKSUM: CHECKSUM_SCHEMA,
     S3_PARAM_CHECKSUM: CASE_SENSITIVE_CHECKSUM_SCHEMA,
 }
@@ -91,11 +92,11 @@ def loadd_from(stage, d_list):
         metric = d.pop(Output.PARAM_METRIC, False)
         plot = d.pop(Output.PARAM_PLOT, False)
         persist = d.pop(Output.PARAM_PERSIST, False)
-        checkpoint = d.pop(Output.PARAM_CHECKPOINT, False)
         remote = d.pop(Output.PARAM_REMOTE, None)
         annot = {field: d.pop(field, None) for field in ANNOTATION_FIELDS}
         files = d.pop(Output.PARAM_FILES, None)
         push = d.pop(Output.PARAM_PUSH, True)
+        hash_name = d.pop(Output.PARAM_HASH, None)
         ret.append(
             _get(
                 stage,
@@ -105,11 +106,11 @@ def loadd_from(stage, d_list):
                 metric=metric,
                 plot=plot,
                 persist=persist,
-                checkpoint=checkpoint,
                 remote=remote,
                 **annot,
                 files=files,
                 push=push,
+                hash_name=hash_name,
             )
         )
     return ret
@@ -122,7 +123,6 @@ def loads_from(
     metric=False,
     plot=False,
     persist=False,
-    checkpoint=False,
     remote=None,
     push=True,
 ):
@@ -135,7 +135,6 @@ def loads_from(
             metric=metric,
             plot=plot,
             persist=persist,
-            checkpoint=checkpoint,
             remote=remote,
             push=push,
         )
@@ -191,7 +190,6 @@ def load_from_pipeline(stage, data, typ="outs"):
             [
                 Output.PARAM_CACHE,
                 Output.PARAM_PERSIST,
-                Output.PARAM_CHECKPOINT,
                 Output.PARAM_REMOTE,
                 Output.PARAM_PUSH,
                 *ANNOTATION_FIELDS,
@@ -232,17 +230,29 @@ def merge_file_meta_from_cloud(entry: Dict) -> Dict:
     return entry
 
 
-def _serialize_tree_obj_to_files(obj: "Tree") -> List[Dict[str, Any]]:
+def _serialize_tree_obj_to_files(obj: Tree) -> List[Dict[str, Any]]:
     key = obj.PARAM_RELPATH
     return sorted(
         (
-            {key: posixpath.sep.join(parts), **hi.to_dict(), **meta.to_dict()}
+            {
+                key: posixpath.sep.join(parts),
+                **_serialize_hi_to_dict(hi),
+                **meta.to_dict(),
+            }
             for parts, meta, hi in obj
         ),
         key=itemgetter(key),
     )
 
 
+def _serialize_hi_to_dict(hash_info: Optional[HashInfo]) -> Dict[str, Any]:
+    if hash_info:
+        if hash_info.name == "md5-dos2unix":
+            return {"md5": hash_info.value}
+        return hash_info.to_dict()
+    return {}
+
+
 class OutputDoesNotExistError(DvcException):
     def __init__(self, path):
         msg = f"output '{path}' does not exist"
@@ -286,7 +296,6 @@ class Output:
 
     PARAM_PATH = "path"
     PARAM_CACHE = "cache"
-    PARAM_CHECKPOINT = "checkpoint"
     PARAM_FILES = "files"
     PARAM_METRIC = "metric"
     PARAM_METRIC_TYPE = "type"
@@ -303,15 +312,7 @@ class Output:
     PARAM_REMOTE = "remote"
     PARAM_PUSH = "push"
     PARAM_CLOUD = "cloud"
-
-    METRIC_SCHEMA = Any(
-        None,
-        bool,
-        {
-            PARAM_METRIC_TYPE: Any(str, None),
-            PARAM_METRIC_XPATH: Any(str, None),
-        },
-    )
+    PARAM_HASH = "hash"
 
     DoesNotExistError: Type[DvcException] = OutputDoesNotExistError
     IsNotFileOrDirError: Type[DvcException] = OutputIsNotFileOrDirError
@@ -327,7 +328,6 @@ class Output:
         metric=False,
         plot=False,
         persist=False,
-        checkpoint=False,
         desc=None,
         type=None,  # noqa: A002, pylint: disable=redefined-builtin
         labels=None,
@@ -337,6 +337,7 @@ class Output:
         fs_config=None,
         files: Optional[List[Dict[str, Any]]] = None,
         push: bool = True,
+        hash_name: Optional[str] = DEFAULT_ALGORITHM,
     ):
         self.annot = Annotation(
             desc=desc, type=type, labels=labels or [], meta=meta or {}
@@ -353,7 +354,11 @@ class Output:
         if fs_config is not None:
             fs_kwargs.update(**fs_config)
 
-        fs_cls, fs_config, fs_path = get_cloud_fs(self.repo, url=path, **fs_kwargs)
+        fs_cls, fs_config, fs_path = get_cloud_fs(
+            self.repo.config if self.repo else {},
+            url=path,
+            **fs_kwargs,
+        )
         self.fs = fs_cls(**fs_config)
 
         if (
@@ -394,7 +399,6 @@ class Output:
         self.metric = False if self.IS_DEPENDENCY else metric
         self.plot = False if self.IS_DEPENDENCY else plot
         self.persist = persist
-        self.checkpoint = checkpoint
         self.can_push = push
 
         self.fs_path = self._parse_path(self.fs, fs_path)
@@ -409,17 +413,30 @@ class Output:
             )
             self.meta.version_id = version_id
 
-        if self.is_in_repo:
-            self.hash_name = "md5"
-        else:
-            self.hash_name = self.fs.PARAM_CHECKSUM
-
-        self.hash_info = HashInfo(
-            name=self.hash_name,
-            value=getattr(self.meta, self.hash_name, None),
-        )
+        self.hash_name, self.hash_info = self._compute_hash_info_from_meta(hash_name)
         self._compute_meta_hash_info_from_files()
 
+    def _compute_hash_info_from_meta(
+        self, hash_name: Optional[str]
+    ) -> Tuple[str, HashInfo]:
+        if self.is_in_repo:
+            if hash_name is None:
+                # Legacy 2.x output, use "md5-dos2unix" but read "md5" from
+                # file meta
+                hash_name = "md5-dos2unix"
+                meta_name = "md5"
+            else:
+                meta_name = hash_name
+        else:
+            hash_name = meta_name = self.fs.PARAM_CHECKSUM
+        assert hash_name
+
+        hash_info = HashInfo(
+            name=hash_name,
+            value=getattr(self.meta, meta_name, None),
+        )
+        return hash_name, hash_info
+
     def _compute_meta_hash_info_from_files(self) -> None:
         if self.files:
             tree = Tree.from_list(self.files, hash_name=self.hash_name)
@@ -434,7 +451,7 @@ class Output:
             self.meta.remote = first(f.get("remote") for f in self.files)
         elif self.meta.nfiles or self.hash_info and self.hash_info.isdir:
             self.meta.isdir = True
-            if not self.hash_info and self.hash_name != "md5":
+            if not self.hash_info and self.hash_name not in ("md5", "md5-dos2unix"):
                 md5 = getattr(self.meta, "md5", None)
                 if md5:
                     self.hash_info = HashInfo("md5", md5)
@@ -458,9 +475,7 @@ class Output:
         return fs.path.abspath(fs.path.normpath(fs_path))
 
     def __repr__(self):
-        return "{class_name}: '{def_path}'".format(
-            class_name=type(self).__name__, def_path=self.def_path
-        )
+        return f"{type(self).__name__}: {self.def_path!r}"
 
     def __str__(self):
         if self.fs.protocol != "local":
@@ -473,6 +488,9 @@ class Output:
         ):
             return str(self.def_path)
 
+        if not self.fs.path.isin(self.fs_path, self.repo.root_dir):
+            return self.fs_path
+
         cur_dir = self.fs.path.getcwd()
         if self.fs.path.isin(cur_dir, self.repo.root_dir):
             return self.fs.path.relpath(self.fs_path, cur_dir)
@@ -498,7 +516,7 @@ class Output:
             return False
 
         return self.repo and self.fs.path.isin(
-            self.fs.path.realpath(self.fs_path),
+            self.fs_path,
             self.repo.root_dir,
         )
 
@@ -511,11 +529,19 @@ class Output:
 
     @property
     def cache(self):
-        odb_name = "repo" if self.is_in_repo else self.protocol
-        odb = getattr(self.repo.cache, odb_name)
-        if self.use_cache and odb is None:
-            raise RemoteCacheRequiredError(self.fs.protocol, self.fs_path)
-        return odb
+        from dvc.cachemgr import LEGACY_HASH_NAMES
+
+        assert self.is_in_repo
+        odb_name = "legacy" if self.hash_name in LEGACY_HASH_NAMES else "repo"
+        return getattr(self.repo.cache, odb_name)
+
+    @property
+    def local_cache(self):
+        from dvc.cachemgr import LEGACY_HASH_NAMES
+
+        if self.hash_name in LEGACY_HASH_NAMES:
+            return self.repo.cache.legacy
+        return self.repo.cache.local
 
     @property
     def cache_path(self):
@@ -527,12 +553,24 @@ class Output:
         _, hash_info = self._get_hash_meta()
         return hash_info
 
+    def _build(
+        self, *args, no_progress_bar=False, **kwargs
+    ) -> Tuple["HashFileDB", "Meta", "HashFile"]:
+        from dvc.ui import ui
+
+        with ui.progress(
+            unit="file",
+            desc=f"Collecting files and computing hashes in {self}",
+            disable=no_progress_bar,
+        ) as pb:
+            return build(*args, callback=pb.as_callback(), **kwargs)
+
     def _get_hash_meta(self):
         if self.use_cache:
             odb = self.cache
         else:
-            odb = self.repo.cache.local
-        _, meta, obj = build(
+            odb = self.local_cache
+        _, meta, obj = self._build(
             odb,
             self.fs_path,
             self.fs,
@@ -658,6 +696,14 @@ class Output:
     # pylint: enable=no-member
 
     def save(self) -> None:
+        if self.use_cache and not self.is_in_repo:
+            raise DvcException(
+                f"Saving cached external output {self!s} is not supported "
+                "since DVC 3.0. See "
+                f"{format_link('https://dvc.org/doc/user-guide/upgrade')} "
+                "for more info."
+            )
+
         if not self.exists:
             raise self.DoesNotExistError(self)
 
@@ -672,8 +718,9 @@ class Output:
         if self.metric:
             self.verify_metric()
 
