repo: cache: support some view options

Related https://github.com/iterative/studio/pull/6541

.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/ISSUE_TEMPLATE/epic_story.md

@@ -0,0 +1,60 @@
+---
+name: Epic/Story Issue Template
+about: Template for "top" level issues - Epics (>2 weeks) / Stories (<2 weeks)
+title: 'Epic: New Feature'
+labels: epic
+assignees:
+
+---
+
+## Summary / Background
+What do you want to achieve, why? business context
+...
+
+## Scope
+
+What will be impacted and what won't be?
+What needs implementation and what is invariant?
+e.g.
+- user should be able to run workflow X from UI
+- enable workflow Y, Z from CLI
+
+## Assumptions
+Product / UX assumptions as well as technical assumptions / limitations
+e.g.
+* Support only Python Runtime
+* Focus on DVC experiments only
+* Deployment environments don't change often and can be picked up from shared configuration
+
+## Open Questions
+e.g.
+- How should access control work for shared artifacts (workflow X)
+- Python runtime assumption - is it really valid? in light of <...>
+
+## Blockers / Dependencies
+List issues or other conditions / blockers
+
+## General Approach
+Invocation example:
+```shell
+$ mapper-run task.tar.gz --ray-cluster <ip>:<port>
+```
+
+## Steps
+
+### Must have (p1)
+- [ ] subissue2
+- [ ] step 2
+ - info
+ - info
+
+### Optional / followup (p2)
+- [ ] ⌛ step 3 wip
+- [ ] step 4
+
+## Timelines
+
+Put your estimations here. Update once certainty changes
+- end of week (Feb 3) for prototype with workflows X, Y
+- Feb 15 - MVP in production
+- Low priority followups can be done later

.github/mergify.yml

@@ -0,0 +1,10 @@
+pull_request_rules:
+  - name: backport patches to 2.x
+    conditions:
+      - base=main
+      - label=backport
+    actions:
+      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/codeql.yml

@@ -2,7 +2,7 @@ name: "CodeQL"
 
 on:
   push:
-    branches: ["main"]
+    branches: ["main", "2.x"]
   pull_request:
     branches: ["main"]
   schedule:

.github/workflows/packages.yaml

@@ -1,112 +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-18.04", asset: "dvc_${{ github.event.release.tag_name }}_amd64.deb"}
-         - {pkg: "rpm", os: "ubuntu-18.04", asset: "dvc-${{ github.event.release.tag_name }}-1.x86_64.rpm"}
-         - {pkg: "osxpkg", os: "macos-10.15", 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
-    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
-      with:
-        password: ${{ secrets.PYPI_TOKEN }}
-
-    - 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:
-        password: ${{ secrets.TEST_PYPI_TOKEN }}
-        repository_url: https://test.pypi.org/legacy/
-        skip_existing: true

.github/workflows/tests.yaml

@@ -2,7 +2,7 @@ name: Tests
 
 on:
   push:
-    branches: [main]
+    branches: [main, 2.x]
   pull_request:
   schedule:
     - cron: '5 1 * * *'  # every day at 01:05
@@ -17,7 +17,7 @@ permissions:
 
 jobs:
   tests:
-    timeout-minutes: 45
+    timeout-minutes: 50
     runs-on: ${{ matrix.os }}
     strategy:
       fail-fast: false
@@ -41,9 +41,11 @@ jobs:
         pip install -e ".[dev]"
     - uses: pre-commit/[email protected]
     - name: run tests
-      timeout-minutes: 30
+      timeout-minutes: 40
+      env:
+        PYTHONUTF8: 1
       run: >
-        pytest -nauto --timeout=300 --durations=100
+        pytest -n=logical --timeout=300 --durations=100
         --cov --cov-report=xml --cov-report=term
     - name: upload coverage report
       uses: codecov/codecov-action@v3

.pre-commit-config.yaml

@@ -23,44 +23,38 @@ 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: 23.1.0
+    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: 'v0.0.238'
+  - 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: v2.2.2
+    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
+    rev: '1.7.5'
+    hooks:
+    - id: bandit
+      args: ["-c", "pyproject.toml", "--quiet", "--format=custom"]
+      files: ^dvc/
+      additional_dependencies: [".[toml]"]
   - repo: local
     hooks:
     - id: mypy

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

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/analytics.py

@@ -47,7 +47,7 @@ def is_enabled():
     enabled = not os.getenv(DVC_NO_ANALYTICS)
     if enabled:
         enabled = to_bool(
-            Config(validate=False).get("core", {}).get("analytics", "true")
+            Config.from_cwd(validate=False).get("core", {}).get("analytics", "true")
         )
 
     logger.debug("Analytics is %sabled.", "en" if enabled else "dis")

dvc/annotations.py

@@ -2,6 +2,7 @@ from dataclasses import asdict, dataclass, field, fields
 from typing import Any, ClassVar, Dict, List, Optional
 
 from funcy import compact
+from voluptuous import Required
 
 
 @dataclass
@@ -16,11 +17,23 @@ 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))
+
+
+@dataclass
+class Artifact:
+    PARAM_PATH: ClassVar[str] = "path"
+    PARAM_DESC: ClassVar[str] = "desc"
+    PARAM_TYPE: ClassVar[str] = "type"
+    PARAM_LABELS: ClassVar[str] = "labels"
+    PARAM_META: ClassVar[str] = "meta"
+
+    path: str
+    desc: Optional[str] = None
+    type: Optional[str] = None  # noqa: A003
+    labels: List[str] = field(default_factory=list)
+    meta: Dict[str, Any] = field(default_factory=dict)
 
     def to_dict(self) -> Dict[str, str]:
         return compact(asdict(self))
@@ -33,3 +46,7 @@ ANNOTATION_SCHEMA = {
     Annotation.PARAM_LABELS: [str],
     Annotation.PARAM_META: object,
 }
+ARTIFACT_SCHEMA = {
+    Required(Artifact.PARAM_PATH): str,
+    **ANNOTATION_SCHEMA,  # type: ignore[arg-type]
+}

dvc/api/__init__.py

@@ -2,14 +2,20 @@ 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 make_checkpoint
-from .params import params_show
+from .experiments import exp_save, exp_show
+from .scm import all_branches, all_commits, all_tags
+from .show import metrics_show, params_show
 
 __all__ = [
+    "all_branches",
+    "all_commits",
+    "all_tags",
+    "exp_save",
+    "exp_show",
     "get_url",
-    "make_checkpoint",
     "open",
     "params_show",
+    "metrics_show",
     "read",
     "DVCFileSystem",
 ]

dvc/api/data.py

@@ -1,12 +1,30 @@
 from contextlib import _GeneratorContextManager as GCM
-from typing import Optional
+from contextlib import contextmanager
+from typing import Any, Dict, Optional
 
 from funcy import reraise
 
-from dvc.exceptions import OutputNotFoundError, PathMissingError
+from dvc.exceptions import FileMissingError, OutputNotFoundError, PathMissingError
 from dvc.repo import Repo
 
 
+@contextmanager
+def _wrap_exceptions(repo, url):
+    from dvc.config import NoRemoteError
+    from dvc.exceptions import NoOutputInExternalRepoError, NoRemoteInExternalRepoError
+
+    try:
+        yield
+    except NoRemoteError as exc:
+        raise NoRemoteInExternalRepoError(url) from exc
+    except OutputNotFoundError as exc:
+        if exc.repo is repo:
+            raise NoOutputInExternalRepoError(exc.output, repo.root_dir, url) from exc
+        raise
+    except FileMissingError as exc:
+        raise PathMissingError(exc.path, url) from exc
+
+
 def get_url(path, repo=None, rev=None, remote=None):
     """
     Returns the URL to the storage location of a data file or directory tracked
@@ -18,19 +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:
-        fs_path = _repo.dvcfs.from_os_path(path)
-        with reraise(FileNotFoundError, PathMissingError(path, repo)):
-            info = _repo.dvcfs.info(fs_path)
+    from dvc.config import NoRemoteError
+    from dvc_data.index import StorageKeyError
 