+        self._update_legacy_hash_name()
         if self.use_cache:
-            _, self.meta, self.obj = build(
+            _, self.meta, self.obj = self._build(
                 self.cache,
                 self.fs_path,
                 self.fs,
@@ -681,8 +728,8 @@ class Output:
                 ignore=self.dvcignore,
             )
         else:
-            _, self.meta, self.obj = build(
-                self.repo.cache.local,
+            _, self.meta, self.obj = self._build(
+                self.local_cache,
                 self.fs_path,
                 self.fs,
                 self.hash_name,
@@ -695,6 +742,10 @@ class Output:
         self.hash_info = self.obj.hash_info
         self.files = None
 
+    def _update_legacy_hash_name(self):
+        if self.hash_name == "md5-dos2unix" and self.changed_checksum():
+            self.hash_name = "md5"
+
     def set_exec(self) -> None:
         if self.isfile() and self.meta.isexec:
             self.cache.set_exec(self.fs_path)
@@ -729,20 +780,25 @@ class Output:
             if granular:
                 obj = self._commit_granular_dir(filter_info, hardlink)
             else:
-                staging, _, obj = build(
+                staging, _, obj = self._build(
                     self.cache,
                     filter_info or self.fs_path,
                     self.fs,
                     self.hash_name,
                     ignore=self.dvcignore,
                 )
-                otransfer(
-                    staging,
-                    self.cache,
-                    {obj.hash_info},
-                    shallow=False,
-                    hardlink=hardlink,
-                )
+                with Callback.as_tqdm_callback(
+                    desc=f"Committing {self} to cache",
+                    unit="file",
+                ) as cb:
+                    otransfer(
+                        staging,
+                        self.cache,
+                        {obj.hash_info},
+                        shallow=False,
+                        hardlink=hardlink,
+                        callback=cb,
+                    )
             if relink:
                 rel = self.fs.path.relpath(filter_info or self.fs_path)
                 with CheckoutCallback(desc=f"Checking out {rel}", unit="files") as cb:
@@ -760,7 +816,7 @@ class Output:
 
     def _commit_granular_dir(self, filter_info, hardlink) -> Optional["HashFile"]:
         prefix = self.fs.path.parts(self.fs.path.relpath(filter_info, self.fs_path))
-        staging, _, obj = build(
+        staging, _, obj = self._build(
             self.cache,
             self.fs_path,
             self.fs,
@@ -771,16 +827,23 @@ class Output:
         save_obj = obj.filter(prefix)
         assert isinstance(save_obj, Tree)
         checkout_obj = save_obj.get_obj(self.cache, prefix)
-        otransfer(
-            staging,
-            self.cache,
-            {save_obj.hash_info} | {oid for _, _, oid in save_obj},
-            shallow=True,
-            hardlink=hardlink,
-        )
+        with Callback.as_tqdm_callback(
+            desc=f"Committing {self} to cache",
+            unit="file",
+        ) as cb:
+            otransfer(
+                staging,
+                self.cache,
+                {save_obj.hash_info} | {oid for _, _, oid in save_obj},
+                shallow=True,
+                hardlink=hardlink,
+                callback=cb,
+            )
         return checkout_obj
 
     def dumpd(self, **kwargs):  # noqa: C901, PLR0912
+        from dvc.cachemgr import LEGACY_HASH_NAMES
+
         ret: Dict[str, Any] = {}
         with_files = (
             (not self.IS_DEPENDENCY or self.stage.is_import)
@@ -791,7 +854,12 @@ class Output:
         if not with_files:
             meta_d = self.meta.to_dict()
             meta_d.pop("isdir", None)
-            ret.update(self.hash_info.to_dict())
+            if self.hash_name in LEGACY_HASH_NAMES:
+                # 2.x checksums get serialized with file meta
+                name = "md5" if self.hash_name == "md5-dos2unix" else self.hash_name
+                ret.update({name: self.hash_info.value})
+            else:
+                ret.update(self.hash_info.to_dict())
             ret.update(split_file_meta_from_cloud(meta_d))
 
         if self.is_in_repo:
@@ -799,6 +867,9 @@ class Output:
         else:
             path = self.def_path
 
+        if self.hash_name not in LEGACY_HASH_NAMES:
+            ret[self.PARAM_HASH] = "md5"
+
         ret[self.PARAM_PATH] = path
 
         if not self.IS_DEPENDENCY:
@@ -822,9 +893,6 @@ class Output:
             if self.persist:
                 ret[self.PARAM_PERSIST] = self.persist
 
-            if self.checkpoint:
-                ret[self.PARAM_CHECKPOINT] = self.checkpoint
-
             if self.remote:
                 ret[self.PARAM_REMOTE] = self.remote
 
@@ -894,7 +962,6 @@ class Output:
         relink: bool = False,
         filter_info: Optional[str] = None,
         allow_missing: bool = False,
-        checkpoint_reset: bool = False,
         **kwargs,
     ) -> Optional[Tuple[bool, Optional[bool]]]:
         # callback passed act as a aggregate callback.
@@ -914,11 +981,6 @@ class Output:
             # backward compatibility
             return None
 
-        if self.checkpoint and checkpoint_reset:
-            if self.exists:
-                self.remove()
-            return None
-
         added = not self.exists
 
         try:
@@ -935,14 +997,17 @@ class Output:
                 **kwargs,
             )
         except CheckoutError:
-            if allow_missing or self.checkpoint:
+            if allow_missing:
                 return None
             raise
         self.set_exec()
         return added, False if added else modified
 
     def remove(self, ignore_remove=False):
-        self.fs.remove(self.fs_path, recursive=True)
+        try:
+            self.fs.remove(self.fs_path, recursive=True)
+        except FileNotFoundError:
+            pass
         if self.protocol != Schemes.LOCAL:
             return
 
@@ -969,7 +1034,9 @@ class Output:
         if odb is None:
             odb = self.cache
 
-        cls, config, from_info = get_cloud_fs(self.repo, url=source)
+        cls, config, from_info = get_cloud_fs(
+            self.repo.config if self.repo else {}, url=source
+        )
         from_fs = cls(**config)
 
         # When running import-url --to-remote / add --to-remote/-o ... we
@@ -983,22 +1050,27 @@ class Output:
 
         upload = not (update and from_fs.isdir(from_info))
         jobs = jobs or min((from_fs.jobs, odb.fs.jobs))
-        staging, self.meta, obj = build(
+        staging, self.meta, obj = self._build(
             odb,
             from_info,
             from_fs,
-            "md5",
+            DEFAULT_ALGORITHM,
             upload=upload,
             no_progress_bar=no_progress_bar,
         )
-        otransfer(
-            staging,
-            odb,
-            {obj.hash_info},
-            jobs=jobs,
-            hardlink=False,
-            shallow=False,
-        )
+        with Callback.as_tqdm_callback(
+            desc=f"Transferring to {odb.fs.unstrip_protocol(odb.path)}",
+            unit="file",
+        ) as cb:
+            otransfer(
+                staging,
+                odb,
+                {obj.hash_info},
+                jobs=jobs,
+                hardlink=False,
+                shallow=False,
+                callback=cb,
+            )
 
         self.hash_info = obj.hash_info
         self.files = None
@@ -1018,7 +1090,7 @@ class Output:
         return len(obj) if obj else 0
 
     def unprotect(self):
-        if self.exists:
+        if self.exists and self.use_cache:
             self.cache.unprotect(self.fs_path)
 
     def get_dir_cache(self, **kwargs) -> Optional["Tree"]:
@@ -1128,7 +1200,9 @@ class Output:
             return {}
 
         if self.remote:
-            remote = self.repo.cloud.get_remote_odb(name=self.remote)
+            remote = self.repo.cloud.get_remote_odb(
+                name=self.remote, hash_name=self.hash_name
+            )
         else:
             remote = None
 
@@ -1167,13 +1241,14 @@ class Output:
             self.hash_name,
             Meta.PARAM_SIZE,
             Meta.PARAM_NFILES,
+            Output.PARAM_HASH,
         ]
 
         for opt in ignored:
             my.pop(opt, None)
             other.pop(opt, None)
 
-        if my != other:
+        if my != other or self.hash_name != out.hash_name:
             raise MergeError("unable to auto-merge outputs with different options")
 
         if not out.is_dir_checksum:
@@ -1238,7 +1313,7 @@ class Output:
             raise FileNotFoundError(  # noqa: B904
                 errno.ENOENT,
                 os.strerror(errno.ENOENT),
-                self.fs.path.relpath(path),
+                path,
             )
 
         new = tree.from_trie(trie)
@@ -1309,13 +1384,14 @@ class Output:
             )
 
         assert self.repo
-        cache = self.cache if self.use_cache else self.repo.cache.local
+        self._update_legacy_hash_name()
+        cache = self.cache if self.use_cache else self.local_cache
         assert isinstance(cache, HashFileDB)
 
         new: "HashFile"
         try:
             assert self.hash_name
-            staging, meta, obj = build(
+            staging, meta, obj = self._build(
                 cache,
                 path,
                 self.fs,
@@ -1357,11 +1433,23 @@ class Output:
 
         assert staging
         assert obj.hash_info
-        otransfer(staging, self.cache, {obj.hash_info}, hardlink=relink, shallow=False)
+        with Callback.as_tqdm_callback(
+            desc=f"Adding {self} to cache",
+            unit="file",
+        ) as cb:
+            otransfer(
+                staging,
+                self.cache,
+                {obj.hash_info},
+                hardlink=relink,
+                shallow=False,
+                callback=cb,
+            )
 
         if relink:
-            rel = self.fs.path.relpath(path)
-            with CheckoutCallback(desc=f"Checking out {rel}", unit="files") as callback:
+            with CheckoutCallback(
+                desc=f"Checking out {path}", unit="files"
+            ) as callback:
                 self._checkout(
                     path,
                     self.fs,
@@ -1434,10 +1522,9 @@ ARTIFACT_SCHEMA = {
     **CHECKSUMS_SCHEMA,
     **META_SCHEMA,
     Required(Output.PARAM_PATH): str,
-    Output.PARAM_PLOT: bool,
     Output.PARAM_PERSIST: bool,
-    Output.PARAM_CHECKPOINT: bool,
     Output.PARAM_CLOUD: CLOUD_SCHEMA,
+    Output.PARAM_HASH: str,
 }
 