-        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):
@@ -46,13 +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.
@@ -94,6 +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
@@ -189,25 +217,86 @@ 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):
-    with Repo.open(repo, rev=rev, subrepos=True, uninitialized=True) as _repo:
-        with _repo.open_by_relpath(
-            path, remote=remote, mode=mode, encoding=encoding
-        ) as fd:
-            yield fd
+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):
+            import os
+            from typing import TYPE_CHECKING, Union
+
+            from dvc.exceptions import IsADirectoryError as DvcIsADirectoryError
+            from dvc.fs.data import DataFileSystem
+            from dvc.fs.dvc import DVCFileSystem
+
+            if TYPE_CHECKING:
+                from dvc.fs import FileSystem
+
+            fs: Union["FileSystem", DataFileSystem, DVCFileSystem]
+            if os.path.isabs(path):
+                fs = DataFileSystem(index=_repo.index.data["local"])
+                fs_path = path
+            else:
+                fs = DVCFileSystem(repo=_repo, subrepos=True)
+                fs_path = fs.from_os_path(path)
+
+            try:
+                with fs.open(
+                    fs_path,
+                    mode=mode,
+                    encoding=encoding,
+                ) as fobj:
+                    yield fobj
+            except FileNotFoundError as exc:
+                raise FileMissingError(path) from exc
+            except IsADirectoryError as exc:
+                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

@@ -1,33 +1,120 @@
-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.stage.monitor import CheckpointTask
+from dvc.repo.experiments.show import tabulate
 
 
-def make_checkpoint():
+def exp_save(
+    name: Optional[str] = None,
+    force: bool = False,
+    include_untracked: Optional[List[str]] = None,
+):
     """
-    Signal DVC to create a checkpoint experiment.
+    Create a new DVC experiment using `exp save`.
 
-    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.
+    See https://dvc.org/doc/command-reference/exp/save.
+
+    Args:
+        name (str, optional): specify a name for this experiment.
+            If `None`, a default one will be generated, such as `urban-sign`.
+            Defaults to `None`.
+        force (bool):  overwrite the experiment if an experiment with the same
+            name already exists.
+            Defaults to `False`.
+        include_untracked (List[str], optional): specify untracked file(s) to
+            be included in the saved experiment.
+            Defaults to `None`.
+
+    Returns:
+        str: The `Git revision`_ of the created experiment.
+
+    Raises:
+        ExperimentExistsError: If an experiment with `name` already exists and
+            `force=False`.
+
+    .. _Git revision:
+        https://git-scm.com/docs/revisions
     """
-    if os.getenv(DVC_CHECKPOINT) is None:
-        return
+    with Repo() as repo:
+        return repo.experiments.save(
+            name=name, force=force, include_untracked=include_untracked
+        )
 