 DIR_FILES_SCHEMA: Dict[str, Any] = {
@@ -1451,7 +1538,6 @@ SCHEMA = {
     **ARTIFACT_SCHEMA,
     **ANNOTATION_SCHEMA,
     Output.PARAM_CACHE: bool,
-    Output.PARAM_METRIC: Output.METRIC_SCHEMA,
     Output.PARAM_REMOTE: str,
     Output.PARAM_PUSH: bool,
     Output.PARAM_FILES: [DIR_FILES_SCHEMA],

dvc/parsing/__init__.py

@@ -2,6 +2,7 @@ import logging
 import os
 from collections.abc import Mapping, Sequence
 from copy import deepcopy
+from itertools import product
 from typing import (
     TYPE_CHECKING,
     Any,
@@ -51,6 +52,7 @@ VARS_KWD = "vars"
 WDIR_KWD = "wdir"
 PARAMS_KWD = "params"
 FOREACH_KWD = "foreach"
+MATRIX_KWD = "matrix"
 DO_KWD = "do"
 
 DEFAULT_PARAMS_FILE = "params.yaml"
@@ -105,7 +107,7 @@ def is_map_or_seq(data: Any) -> bool:
     return not isinstance(data, str) and _is_map_or_seq(data)
 
 
-def split_foreach_name(name: str) -> Tuple[str, Optional[str]]:
+def split_group_name(name: str) -> Tuple[str, Optional[str]]:
     group, *keys = name.rsplit(JOIN, maxsplit=1)
     return group, first(keys)
 
@@ -120,13 +122,15 @@ def check_interpolations(data: "DictStrAny", where: str, path: str):
     return recurse(func)(data)
 
 
-Definition = Union["ForeachDefinition", "EntryDefinition"]
+Definition = Union["ForeachDefinition", "EntryDefinition", "MatrixDefinition"]
 
 
 def make_definition(
     resolver: "DataResolver", name: str, definition: "DictStrAny", **kwargs
 ) -> Definition:
     args = resolver, resolver.context, name, definition
+    if MATRIX_KWD in definition:
+        return MatrixDefinition(*args, **kwargs)
     if FOREACH_KWD in definition:
         return ForeachDefinition(*args, **kwargs)
     return EntryDefinition(*args, **kwargs)
@@ -135,6 +139,7 @@ def make_definition(
 class DataResolver:
     def __init__(self, repo: "Repo", wdir: str, d: dict):
         self.fs = fs = repo.fs
+        self.parsing_config = repo.config.get("parsing", {})
 
         if os.path.isabs(wdir):
             wdir = fs.path.relpath(wdir)
@@ -158,7 +163,8 @@ class DataResolver:
         self.tracked_vars: Dict[str, Mapping] = {}
 
         stages_data = d.get(STAGES_KWD, {})
-        # we wrap the definitions into ForeachDefinition and EntryDefinition,
+        # we wrap the definitions into:
+        # ForeachDefinition, MatrixDefinition, and EntryDefinition
         # that helps us to optimize, cache and selectively load each one of
         # them as we need, and simplify all of this DSL/parsing logic.
         self.definitions: Dict[str, Definition] = {
@@ -167,12 +173,13 @@ class DataResolver:
         }
 
     def resolve_one(self, name: str):
-        group, key = split_foreach_name(name)
+        group, key = split_group_name(name)
 
         if not self._has_group_and_key(group, key):
             raise EntryNotFound(f"Could not find '{name}'")
 
-        # all of the checks for `key` not being None for `ForeachDefinition`
+        # all of the checks for `key` not being None for
+        # `ForeachDefinition`/`MatrixDefinition`
         # and/or `group` not existing in the `interim`, etc. should be
         # handled by the `self.has_key()` above.
         definition = self.definitions[group]
@@ -189,7 +196,7 @@ class DataResolver:
         return {STAGES_KWD: data}
 
     def has_key(self, key: str):
-        return self._has_group_and_key(*split_foreach_name(key))
+        return self._has_group_and_key(*split_group_name(key))
 
     def _has_group_and_key(self, group: str, key: Optional[str] = None):
         try:
@@ -197,16 +204,14 @@ class DataResolver:
         except KeyError:
             return False
 
-        if key:
-            return isinstance(definition, ForeachDefinition) and definition.has_member(
-                key
-            )
-        return not isinstance(definition, ForeachDefinition)
+        if not isinstance(definition, (ForeachDefinition, MatrixDefinition)):
+            return key is None
+        return key is not None and definition.has_member(key)
 
     @collecting
     def get_keys(self):
         for name, definition in self.definitions.items():
-            if isinstance(definition, ForeachDefinition):
+            if isinstance(definition, (ForeachDefinition, MatrixDefinition)):
                 yield from definition.get_generated_names()
                 continue
             yield name
@@ -255,8 +260,9 @@ class EntryDefinition:
         name = self.name
         if not skip_checks:
             # we can check for syntax errors as we go for interpolated entries,
-            # but for foreach-generated ones, once is enough, which it does
-            # that itself. See `ForeachDefinition.do_definition`.
+            # but for foreach and matrix generated ones, once is enough, which it does
+            # that itself. See `ForeachDefinition.template`
+            # and `MatrixDefinition.template`.
             check_syntax_errors(self.definition, name, self.relpath)
 
         # we need to pop vars from generated/evaluated data
@@ -301,7 +307,10 @@ class EntryDefinition:
     ) -> "DictStrAny":
         try:
             return context.resolve(
-                value, skip_interpolation_checks=skip_checks, key=key
+                value,
+                skip_interpolation_checks=skip_checks,
+                key=key,
+                config=self.resolver.parsing_config,
             )
         except (ParseError, KeyNotInContext) as exc:
             format_and_raise(exc, f"'{self.where}.{self.name}.{key}'", self.relpath)
@@ -327,17 +336,18 @@ class ForeachDefinition:
         self.name = name
 
         assert DO_KWD in definition
+        assert MATRIX_KWD not in definition
         self.foreach_data = definition[FOREACH_KWD]
-        self._do_definition = definition[DO_KWD]
+        self._template = definition[DO_KWD]
 
         self.pair = IterationPair()
         self.where = where
 
     @cached_property
-    def do_definition(self):
+    def template(self):
         # optimization: check for syntax errors only once for `foreach` stages
-        check_syntax_errors(self._do_definition, self.name, self.relpath)
-        return self._do_definition
+        check_syntax_errors(self._template, self.name, self.relpath)
+        return self._template
 
     @cached_property
     def resolved_iterable(self):
@@ -440,16 +450,111 @@ class ForeachDefinition:
             # i.e. quadratic complexity).
             generated = self._generate_name(key)
             entry = EntryDefinition(
-                self.resolver, self.context, generated, self.do_definition
+                self.resolver, self.context, generated, self.template
             )
             try:
                 # optimization: skip checking for syntax errors on each foreach
-                # generated stages. We do it once when accessing do_definition.
+                # generated stages. We do it once when accessing template.
                 return entry.resolve_stage(skip_checks=True)
             except ContextError as exc:
                 format_and_raise(exc, f"stage '{generated}'", self.relpath)
 
-            # let mypy know that this state is unreachable as format_and_raise
-            # does not return at all (it's not able to understand it for some
-            # reason)
-            raise AssertionError("unreachable")
+
+class MatrixDefinition:
+    def __init__(
+        self,
+        resolver: DataResolver,
+        context: Context,
+        name: str,
+        definition: "DictStrAny",
+        where: str = STAGES_KWD,
+    ):
+        self.resolver = resolver
+        self.relpath = self.resolver.relpath
+        self.context = context
+        self.name = name
+
+        assert MATRIX_KWD in definition
+        assert DO_KWD not in definition
+        assert FOREACH_KWD not in definition
+
+        self._template = definition.copy()
+        self.matrix_data = self._template.pop(MATRIX_KWD)
+
+        self.pair = IterationPair()
+        self.where = where
+
+    @cached_property
+    def template(self) -> "DictStrAny":
+        # optimization: check for syntax errors only once for `matrix` stages
+        check_syntax_errors(self._template, self.name, self.relpath)
+        return self._template
+
+    @cached_property
+    def resolved_iterable(self) -> Dict[str, List]:
+        return self._resolve_matrix_data()
+
+    def _resolve_matrix_data(self) -> Dict[str, List]:
+        try:
+            iterable = self.context.resolve(self.matrix_data, unwrap=False)
+        except (ContextError, ParseError) as exc:
+            format_and_raise(exc, f"'{self.where}.{self.name}.matrix'", self.relpath)
+        return iterable
+
+    @cached_property
+    def normalized_iterable(self) -> Dict[str, "DictStrAny"]:
+        """Convert sequence to Mapping with keys normalized."""
+        iterable = self.resolved_iterable
+        assert isinstance(iterable, Mapping)
+
+        ret: Dict[str, "DictStrAny"] = {}
+        matrix = {key: enumerate(v) for key, v in iterable.items()}
+        for combination in product(*matrix.values()):
+            d: "DictStrAny" = {}
+            fragments: List[str] = []
+            for k, (i, v) in zip(matrix.keys(), combination):
+                d[k] = v
+                fragments.append(f"{k}{i}" if is_map_or_seq(v) else to_str(v))
+
+            key = "-".join(fragments)
+            ret[key] = d
+        return ret
+
+    def has_member(self, key: str) -> bool:
+        return key in self.normalized_iterable
+
+    def get_generated_names(self) -> List[str]:
+        return list(map(self._generate_name, self.normalized_iterable))
+
+    def _generate_name(self, key: str) -> str:
+        return f"{self.name}{JOIN}{key}"
+
+    def resolve_all(self) -> "DictStrAny":
+        return join(map(self.resolve_one, self.normalized_iterable))
+
+    def resolve_one(self, key: str) -> "DictStrAny":
+        return self._each_iter(key)
+
+    def _each_iter(self, key: str) -> "DictStrAny":
+        err_message = f"Could not find '{key}' in matrix group '{self.name}'"
+        with reraise(KeyError, EntryNotFound(err_message)):
+            value = self.normalized_iterable[key]
+
+        temp_dict = {self.pair.value: value}
+        with self.context.set_temporarily(temp_dict, reserve=True):
+            # optimization: item and key can be removed on __exit__() as they
+            # are top-level values, and are not merged recursively.
+            # This helps us avoid cloning context, which is slower
+            # (increasing the size of the context might increase
+            # the no. of items to be generated which means more cloning,
+            # i.e. quadratic complexity).
+            generated = self._generate_name(key)
+            entry = EntryDefinition(
+                self.resolver, self.context, generated, self.template
+            )
+            try:
+                # optimization: skip checking for syntax errors on each matrix
+                # generated stages. We do it once when accessing template.
+                return entry.resolve_stage(skip_checks=True)
+            except ContextError as exc:
+                format_and_raise(exc, f"stage '{generated}'", self.relpath)

dvc/parsing/context.py

@@ -495,7 +495,12 @@ class Context(CtxDict):
                 self.data.pop(key, None)
 
     def resolve(
-        self, src, unwrap=True, skip_interpolation_checks=False, key=None
+        self,
+        src,
+        unwrap=True,
+        skip_interpolation_checks=False,
+        key=None,
+        config=None,
     ) -> Any:
         """Recursively resolves interpolation and returns resolved data.
 
@@ -511,10 +516,15 @@ class Context(CtxDict):
         {'lst': [1, 2, 3]}
         """
         func = recurse(self.resolve_str)
-        return func(src, unwrap, skip_interpolation_checks, key)
+        return func(src, unwrap, skip_interpolation_checks, key, config)
 
     def resolve_str(
-        self, src: str, unwrap=True, skip_interpolation_checks=False, key=None
+        self,
+        src: str,
+        unwrap=True,
+        skip_interpolation_checks=False,
+        key=None,
+        config=None,
     ) -> str:
         """Resolves interpolated string to it's original value,
         or in case of multiple interpolations, a combined string.
@@ -535,7 +545,12 @@ class Context(CtxDict):
             return value
         # but not "${num} days"
         return str_interpolate(
-            src, matches, self, skip_checks=skip_interpolation_checks, key=key
+            src,
+            matches,
+            self,
+            skip_checks=skip_interpolation_checks,
+            key=key,
+            config=config,
         )
 
 

dvc/parsing/interpolate.py

@@ -82,20 +82,18 @@ def escape_str(value):
 
 
 @singledispatch
-def to_str(obj) -> str:
+def to_str(obj, config=None) -> str:  # noqa: ARG001, pylint: disable=unused-argument
     return str(obj)
 
 
 @to_str.register(bool)
-def _(obj: bool):
+def _(obj: bool, config=None):  # noqa: ARG001, pylint: disable=unused-argument
     return "true" if obj else "false"
 
 
 @to_str.register(dict)
-def _(obj: dict):  # noqa: C901
-    from dvc.config import Config
-
-    config = Config.from_cwd().get("parsing", {})
+def _(obj: dict, config=None):  # noqa: C901
+    config = config or {}
 
     result = ""
     for k, v in flatten(obj).items():
@@ -211,6 +209,7 @@ def str_interpolate(
     context: "Context",
     skip_checks: bool = False,
     key=None,
+    config=None,
 ):
     index, buf = 0, ""
     for match in matches:
@@ -218,7 +217,7 @@ def str_interpolate(
         expr = get_expression(match, skip_checks=skip_checks)
         value = context.select(expr, unwrap=True)
         validate_value(value, key)
-        buf += template[index:start] + to_str(value)
+        buf += template[index:start] + to_str(value, config=config)
         index = end
     buf += template[index:]
     # regex already backtracks and avoids any `${` starting with

dvc/parsing/versions.py

@@ -1,39 +0,0 @@
-import enum
-from collections.abc import Mapping
-
-from voluptuous import validators
-
-SCHEMA_KWD = "schema"
-META_KWD = "meta"
-
-
-def lockfile_version_schema(value):
-    expected = [LOCKFILE_VERSION.V2.value]  # pylint: disable=no-member
-    msg = f"invalid schema version {value}, expected one of {expected}"
-    return validators.Any(*expected, msg=msg)(value)
-
-
-class VersionEnum(str, enum.Enum):
-    @classmethod
-    def all_versions(cls):
-        return [v.value for v in cls]
-
-
-class LOCKFILE_VERSION(VersionEnum):  # noqa: N801
-    V1 = "1.0"
-    V2 = "2.0"
-
-    @classmethod
-    def from_dict(cls, data):
-        # 1) if it's empty or or is not a dict, use the latest one (V2).
-        # 2) use the `schema` identifier if it exists and is a supported
-        # version
-        # 3) if it's not in any of the supported version, use the latest one
-        # 4) if there's no identifier, it's a V1
-        if not data or not isinstance(data, Mapping):
-            return cls(cls.V2)
-
-        version = data.get(SCHEMA_KWD)
-        if version:
-            return cls(version if version in cls.all_versions() else cls.V2)
-        return cls(cls.V1)

dvc/progress.py

@@ -2,12 +2,16 @@
 import logging
 import sys
 from threading import RLock
+from typing import TYPE_CHECKING
 
 from tqdm import tqdm
 
 from dvc.env import DVC_IGNORE_ISATTY
 from dvc.utils import env2bool
 
+if TYPE_CHECKING:
+    from dvc.fs.callbacks import TqdmCallback
+
 logger = logging.getLogger(__name__)
 tqdm.set_lock(RLock())
 
@@ -158,3 +162,8 @@ class Tqdm(tqdm):
             d["ncols_desc"] = d["ncols_info"] = 1
             d["prefix"] = ""
         return d
+
+    def as_callback(self) -> "TqdmCallback":
+        from dvc.fs.callbacks import TqdmCallback
+
+        return TqdmCallback(progress_bar=self)

dvc/render/match.py

@@ -1,3 +1,4 @@
+import logging
 import os
 from collections import defaultdict
 from typing import TYPE_CHECKING, DefaultDict, Dict, List, NamedTuple, Optional
@@ -17,6 +18,7 @@ if TYPE_CHECKING:
 
 
 dpath.options.ALLOW_EMPTY_STRING_KEYS = True
+logger = logging.getLogger(__name__)
 
 
 def _squash_plots_properties(data: List) -> Dict:
@@ -113,6 +115,7 @@ def match_defs_renderers(  # noqa: C901, PLR0912
             try:
                 dps, rev_props = converter.flat_datapoints(rev)
             except Exception as e:  # noqa: BLE001, pylint: disable=broad-except
+                logger.warning("In %r, %s", rev, str(e).lower())
                 def_errors[rev] = e
                 continue
 

dvc/repo/__init__.py

@@ -8,7 +8,6 @@ from typing import (
     Callable,
     ContextManager,
     Iterable,
-    List,
     Optional,
     Tuple,
     Union,
@@ -16,7 +15,6 @@ from typing import (
 
 from dvc.exceptions import NotDvcRepoError, OutputNotFoundError
 from dvc.ignore import DvcIgnoreFilter
-from dvc.utils import env2bool
 from dvc.utils.fs import path_isin
 from dvc.utils.objects import cached_property
 
@@ -25,12 +23,11 @@ if TYPE_CHECKING:
     from dvc.fs.data import DataFileSystem
     from dvc.fs.dvc import DVCFileSystem
     from dvc.lock import LockBase
-    from dvc.machine import MachineManager
     from dvc.scm import Git, NoSCM
     from dvc.stage import Stage
     from dvc.types import DictStrAny
     from dvc_data.hashfile.state import StateBase
-    from dvc_data.index import DataIndex
+    from dvc_data.index import DataIndex, DataIndexEntry
 
     from .experiments import Experiments
     from .index import Index
@@ -94,6 +91,7 @@ class Repo:
     from dvc.repo.status import status  # type: ignore[misc]
     from dvc.repo.update import update  # type: ignore[misc]
 
+    from .cache import check_missing as cache_check_missing  # type: ignore[misc]
     from .data import status as data_status  # type: ignore[misc]
 
     ls = staticmethod(_ls)
@@ -134,7 +132,7 @@ class Repo:
         assert root_dir
         return root_dir, dvc_dir
 
-    def __init__(  # noqa: PLR0915
+    def __init__(  # noqa: PLR0915, PLR0913
         self,
         root_dir: Optional[str] = None,
         fs: Optional["FileSystem"] = None,
@@ -145,9 +143,10 @@ class Repo:
         url: Optional[str] = None,
         repo_factory: Optional[Callable] = None,
         scm: Optional[Union["Git", "NoSCM"]] = None,
+        remote: Optional[str] = None,
+        remote_config: Optional["DictStrAny"] = None,
     ):
         from dvc.cachemgr import CacheManager
-        from dvc.config import Config
         from dvc.data_cloud import DataCloud
         from dvc.fs import GitFileSystem, LocalFileSystem, localfs
         from dvc.lock import LockNoop, make_lock
@@ -164,6 +163,9 @@ class Repo:
         self._fs_conf = {"repo_factory": repo_factory}
         self._fs = fs or localfs
         self._scm = scm
+        self._config = config
+        self._remote = remote
+        self._remote_config = remote_config
         self._data_index = None
 
         if rev and not fs:
@@ -183,7 +185,6 @@ class Repo:
             scm=scm,
         )
 
-        self.config: Config = Config(self.dvc_dir, fs=self.fs, config=config)
         self._uninitialized = uninitialized
 
         # used by DVCFileSystem to determine if it should traverse subrepos
@@ -235,6 +236,19 @@ class Repo:
     def __str__(self):
         return self.url or self.root_dir
 
+    @cached_property
+    def config(self):
+        from dvc.config import Config
+
+        return Config(
+            self.dvc_dir,
+            local_dvc_dir=self.local_dvc_dir,
+            fs=self.fs,
+            config=self._config,
+            remote=self._remote,
+            remote_config=self._remote_config,
+        )
+
     @cached_property
     def local_dvc_dir(self):
         from dvc.fs import GitFileSystem, LocalFileSystem
@@ -286,7 +300,7 @@ class Repo:
             new.check_graph()
 
     @staticmethod
-    def open(url, *args, **kwargs):  # noqa: A003
+    def open(url: Optional[str], *args, **kwargs) -> "Repo":  # noqa: A003
         from .open_repo import open_repo
 
         return open_repo(url, *args, **kwargs)
@@ -336,16 +350,6 @@ class Repo:
 
         return Experiments(self)
 
-    @cached_property
-    def machine(self) -> Optional["MachineManager"]:
-        from dvc.machine import MachineManager
-
-        if self.tmp_dir and (
-            self.config["feature"].get("machine", False) or env2bool("DVC_TEST")
-        ):
-            return MachineManager(self)
-        return None
-
     @property
     def fs(self) -> "FileSystem":
         return self._fs
@@ -368,6 +372,36 @@ class Repo:
 
         return self._data_index
 
+    def drop_data_index(self) -> None:
+        try:
+            self.data_index.delete_node(("tree",))
+        except KeyError:
+            pass
+        self.data_index.commit()
+        self._reset()
+
+    def get_data_index_entry(
+        self,
+        path: str,
+        workspace: str = "repo",
+    ) -> Tuple["DataIndex", "DataIndexEntry"]:
+        if self.subrepos:
+            fs_path = self.dvcfs.from_os_path(path)
+            fs = self.dvcfs.fs
+            # pylint: disable-next=protected-access
+            key = fs._get_key_from_relative(fs_path)
+            # pylint: disable-next=protected-access
+            subrepo, _, key = fs._get_subrepo_info(key)
+            index = subrepo.index.data[workspace]
+        else:
+            index = self.index.data[workspace]
+            key = self.fs.path.relparts(path, self.root_dir)
+
+        try:
+            return index, index[key]
+        except KeyError as exc:
+            raise OutputNotFoundError(path, self) from exc
+
     def __repr__(self):
         return f"{self.__class__.__name__}: '{self.root_dir}'"
 