-    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 _postprocess(exp_rows):
+    for exp_row in exp_rows:
+        for k, v in exp_row.items():
+            if isinstance(v, Text):
+                v_str = str(v)
+                try:
+                    exp_row[k] = float(v_str)
+                except ValueError:
+                    exp_row[k] = v_str
+
+            if not exp_row[k]:
+                exp_row[k] = None
+
+    return exp_rows
+
+
+def exp_show(
+    repo: Optional[str] = None,
+    revs: Optional[Union[str, List[str]]] = None,
+    num: int = 1,
+    param_deps: bool = False,
+    force: bool = False,
+    config: Optional[Dict] = None,
+) -> List[Dict]:
+    """Get DVC experiments tracked in `repo`.
+
+    Without arguments, this function will retrieve all experiments derived from
+    the Git `HEAD`.
+
+    See the options below to customize the experiments retrieved.
+
+    Args:
+        repo (str, optional): location of the DVC repository.
+            Defaults to the current project (found by walking up from the
+            current working directory tree).
+            It can be a URL or a file system path.
+            Both HTTP and SSH protocols are supported for online Git repos
+            (e.g. [user@]server:project.git).
+        revs (Union[str, List[str]], optional): Git revision(s) (e.g. branch,
+            tag, SHA commit) to use as a reference point to start listing
+            experiments.
+            Defaults to `None`, which will use `HEAD` as starting point.
+        num (int, optional): show experiments from the last `num` commits
+            (first parents) starting from the `revs` baseline.
+            Give a negative value to include all first-parent commits (similar
+            to `git log -n`).
+            Defaults to 1.
+        param_deps (bool, optional): include only parameters that are stage
+            dependencies.
+            Defaults to `False`.
+        force (bool, optional): force re-collection of experiments instead of
+            loading from internal experiments cache.
+            DVC caches `exp_show` data for completed experiments to improve
+            performance of subsequent calls.
+            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, config=config) as _repo:
+        experiments = _repo.experiments.show(
+            revs=revs,
+            num=num,
+            param_deps=param_deps,
+            force=force,
+        )
+        td, _ = tabulate(experiments, fill_value=None)
+
+        return _postprocess(td.as_dict())

dvc/api/scm.py

@@ -0,0 +1,54 @@
+from typing import List, Optional
+
+from dvc.repo import Repo
+
+
+def all_branches(repo: Optional[str] = None) -> List[str]:
+    """Get all Git branches in a DVC repository.
+
+    Args:
+        repo (str, optional): location of the DVC repository.
+            Defaults to the current project (found by walking up from the
+            current working directory tree).
+            It can be a URL or a file system path.
+            Both HTTP and SSH protocols are supported for online Git repos
+            (e.g. [user@]server:project.git).
+    Returns:
+        List[str]: Names of the Git branches.
+    """
+    with Repo.open(repo) as _repo:
+        return _repo.scm.list_branches()
+
+
+def all_commits(repo: Optional[str] = None) -> List[str]:
+    """Get all Git commits in a DVC repository.
+
+    Args:
+        repo (str, optional): location of the DVC repository.
+            Defaults to the current project (found by walking up from the
+            current working directory tree).
+            It can be a URL or a file system path.
+            Both HTTP and SSH protocols are supported for online Git repos
+            (e.g. [user@]server:project.git).
+    Returns:
+        List[str]: SHAs of the Git commits.
+    """
+    with Repo.open(repo) as _repo:
+        return _repo.scm.list_all_commits()
+
+
+def all_tags(repo: Optional[str] = None) -> List[str]:
+    """Get all Git tags in a DVC repository.
+
+    Args:
+        repo (str, optional): location of the DVC repository.
+            Defaults to the current project (found by walking up from the
+            current working directory tree).
+            It can be a URL or a file system path.
+            Both HTTP and SSH protocols are supported for online Git repos
+            (e.g. [user@]server:project.git).
+    Returns:
+        List[str]: Names of the Git tags.
+    """
+    with Repo.open(repo) as _repo:
+        return _repo.scm.list_tags()

dvc/api/show.py

@@ -4,16 +4,163 @@ from typing import Dict, Iterable, Optional, Union
 
 from funcy import first
 
-from dvc.exceptions import DvcException
 from dvc.repo import Repo
 
 
+def _onerror_raise(exception: Exception, *args, **kwargs):
+    raise exception
+
+
+def _postprocess(results):
+    processed: Dict[str, Dict] = {}
+    for rev, rev_data in results.items():
+        if not rev_data:
+            continue
+
+        processed[rev] = {}
+
+        counts: typing.Counter[str] = Counter()
+        for file_data in rev_data["data"].values():
+            for k in file_data["data"]:
+                counts[k] += 1
+
+        for file_name, file_data in rev_data["data"].items():
+            to_merge = {
+                (k if counts[k] == 1 else f"{file_name}:{k}"): v
+                for k, v in file_data["data"].items()
+            }
+            processed[rev] = {**processed[rev], **to_merge}
+
+    if "workspace" in processed:
+        del processed["workspace"]
+
+    return processed
+
+
+def metrics_show(
+    *targets: str,
+    repo: Optional[str] = None,
+    rev: Optional[str] = None,
+    config: Optional[Dict] = None,
+) -> Dict:
+    """Get metrics tracked in `repo`.
+
+    Without arguments, this function will retrieve all metrics from all tracked
+    metric files, for the current working tree.
+
+    See the options below to restrict the metrics retrieved.
+
+    Args:
+        *targets (str, optional): Names of the metric files to retrieve
+        metrics from. For example, "classifier_eval.json,
+        clustering_eval.json".
+        If no `targets` are provided, all metric files tracked in `dvc.yaml`
+        will be used.
+        Note that targets don't necessarily have to be defined in `dvc.yaml`.
+        repo (str, optional): Location of the DVC repository.
+            Defaults to the current project (found by walking up from the
+            current working directory tree).
+            It can be a URL or a file system path.
+            Both HTTP and SSH protocols are supported for online Git repos
+            (e.g. [user@]server:project.git).
+        rev (str, optional): Name of the `Git revision`_ to retrieve metrics
+            from.
+            Defaults to `None`.
+            An example of git revision can be a branch or tag name, a commit
+            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.
+
+    Examples:
+
+        - No arguments.
+
+        Working on https://github.com/iterative/example-get-started
+
+        >>> import dvc.api
+        >>> metrics = dvc.api.metrics_show()
+        >>> print(json.dumps(metrics, indent=4))
+        {
+            "avg_prec": 0.9249974999612706,
+            "roc_auc": 0.9460213440787918
+        }
+
+        ---
+
+        - Using `rev`.
+
+        Working on https://github.com/iterative/example-get-started
+
+        >>> import json
+        >>> import dvc.api
+        >>> metrics = dvc.api.metrics_show(rev="tune-hyperparams")
+        >>> print(json.dumps(metrics, indent=4))
+        {
+            "avg_prec": 0.9268792615819422,
+            "roc_auc": 0.945093365854111
+        }
+
+        ---
+
+        - Using `targets`.
+
+        Working on https://github.com/iterative/example-get-started
+
+        >>> import json
+        >>> import dvc.api
+        >>> metrics = dvc.api.metrics_show("evaluation.json")
+        >>> print(json.dumps(metrics, indent=4))
+        {
+            "avg_prec": 0.9249974999612706,
+            "roc_auc": 0.9460213440787918
+        }
+
+        ---
+
+        - Git URL as `repo`.
+
+        >>> import json
+        >>> import dvc.api
+        >>> metrics = dvc.api.metrics_show(
+        ...     repo="https://github.com/iterative/demo-fashion-mnist")
+        >>> print(json.dumps(metrics, indent=4))
+        {
+            "loss": 0.25284987688064575,
+            "accuracy": 0.9071000218391418
+        }
+
+
+    .. _Git revision:
+        https://git-scm.com/docs/revisions
+    """
+
+    with Repo.open(repo, config=config) as _repo:
+        metrics = _repo.metrics.show(
+            targets=targets,
+            revs=rev if rev is None else [rev],
+            onerror=_onerror_raise,
+        )
+
+    metrics = _postprocess(metrics)
+
+    if not metrics:
+        return {}
+
+    return metrics[first(metrics)]
+
+
 def params_show(
     *targets: str,
     repo: Optional[str] = None,
     stages: Optional[Union[str, Iterable[str]]] = None,
     rev: Optional[str] = None,
     deps: bool = False,
+    config: Optional[Dict] = None,
 ) -> Dict:
     """Get parameters tracked in `repo`.
 
@@ -52,13 +199,12 @@ 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.
 