@@ -377,7 +411,7 @@ class Repo:
 
         fs = fs or localfs
         root = root or os.curdir
-        root_dir = fs.path.realpath(root)
+        root_dir = fs.path.abspath(root)
 
         if not fs.isdir(root_dir):
             raise NotDvcRepoError(f"directory '{root}' does not exist")
@@ -421,8 +455,8 @@ class Repo:
         flist = [self.config.files["local"]]
         if tmp_dir := self.tmp_dir:
             flist.append(tmp_dir)
-        if path_isin(self.cache.repo.path, self.root_dir):
-            flist.append(self.cache.repo.path)
+        if path_isin(self.cache.legacy.path, self.root_dir):
+            flist.append(self.cache.legacy.path)
 
         for file in flist:
             self.scm_context.ignore(file)
@@ -500,12 +534,6 @@ class Repo:
 
         return used
 
-    @property
-    def stages(
-        self,
-    ) -> List["Stage"]:  # obsolete, only for backward-compatibility
-        return self.index.stages
-
     def find_outs_by_path(self, path, outs=None, recursive=False, strict=True):
         # using `outs_graph` to ensure graph checks are run
         outs = outs or self.index.outs_graph
@@ -556,6 +584,7 @@ class Repo:
 
         from dvc.dirs import site_cache_dir
         from dvc.fs import GitFileSystem
+        from dvc.version import version_tuple
 
         cache_dir = self.config["core"].get("site_cache_dir") or site_cache_dir()
 
@@ -577,7 +606,13 @@ class Repo:
             os.umask(umask)
 
         md5 = hashlib.md5(  # noqa: S324  # nosec B324, B303
-            str((root_dir, getpass.getuser())).encode()
+            str(
+                (
+                    root_dir,
+                    getpass.getuser(),
+                    version_tuple[0],
+                )
+            ).encode()
         )
         repo_token = md5.hexdigest()
         return os.path.join(repos_dir, repo_token)
@@ -585,16 +620,21 @@ class Repo:
     def close(self):
         self.scm.close()
         self.state.close()
+        if "dvcfs" in self.__dict__:
+            self.dvcfs.close()
         if self._data_index is not None:
             self._data_index.close()
 
     def _reset(self):
         self.scm._reset()  # pylint: disable=protected-access
         self.state.close()
+        if "dvcfs" in self.__dict__:
+            self.dvcfs.close()
         self.__dict__.pop("index", None)
         self.__dict__.pop("dvcignore", None)
         self.__dict__.pop("dvcfs", None)
         self.__dict__.pop("datafs", None)
+        self.__dict__.pop("config", None)
 
     def __enter__(self):
         return self

dvc/repo/add.py

@@ -1,34 +1,33 @@
-import logging
 import os
 from contextlib import contextmanager
-from itertools import tee
-from typing import TYPE_CHECKING, Any, Iterator, List, NamedTuple, Optional
-
-import colorama
+from typing import (
+    TYPE_CHECKING,
+    Dict,
+    Iterator,
+    List,
+    NamedTuple,
+    Optional,
+    Tuple,
+    Union,
+)
 
 from dvc.exceptions import (
     CacheLinkError,
     DvcException,
-    InvalidArgumentError,
     OutputDuplicationError,
     OutputNotFoundError,
     OverlappingOutputPathsError,
-    RecursiveAddingWhileUsingFilename,
 )
 from dvc.repo.scm_context import scm_context
 from dvc.ui import ui
-from dvc.utils import LARGE_DIR_SIZE, glob_targets, resolve_output, resolve_paths
-from dvc.utils.collections import ensure_list, validate
+from dvc.utils import glob_targets, resolve_output, resolve_paths
 
 from . import locked
 
 if TYPE_CHECKING:
     from dvc.repo import Repo
     from dvc.stage import Stage
-    from dvc.types import TargetType
-
-Stages = List["Stage"]
-logger = logging.getLogger(__name__)
+    from dvc.types import StrOrBytesPath
 
 
 class StageInfo(NamedTuple):
@@ -36,6 +35,54 @@ class StageInfo(NamedTuple):
     output_exists: bool
 
 
+def find_targets(
+    targets: Union["StrOrBytesPath", Iterator["StrOrBytesPath"]], glob: bool = False
+) -> List[str]:
+    if isinstance(targets, (str, bytes, os.PathLike)):
+        targets_list = [os.fsdecode(targets)]
+    else:
+        targets_list = [os.fsdecode(target) for target in targets]
+    return glob_targets(targets_list, glob=glob)
+
+
+PIPELINE_TRACKED_UPDATE_FMT = (
+    "cannot update {out!r}: overlaps with an output of {stage} in '{path}'.\n"
+    "Run the pipeline or use 'dvc commit' to force update it."
+)
+
+
+def get_or_create_stage(
+    repo: "Repo",
+    target: str,
+    out: Optional[str] = None,
+    to_remote: bool = False,
+    force: bool = False,
+) -> StageInfo:
+    if out:
+        target = resolve_output(target, out, force=force)
+    path, wdir, out = resolve_paths(repo, target, always_local=to_remote and not out)
+
+    try:
+        (out_obj,) = repo.find_outs_by_path(target, strict=False)
+        stage = out_obj.stage
+        if not stage.is_data_source:
+            msg = PIPELINE_TRACKED_UPDATE_FMT.format(
+                out=out, stage=stage, path=stage.relpath
+            )
+            raise DvcException(msg)
+        return StageInfo(stage, output_exists=True)
+    except OutputNotFoundError:
+        stage = repo.stage.create(
+            single_stage=True,
+            validate=False,
+            fname=path,
+            wdir=wdir,
+            outs=[out],
+            force=force,
+        )
+        return StageInfo(stage, output_exists=False)
+
+
 OVERLAPPING_CHILD_FMT = (
     "Cannot add '{out}', because it is overlapping with other "
     "DVC tracked output: '{parent}'.\n"
@@ -51,52 +98,8 @@ OVERLAPPING_PARENT_FMT = (
 )
 
 
-def check_recursive_and_fname(args):
-    if args.recursive and args.fname:
-        raise RecursiveAddingWhileUsingFilename()
-
-
-def transform_targets(args):
-    from funcy import count_reps
-
-    counts = count_reps(ensure_list(args.targets))
-    dupes = [key for key, count in counts.items() if count > 1]
-    if dupes:
-        msg = ", ".join(f"[b]{key}[/]" for key in dupes)
-        ui.error_write(f"ignoring duplicated targets: {msg}", styled=True)
-    args.targets = list(counts)
-
-
-def check_arg_combinations(args):
-    kwargs = args.kwargs
-    invalid_opt = None
-    to_remote = args.to_remote
-    to_cache = kwargs.get("out") and not to_remote
-
-    if to_remote or to_cache:
-        message = "{option} can't be used with "
-        message += "--to-remote" if to_remote else "-o"
-        if len(args.targets) != 1:
-            invalid_opt = "multiple targets"
-        elif args.no_commit:
-            invalid_opt = "--no-commit option"
-        elif args.recursive:
-            invalid_opt = "--recursive option"
-        elif kwargs.get("external"):
-            invalid_opt = "--external option"
-    else:
-        message = "{option} can't be used without --to-remote"
-        if kwargs.get("remote"):
-            invalid_opt = "--remote"
-        elif kwargs.get("jobs"):
-            invalid_opt = "--jobs"
-
-    if invalid_opt is not None:
-        raise InvalidArgumentError(message.format(option=invalid_opt))
-
-
 @contextmanager
-def translate_graph_error(stages: Stages) -> Iterator[None]:
+def translate_graph_error(stages: List["Stage"]) -> Iterator[None]:
     try:
         yield
     except OverlappingOutputPathsError as exc:
@@ -121,18 +124,21 @@ def translate_graph_error(stages: Stages) -> Iterator[None]:
         )
 
 
-def progress_iter(stages: List[StageInfo]) -> Iterator["StageInfo"]:
+def progress_iter(stages: Dict[str, StageInfo]) -> Iterator[Tuple[str, StageInfo]]:
     total = len(stages)
     desc = "Adding..."
-    with ui.progress(stages, total=total, desc=desc, unit="file", leave=True) as pbar:
+    with ui.progress(
+        stages.items(), total=total, desc=desc, unit="file", leave=True
+    ) as pbar:
         if total == 1:
             pbar.bar_format = desc
             pbar.refresh()
 
-        for item in pbar:
+        for item, stage_info in pbar:
             if total > 1:
-                pbar.set_msg(f"{item.stage.outs[0]}")
-            yield item
+                pbar.set_msg(str(stage_info.stage.outs[0]))
+                pbar.refresh()
+            yield item, stage_info
             if total == 1:  # restore bar format for stats
                 # pylint: disable=no-member
                 pbar.bar_format = pbar.BAR_FMT_DEFAULT
@@ -159,145 +165,75 @@ def warn_link_failures() -> Iterator[List[str]]:
             ui.error_write(msg)
 
 
-VALIDATORS = (
-    check_recursive_and_fname,
-    transform_targets,
-    check_arg_combinations,
-)
+def _add_transfer(
+    stage: "Stage",
+    source: str,
+    remote: Optional[str] = None,
+    to_remote: bool = False,
+    jobs: Optional[int] = None,
+    force: bool = False,
+) -> None:
+    odb = None
+    if to_remote:
+        odb = stage.repo.cloud.get_remote_odb(remote, "add")
+    stage.transfer(source, odb=odb, to_remote=to_remote, jobs=jobs, force=force)
+    stage.dump()
+
+
+def _add(stage: "Stage", source: Optional[str] = None, no_commit: bool = False) -> None:
+    out = stage.outs[0]
+    path = out.fs.path.abspath(source) if source else None
+    try:
+        stage.add_outs(path, no_commit=no_commit)
+    except CacheLinkError:
+        stage.dump()
+        raise
+    stage.dump()
 
 
-@validate(*VALIDATORS)
 @locked
 @scm_context
 def add(
     repo: "Repo",
-    targets: "TargetType",
-    recursive: bool = False,
+    targets: Union["StrOrBytesPath", Iterator["StrOrBytesPath"]],
     no_commit: bool = False,
-    fname: Optional[str] = None,
+    glob: bool = False,
+    out: Optional[str] = None,
+    remote: Optional[str] = None,
     to_remote: bool = False,
-    **kwargs: Any,
-):
-    to_cache = bool(kwargs.get("out")) and not to_remote
-    transfer = to_remote or to_cache
+    remote_jobs: Optional[int] = None,
+    force: bool = False,
+) -> List["Stage"]:
+    add_targets = find_targets(targets, glob=glob)
+    if not add_targets:
+        return []
 
-    glob = kwargs.get("glob", False)
-    add_targets = collect_targets(repo, targets, recursive, glob)
-    # pass one for creating stages, other one is used for iterating here
-    add_targets, sources = tee(add_targets)
-
-    # collect targets and build stages as we go
-    desc = "Collecting targets"
-    stages_it = create_stages(repo, add_targets, fname, transfer, **kwargs)
-    stages = list(ui.progress(stages_it, desc=desc, unit="file"))
-
-    stages_list = [stage for stage, _ in stages]
-    msg = "Collecting stages from the workspace"
-    with translate_graph_error(stages_list), ui.status(msg) as status:
-        # remove existing stages that are to-be replaced with these
-        # new stages for the graph checks.
-        repo.check_graph(
-            stages=stages_list, callback=lambda: status.update("Checking graph")
+    stages_with_targets = {
+        target: get_or_create_stage(
+            repo,
+            target,
+            out=out,
+            to_remote=to_remote,
+            force=force,
         )
+        for target in add_targets
+    }
 
-    odb = None
-    if to_remote:
-        odb = repo.cloud.get_remote_odb(kwargs.get("remote"), "add")
+    stages = [stage for stage, _ in stages_with_targets.values()]
+    msg = "Collecting stages from the workspace"
+    with translate_graph_error(stages), ui.status(msg) as st:
+        repo.check_graph(stages=stages, callback=lambda: st.update("Checking graph"))
+
+    if to_remote or out:
+        assert len(stages_with_targets) == 1, "multiple targets are unsupported"
+        (source, (stage, _)) = next(iter(stages_with_targets.items()))
+        _add_transfer(stage, source, remote, to_remote, jobs=remote_jobs, force=force)
+        return [stage]
 
     with warn_link_failures() as link_failures:
-        for (stage, output_exists), source in zip(progress_iter(stages), sources):
-            out = stage.outs[0]
-            if to_remote or to_cache:
-                stage.transfer(source, to_remote=to_remote, odb=odb, **kwargs)
-            else:
-                try:
-                    path = out.fs.path.abspath(source) if output_exists else None
-                    stage.add_outs(path, no_commit=no_commit)
-                except CacheLinkError:
-                    link_failures.append(str(stage.relpath))
-            stage.dump()
-    return stages_list
-
-
-LARGE_DIR_RECURSIVE_ADD_WARNING = (
-    "You are adding a large directory '{target}' recursively.\n"
-    "Consider tracking it as a whole instead with "
-    "`{cyan}dvc add {target}{nc}`."
-)
-
-
-def collect_targets(
-    repo: "Repo",
-    targets: "TargetType",
-    recursive: bool = False,
-    glob: bool = False,
-) -> Iterator[str]:
-    for target in glob_targets(ensure_list(targets), glob=glob):
-        expanded_targets = _find_all_targets(repo, target, recursive=recursive)
-        for index, path in enumerate(expanded_targets):
-            if index == LARGE_DIR_SIZE:
-                msg = LARGE_DIR_RECURSIVE_ADD_WARNING.format(
-                    cyan=colorama.Fore.CYAN,
-                    nc=colorama.Style.RESET_ALL,
-                    target=target,
-                )
-                ui.error_write(msg)
-            yield path
-
-
-def _find_all_targets(
-    repo: "Repo", target: str, recursive: bool = False
-) -> Iterator[str]:
-    from dvc.dvcfile import is_dvc_file
-
-    if os.path.isdir(target) and recursive:
-        files = repo.dvcignore.find(repo.fs, target)
-        yield from (
-            path
-            for path in files
-            if not repo.is_dvc_internal(path)
-            if not is_dvc_file(path)
-            if not repo.scm.belongs_to_scm(path)
-            if not repo.scm.is_tracked(path)
-        )
-    else:
-        yield target
-
-
-def create_stages(
-    repo: "Repo",
-    targets: Iterator[str],
-    fname: Optional[str] = None,
-    transfer: bool = False,
-    external: bool = False,
-    force: bool = False,
-    **kwargs: Any,
-) -> Iterator[StageInfo]:
-    for target in targets:
-        if kwargs.get("out"):
-            target = resolve_output(target, kwargs["out"], force=force)
-        path, wdir, out = resolve_paths(
-            repo, target, always_local=transfer and not kwargs.get("out")
-        )
-
-        try:
-            (out_obj,) = repo.find_outs_by_path(target, strict=False)
-            stage = out_obj.stage
-            if not stage.is_data_source:
-                raise DvcException(f"cannot update {out!r}: not a data source")
-            output_exists = True
-        except OutputNotFoundError:
-            stage = repo.stage.create(
-                single_stage=True,
-                validate=False,
-                fname=fname or path,
-                wdir=wdir,
-                outs=[out],
-                external=external,
-                force=force,
-            )
-            output_exists = False
-
-        out_obj = stage.outs[0]
-        out_obj.annot.update(**kwargs)
-        yield StageInfo(stage, output_exists)
+        for source, (stage, output_exists) in progress_iter(stages_with_targets):
+            try:
+                _add(stage, source if output_exists else None, no_commit=no_commit)
+            except CacheLinkError:
+                link_failures.append(stage.relpath)
+    return stages

dvc/repo/brancher.py

@@ -67,6 +67,7 @@ def brancher(  # noqa: E302
 
     saved_fs = self.fs
     saved_root = self.root_dir
+    saved_dvc_dir = self.dvc_dir
 
     scm = self.scm
 
@@ -101,6 +102,8 @@ def brancher(  # noqa: E302
     finally:
         self.fs = saved_fs
         self.root_dir = saved_root
+        self.dvc_dir = saved_dvc_dir
+        self._reset()  # pylint: disable=protected-access
 
 
 def _switch_fs(
@@ -125,6 +128,8 @@ def _switch_fs(
 
     repo.fs = fs
     repo.root_dir = root_dir
+    repo.dvc_dir = fs.path.join(root_dir, repo.DVC_DIR)
+    repo._reset()  # pylint: disable=protected-access
 
     if cwd_parts:
         cwd = repo.fs.path.join("/", *cwd_parts)
@@ -149,9 +154,12 @@ def switch(repo: "Repo", rev: str) -> Iterator[str]:
 
     saved_fs = repo.fs
     saved_root = repo.root_dir
+    saved_dvc_dir = repo.dvc_dir
     try:
         _switch_fs(repo, rev, repo_root_parts, cwd_parts)
         yield rev
     finally:
         repo.fs = saved_fs
         repo.root_dir = saved_root
+        repo.dvc_dir = saved_dvc_dir
+        repo._reset()  # pylint: disable=protected-access

dvc/repo/cache.py

@@ -0,0 +1,31 @@
+import os
+
+
+def check_missing(repo, rev=None, max_size=None, types=None):
+    from dvc_data.index import StorageKeyError
+
+    with repo.switch(rev or "workspace"):
+        idx = repo.index.targets_view(
+            None,
+            max_size=max_size,
+            types=types,
+        )
+
+        index = idx.data["repo"]
+
+    def onerror(_entry, _exc):
+        pass
+
+    index.onerror = onerror
+
+    ret = []
+    for _, entry in index.iteritems():
+        try:
+            fs, path = index.storage_map.get_cache(entry)
+        except (StorageKeyError, ValueError):
+            continue
+
+        if not fs.exists(path):
+            ret.append(os.path.join(*entry.key))
+
+    return ret

dvc/repo/checkout.py

@@ -1,15 +1,21 @@
 import logging
 import os
-from typing import TYPE_CHECKING, Dict, List, Set
+from typing import TYPE_CHECKING, Dict, List
 
-from dvc.exceptions import CheckoutError, CheckoutErrorSuggestGit, NoOutputOrStageError
+from dvc.exceptions import (
+    CheckoutError,
+    CheckoutErrorSuggestGit,
+    DvcException,
+    NoOutputOrStageError,
+)
+from dvc.ui import ui
 from dvc.utils import relpath
 
 from . import locked
 
 if TYPE_CHECKING:
-    from . import Repo
-    from .stage import StageInfo
+    from dvc_data.index import BaseDataIndex, DataIndexEntry
+    from dvc_objects.fs.base import FileSystem
 
 logger = logging.getLogger(__name__)
 
@@ -30,37 +36,64 @@ def _remove_unused_links(repo):
     return ret
 
 
-def get_all_files_numbers(pairs):
-    return sum(stage.get_all_files_number(filter_info) for stage, filter_info in pairs)
+def _build_out_changes(index, changes):
+    from dvc_data.index.checkout import MODIFY
+
+    out_keys = []
+    for out in index.outs:
+        if not out.use_cache:
+            continue
+
+        ws, key = out.index_key
+        if ws != "repo":
+            continue
+
+        out_keys.append(key)
+
+    out_changes = {}
+    for key, change in changes.items():
+        for out_key in out_keys:
+            if len(out_key) > len(key) or key[: len(out_key)] != out_key:
+                continue
+
+            if key == out_key:
+                out_changes[out_key] = change.typ
+            elif not out_changes.get(out_key):
+                out_changes[out_key] = MODIFY
+            break
+
+    return out_changes
 
 
-def _collect_pairs(
-    self: "Repo", targets, with_deps: bool, recursive: bool
-) -> Set["StageInfo"]:
-    from dvc.stage.exceptions import StageFileBadNameError, StageFileDoesNotExistError
-
-    pairs: Set["StageInfo"] = set()
-    for target in targets:
+def _check_can_delete(
+    entries: List["DataIndexEntry"],
+    index: "BaseDataIndex",
+    path: str,
+    fs: "FileSystem",
+):
+    entry_paths = []
+    for entry in entries:
         try:
-            pairs.update(
-                self.stage.collect_granular(
-                    target, with_deps=with_deps, recursive=recursive
-                )
-            )
-        except (
-            StageFileDoesNotExistError,
-            StageFileBadNameError,
-            NoOutputOrStageError,
-        ) as exc:
-            if not target:
-                raise
-            raise CheckoutErrorSuggestGit(target) from exc
+            cache_fs, cache_path = index.storage_map.get_cache(entry)
+        except ValueError:
+            continue
 
-    return pairs
+        if cache_fs.exists(cache_path):
+            continue
+
+        entry_paths.append(fs.path.join(path, *entry.key))
+
+    if not entry_paths:
+        return
+
+    raise DvcException(
+        "Can't remove the following unsaved files without confirmation. "
+        "Use `--force` to force.\n" + "\n".join(entry_paths)
+    )
 
 
 @locked
-def checkout(
+def checkout(  # noqa: C901
     self,
     targets=None,
     with_deps=False,
@@ -70,13 +103,14 @@ def checkout(
     allow_missing=False,
     **kwargs,
 ):
-    from dvc.fs.callbacks import Callback
+    from dvc.repo.index import build_data_index
+    from dvc.stage.exceptions import StageFileBadNameError, StageFileDoesNotExistError
+    from dvc_data.index.checkout import ADD, DELETE, MODIFY, apply, compare
 
     stats: Dict[str, List[str]] = {
         "added": [],
         "deleted": [],
         "modified": [],
-        "failed": [],
     }
     if not targets:
         targets = [None]
@@ -85,28 +119,88 @@ def checkout(
     if isinstance(targets, str):
         targets = [targets]
 
-    pairs = _collect_pairs(self, targets, with_deps, recursive)
-    total = get_all_files_numbers(pairs)
-    with Callback.as_tqdm_callback(
+    def onerror(target, exc):
+        if target and isinstance(
+            exc,
+            (
+                StageFileDoesNotExistError,
+                StageFileBadNameError,
+                NoOutputOrStageError,
+            ),
+        ):
+            raise CheckoutErrorSuggestGit(target) from exc
+        raise  # pylint: disable=misplaced-bare-raise
+
+    view = self.index.targets_view(
+        targets,
+        recursive=recursive,
+        with_deps=with_deps,
+        onerror=onerror,
+    )
+
+    with ui.progress(
+        unit="entry",
+        desc="Building data index",
+    ) as pb:
+        old = build_data_index(
+            view,
+            self.root_dir,
+            self.fs,
+            compute_hash=True,
+            callback=pb.as_callback(),
+        )
+
+    new = view.data["repo"]
+
+    with ui.progress(
+        desc="Comparing indexes",
+        unit="entry",
+    ) as pb:
+        diff = compare(old, new, relink=relink, delete=True, callback=pb.as_callback())
+
+    if not force:
+        _check_can_delete(diff.files_delete, new, self.root_dir, self.fs)
+
+    failed = set()
+    out_paths = [out.fs_path for out in view.outs if out.use_cache and out.is_in_repo]
+
+    def checkout_onerror(src_path, dest_path, _exc):
+        logger.debug(
+            "failed to create '%s' from '%s'", dest_path, src_path, exc_info=True
+        )
+
+        for out_path in out_paths:
+            if self.fs.path.isin_or_eq(dest_path, out_path):
+                failed.add(out_path)
+
+    with ui.progress(
         unit="file",
         desc="Checkout",
-        disable=total == 0,
-    ) as cb:
-        cb.set_size(total)
-        for stage, filter_info in pairs:
-            result = stage.checkout(
-                force=force,
-                progress_callback=cb,
-                relink=relink,
-                filter_info=filter_info,
-                allow_missing=allow_missing,
-                **kwargs,
-            )
-            for key, items in result.items():
-                stats[key].extend(_fspath_dir(path) for path in items)
+    ) as pb:
+        apply(
+            diff,
+            self.root_dir,
+            self.fs,
+            callback=pb.as_callback(),
+            update_meta=False,
+            onerror=checkout_onerror,
+            state=self.state,
+            **kwargs,
+        )
 
-    if stats.get("failed"):
-        raise CheckoutError(stats["failed"], stats)
+    out_changes = _build_out_changes(view, diff.changes)
+
+    typ_map = {ADD: "added", DELETE: "deleted", MODIFY: "modified"}
+    for key, typ in out_changes.items():
+        out_path = self.fs.path.join(self.root_dir, *key)
+
+        if out_path in failed:
+            self.fs.remove(out_path, recursive=True)
+        else:
+            self.state.save_link(out_path, self.fs)
+            stats[typ_map[typ]].append(_fspath_dir(out_path))
+
+    if failed and not allow_missing:
+        raise CheckoutError([relpath(out_path) for out_path in failed], stats)
 
-    del stats["failed"]
     return stats

dvc/repo/commit.py

@@ -37,7 +37,7 @@ def prompt_to_commit(stage, changes, force=False):
 @locked
 def commit(
     self,
-    target,
+    target=None,
     with_deps=False,
     recursive=False,
     force=False,

dvc/repo/data.py

@@ -1,21 +1,12 @@
 import os
 import posixpath
-from itertools import chain
-from typing import (
-    TYPE_CHECKING,
-    Any,
-    Dict,
-    Iterable,
-    Iterator,
-    List,
-    Optional,
-    TypedDict,
-    Union,
-)
+from typing import TYPE_CHECKING, Any, Dict, Iterable, List, TypedDict, Union
 
+from dvc.fs.callbacks import DEFAULT_CALLBACK
 from dvc.ui import ui
 
 if TYPE_CHECKING:
+    from dvc.fs.callbacks import Callback
     from dvc.repo import Repo
     from dvc.scm import Git, NoSCM
     from dvc_data.index import DataIndex
@@ -61,6 +52,7 @@ def _diff(
     not_in_cache: bool = False,
     not_in_remote: bool = False,
     remote_refresh: bool = False,
+    callback: "Callback" = DEFAULT_CALLBACK,
 ) -> Dict[str, List[str]]:
     from dvc_data.index import StorageError
     from dvc_data.index.diff import UNCHANGED, UNKNOWN, diff
@@ -81,13 +73,12 @@ def _diff(
         shallow=not granular,
         hash_only=True,
         with_unknown=True,
+        callback=callback,
     ):
         if (
             change.typ == UNCHANGED
-            and change.old
-            and change.new
-            and not change.old.hash_info
-            and not change.new.hash_info
+            and (not change.old or not change.old.hash_info)
+            and (not change.new or not change.new.hash_info)
         ):
             # NOTE: emulating previous behaviour
             continue
@@ -162,16 +153,27 @@ def _git_info(scm: Union["Git", "NoSCM"], untracked_files: str = "all") -> GitIn
 def _diff_index_to_wtree(repo: "Repo", **kwargs: Any) -> Dict[str, List[str]]:
     from .index import build_data_index
 
-    with ui.status("Building workspace index"):
+    with ui.progress(
+        desc="Building workspace index",
+        unit="entry",
+    ) as pb:
         workspace = build_data_index(
-            repo.index, repo.root_dir, repo.fs, compute_hash=True
+            repo.index,
+            repo.root_dir,
+            repo.fs,
+            compute_hash=True,
+            callback=pb.as_callback(),
         )
 
-    with ui.status("Calculating diff between index/workspace"):
+    with ui.progress(
+        desc="Calculating diff between index/workspace",
+        unit="entry",
+    ) as pb:
         return _diff(
             repo.index.data["repo"],
             workspace,
             not_in_cache=True,
+            callback=pb.as_callback(),
             **kwargs,
         )
 
@@ -184,8 +186,11 @@ def _diff_head_to_index(
     with repo.switch(head):
         head_index = repo.index.data["repo"]
 
-    with ui.status("Calculating diff between head/index"):
-        return _diff(head_index, index, **kwargs)
+    with ui.progress(
+        desc="Calculating diff between head/index",
+        unit="entry",
+    ) as pb:
+        return _diff(head_index, index, callback=pb.as_callback(), **kwargs)
 
 
 class Status(TypedDict):
@@ -251,17 +256,3 @@ def status(
         unchanged=list(unchanged),
         git=git_info,
     )
-
-
-def ls(
-    repo: "Repo",
-    targets: Optional[List[Optional[str]]] = None,
-    recursive: bool = False,
-) -> Iterator[Dict[str, Any]]:
-    targets = targets or [None]
-    pairs = chain.from_iterable(
-        repo.stage.collect_granular(target, recursive=recursive) for target in targets
-    )
-    for stage, filter_info in pairs:
-        for out in stage.filter_outs(filter_info):
-            yield {"path": str(out), **out.annot.to_dict()}

dvc/repo/diff.py

@@ -156,8 +156,13 @@ def diff(
         for target in old_missing & new_missing:
             raise FileNotFoundError(errno.ENOENT, os.strerror(errno.ENOENT), target)
 
-    old = indexes[a_rev]
-    new = indexes[b_rev]
+    if len(indexes.keys()) == 1:
+        # both a_rev and b_rev point to the same sha, nothing to compare
+        old = None
+        new = None
+    else:
+        old = indexes[a_rev]
+        new = indexes[b_rev]
 
     with ui.status("Calculating diff"):
         return _diff(old, new, with_missing=with_missing)

dvc/repo/experiments/__init__.py

@@ -5,7 +5,6 @@ from typing import TYPE_CHECKING, Dict, Iterable, List, Optional
 
 from funcy import chain, first
 
-from dvc.exceptions import DvcException
 from dvc.ui import ui
 from dvc.utils import relpath
 from dvc.utils.objects import cached_property
@@ -22,7 +21,6 @@ from .refs import (
     CELERY_FAILED_STASH,
     CELERY_STASH,
     EXEC_APPLY,
-    EXEC_CHECKPOINT,
     EXEC_NAMESPACE,
     EXPS_NAMESPACE,
     WORKSPACE_STASH,
@@ -48,10 +46,7 @@ class Experiments:
         repo (dvc.repo.Repo): repo instance that these experiments belong to.
     """
 
-    BRANCH_RE = re.compile(
-        r"^(?P<baseline_rev>[a-f0-9]{7})-(?P<exp_sha>[a-f0-9]+)"
-        r"(?P<checkpoint>-checkpoint)?$"
-    )
+    BRANCH_RE = re.compile(r"^(?P<baseline_rev>[a-f0-9]{7})-(?P<exp_sha>[a-f0-9]+)")
 
     def __init__(self, repo):
         from dvc.scm import NoSCMError
@@ -138,47 +133,14 @@ class Experiments:
     def queue_one(
         self,
         queue: "BaseStashQueue",
-        checkpoint_resume: Optional[str] = None,
-        reset: bool = False,
         **kwargs,
     ) -> "QueueEntry":
         """Queue a single experiment."""
-        if reset:
-            self.reset_checkpoints()
-
-        if kwargs.pop("machine", None) is not None:
-            # TODO: decide how to handle queued remote execution
-            raise NotImplementedError
-
-        if checkpoint_resume:
-            from dvc.scm import resolve_rev
-
-            resume_rev = resolve_rev(self.scm, checkpoint_resume)
-            try:
-                self.check_baseline(resume_rev)
-                checkpoint_resume = resume_rev
-            except BaselineMismatchError as exc:
-                raise DvcException(
-                    f"Cannot resume from '{checkpoint_resume}' as it is not "
-                    "derived from your current workspace."
-                ) from exc
-        else:
-            checkpoint_resume = self._workspace_resume_rev()
-
         return self.new(
             queue,
-            checkpoint_resume=checkpoint_resume,
-            reset=reset,
             **kwargs,
         )
 
-    def _workspace_resume_rev(self) -> Optional[str]:
-        last_checkpoint = self._get_last_checkpoint()
-        last_applied = self._get_last_applied()
-        if last_checkpoint and last_applied:
-            return last_applied
-        return None
-
     def reproduce_celery(  # noqa: C901
         self, entries: Optional[Iterable["QueueEntry"]] = None, **kwargs
     ) -> Dict[str, str]:
@@ -257,17 +219,12 @@ class Experiments:
         self,
         queue: "BaseStashQueue",
         *args,
-        checkpoint_resume: Optional[str] = None,
         **kwargs,
     ) -> "QueueEntry":
         """Create and enqueue a new experiment.
 
         Experiment will be derived from the current workspace.
         """
-        if checkpoint_resume is not None:
-            return self._resume_checkpoint(
-                queue, *args, resume_rev=checkpoint_resume, **kwargs
-            )
 
         name = kwargs.get("name", None)
         baseline_sha = kwargs.get("baseline_rev") or self.repo.scm.get_rev()
@@ -281,59 +238,6 @@ class Experiments:
 
         return queue.put(*args, **kwargs)
 
-    def _resume_checkpoint(
-        self,
-        queue: "BaseStashQueue",
-        *args,
-        resume_rev: Optional[str] = None,
-        **kwargs,
-    ) -> "QueueEntry":
-        """Create and queue a resumed checkpoint experiment."""
-        assert resume_rev
-
-        branch: Optional[str] = None
-        try:
-            allow_multiple = bool(kwargs.get("params", None))
-            branch = self.get_branch_by_rev(resume_rev, allow_multiple=allow_multiple)
-            if not branch:
-                raise DvcException(
-                    f"Could not find checkpoint experiment '{resume_rev[:7]}'"
-                )
-            baseline_rev = self._get_baseline(branch)
-        except MultipleBranchError as exc:
-            baselines = {
-                info.baseline_sha for info in exc.ref_infos if info.baseline_sha
-            }
-            if len(baselines) == 1:
-                baseline_rev = baselines.pop()
-            else:
-                raise
-
-        logger.debug(
-            "Checkpoint run from '%s' with baseline '%s'",
-            resume_rev[:7],
-            baseline_rev,
-        )
-        return queue.put(
-            *args,
-            resume_rev=resume_rev,
-            baseline_rev=baseline_rev,
-            branch=branch,
-            **kwargs,
-        )
-
-    def _get_last_checkpoint(self) -> Optional[str]:
-        try:
-            last_checkpoint = self.scm.get_ref(EXEC_CHECKPOINT)
-            if last_checkpoint:
-                self.check_baseline(last_checkpoint)
-            return last_checkpoint
-        except BaselineMismatchError:
-            # If HEAD has moved since the the last checkpoint run,
-            # the specified checkpoint is no longer relevant
-            self.scm.remove_ref(EXEC_CHECKPOINT)
-        return None
-
     def _get_last_applied(self) -> Optional[str]:
         try:
             last_applied = self.scm.get_ref(EXEC_APPLY)
@@ -346,10 +250,6 @@ class Experiments:
             self.scm.remove_ref(EXEC_APPLY)
         return None
 
-    def reset_checkpoints(self):
-        self.scm.remove_ref(EXEC_CHECKPOINT)
-        self.scm.remove_ref(EXEC_APPLY)
-
     @unlocked_repo
     def _reproduce_queue(
         self,
@@ -449,17 +349,6 @@ class Experiments:
             result[rev] = name
         return result
 
-    def get_running_exps(self, fetch_refs: bool = True) -> Dict[str, Dict]:
-        """Return info for running experiments."""
-        result = {}
-        for queue in (
-            self.workspace_queue,
-            self.tempdir_queue,
-            self.celery_queue,
-        ):
-            result.update(queue.get_running_exps(fetch_refs))
-        return result
-
     def apply(self, *args, **kwargs):
         from dvc.repo.experiments.apply import apply
 
@@ -490,11 +379,6 @@ class Experiments:
 
         return save(self.repo, *args, **kwargs)
 
-    def gc(self, *args, **kwargs):
-        from dvc.repo.experiments.gc import gc
-
-        return gc(self.repo, *args, **kwargs)
-
     def push(self, *args, **kwargs):
         from dvc.repo.experiments.push import push
 

dvc/repo/experiments/brancher.py

@@ -0,0 +1,47 @@
+from contextlib import ExitStack, contextmanager
+from typing import TYPE_CHECKING, Iterator, Tuple
+
+from dvc.repo.experiments.exceptions import InvalidExpRevError
+from dvc.scm import RevError
+
+if TYPE_CHECKING:
+    from dvc.repo import Repo
+
+
+@contextmanager
+def switch_repo(
+    repo: "Repo",
+    rev: str,
+) -> Iterator[Tuple["Repo", str]]:
+    """Return a repo instance (brancher) switched to rev.
+
+    If rev is the name of a running experiment, the returned instance will be
+    the live repo wherever the experiment is running.
+
+    NOTE: This will not resolve git SHA's that only exist in queued exp workspaces
+    (it will only match queued exp names).
+    """
+    try:
+        with repo.switch(rev):
+            yield repo, rev
+        return
+    except RevError as exc:
+        orig_exc = exc
+    exps = repo.experiments
+
+    if rev == exps.workspace_queue.get_running_exp():
+        yield repo, "workspace"
+        return
+
+    for queue in (exps.tempdir_queue, exps.celery_queue):
+        try:
+            active_repo = queue.active_repo(rev)
+        except InvalidExpRevError:
+            continue
+        stack = ExitStack()
+        stack.enter_context(active_repo)
+        stack.enter_context(active_repo.switch("workspace"))
+        with stack:
+            yield active_repo, rev
+        return
+    raise orig_exc

dvc/repo/experiments/collect.py

@@ -1,6 +1,7 @@
 import itertools
 import logging
 import os
+from dataclasses import fields
 from datetime import datetime
 from typing import (
     TYPE_CHECKING,
@@ -10,6 +11,7 @@ from typing import (
     Iterator,
     List,
     Optional,
+    Tuple,
     Union,
 )
 
@@ -21,6 +23,7 @@ from dvc.scm import Git, SCMError, iter_revs
 from .exceptions import InvalidExpRefError
 from .refs import EXEC_BRANCH, ExpRefInfo
 from .serialize import ExpRange, ExpState, SerializableError, SerializableExp
+from .utils import describe
 
 if TYPE_CHECKING:
     from dvc.repo import Repo
@@ -151,7 +154,23 @@ def collect_queued(
     """
     if not baseline_revs:
         return {}
-    return repo.experiments.celery_queue.collect_queued_data(baseline_revs, **kwargs)
+    queued_data = {}
+    for rev, ranges in repo.experiments.celery_queue.collect_queued_data(
+        baseline_revs, **kwargs
+    ).items():
+        for exp_range in ranges:
+            for exp_state in exp_range.revs:
+                if exp_state.data:
+                    attrs = [f.name for f in fields(SerializableExp)]
+                    exp_state.data = SerializableExp(
+                        **{
+                            attr: getattr(exp_state.data, attr)
+                            for attr in attrs
+                            if attr != "metrics"
+                        }
+                    )
+        queued_data[rev] = ranges
+    return queued_data
 
 
 def collect_active(
@@ -290,7 +309,9 @@ def collect(
     if sha_only:
         baseline_names: Dict[str, Optional[str]] = {}
     else:
-        baseline_names = _describe(repo.scm, baseline_revs, refs=cached_refs)
+        baseline_names = describe(
+            repo.scm, baseline_revs, refs=cached_refs, logger=logger
+        )
 
     workspace_data = collect_rev(repo, "workspace", **kwargs)
     result: List["ExpState"] = [workspace_data]
@@ -319,62 +340,14 @@ def collect(
     return result
 
 
-def _describe(
-    scm: "Git",
-    revs: Iterable[str],
-    refs: Optional[Iterable[str]] = None,
-) -> Dict[str, Optional[str]]:
-    """Describe revisions using a tag, branch.
-
-    The first matching name will be returned for each rev. Names are preferred in this
-    order:
-        - current branch (if rev matches HEAD and HEAD is a branch)
-        - tags
-        - branches
-
-    Returns:
-        Dict mapping revisions from revs to a name.
-    """
-
-    head_rev = scm.get_rev()
-    head_ref = scm.get_ref("HEAD", follow=False)
-    if head_ref and head_ref.startswith("refs/heads/"):
-        head_branch = head_ref[len("refs/heads/") :]
-    else:
-        head_branch = None
-
-    tags = {}
-    branches = {}
-    ref_it = iter(refs) if refs else scm.iter_refs()
-    for ref in ref_it:
-        is_tag = ref.startswith("refs/tags/")
-        is_branch = ref.startswith("refs/heads/")
-        if not (is_tag or is_branch):
-            continue
-        rev = scm.get_ref(ref)
-        if not rev:
-            logger.debug("unresolved ref %s", ref)
-            continue
-        if is_tag and rev not in tags:
-            tags[rev] = ref[len("refs/tags/") :]
-        if is_branch and rev not in branches:
-            branches[rev] = ref[len("refs/heads/") :]
-    names: Dict[str, Optional[str]] = {}
-    for rev in revs:
-        if rev == head_rev and head_branch:
-            names[rev] = head_branch
-        else:
-            names[rev] = tags.get(rev) or branches.get(rev)
-    return names
-
-
 def _sorted_ranges(exp_ranges: Iterable["ExpRange"]) -> List["ExpRange"]:
-    """Return list of ExpRange sorted by timestamp."""
+    """Return list of ExpRange sorted by (timestamp, rev)."""
 
-    def _head_timestamp(exp_range: "ExpRange") -> datetime:
+    def _head_timestamp(exp_range: "ExpRange") -> Tuple[datetime, str]:
         head_exp = first(exp_range.revs)
         if head_exp and head_exp.data and head_exp.data.timestamp:
-            return head_exp.data.timestamp
-        return datetime.fromtimestamp(0)
+            return head_exp.data.timestamp, head_exp.rev
+
+        return datetime.fromtimestamp(0), ""
 
     return sorted(exp_ranges, key=_head_timestamp, reverse=True)