-    Raises:
-        DvcException: If no params are found in `repo`.
-
     Examples:
 
         - No arguments.
@@ -239,35 +385,7 @@ def params_show(
     if isinstance(stages, str):
         stages = [stages]
 
-    def _onerror_raise(_result: Dict, exception: Exception, *args, **kwargs):
-        raise exception
-
-    def _postprocess(params):
-        processed: Dict[str, Dict] = {}
-        for rev, rev_data in params.items():
-            processed[rev] = {}
-
-            counts: typing.Counter[str] = Counter()
-            for file_data in rev_data["data"].values():
-                for k in file_data["data"]:
-                    counts[k] += 1
-
-            for file_name, file_data in rev_data["data"].items():
-                to_merge = {
-                    (k if counts[k] == 1 else f"{file_name}:{k}"): v
-                    for k, v in file_data["data"].items()
-                }
-                processed[rev] = {**processed[rev], **to_merge}
-
-        if "workspace" in processed:
-            del processed["workspace"]
-
-        if not processed:
-            raise DvcException("No params found")
-
-        return processed[first(processed)]
-
-    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,
@@ -276,4 +394,9 @@ def params_show(
             stages=stages,
         )
 
-    return _postprocess(params)
+    params = _postprocess(params)
+
+    if not params:
+        return {}
+
+    return params[first(params)]

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

@@ -0,0 +1,138 @@
+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,
+    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.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"
+    FILES_DIR = "files"
+
+    def __init__(self, repo):
+        self._repo = repo
+        self.config = config = repo.config["cache"]
+        self._odb = {}
+
+        default = None
+        if repo and repo.local_dvc_dir:
+            default = os.path.join(repo.local_dvc_dir, self.CACHE_DIR)
+
+        local = config.get("local")
+
+        if local:
+            settings = {"name": local}
+        elif "dir" not in config and not default:
+            settings = None
+        else:
+            from dvc.config_schema import LOCAL_COMMON
+
+            url = config.get("dir") or default
+            settings = {"url": url}
+            for opt in LOCAL_COMMON:
+                if opt in config:
+                    settings[str(opt)] = config.get(opt)
+
+        kwargs = {}
+        if not isinstance(repo.fs, GitFileSystem):
+            kwargs["fs"] = repo.fs
+
+        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,
+                prefix=(self.FILES_DIR, DEFAULT_ALGORITHM),
+            )
+
+    def __getattr__(self, name):
+        try:
+            return self._odb[name]
+        except KeyError as exc:
+            raise AttributeError from exc
+
+    def by_scheme(self):
+        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

@@ -39,12 +39,17 @@ def parse_args(argv=None):
 
 def _log_unknown_exceptions() -> None:
     from dvc.info import get_dvc_info
-    from dvc.logger import FOOTER
+    from dvc.ui import ui
+    from dvc.utils import colorize
 
     logger.exception("unexpected error")
     if logger.isEnabledFor(logging.DEBUG):
         logger.debug("Version info for developers:\n%s", get_dvc_info())
-    logger.info(FOOTER)
+
+    q = colorize("Having any troubles?", "yellow")
+    link = colorize("https://dvc.org/support", "blue")
+    footer = f"\n{q} Hit us up at {link}, we are always happy to help!"
+    ui.error_write(footer)
 
 
 def _log_exceptions(exc: Exception) -> Optional[int]:
@@ -70,7 +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 = {
@@ -78,9 +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"
@@ -141,7 +146,7 @@ def _log_exceptions(exc: Exception) -> Optional[int]:
     return None
 
 
-def main(argv=None):  # noqa: C901
+def main(argv=None):  # noqa: C901, PLR0912, PLR0915
     """Main entry point for dvc CLI.
 
     Args:
@@ -153,7 +158,7 @@ def main(argv=None):  # noqa: C901
     from dvc._debug import debugtools
     from dvc.config import ConfigError
     from dvc.exceptions import DvcException, NotDvcRepoError
-    from dvc.logger import disable_other_loggers, set_loggers_level
+    from dvc.logger import set_loggers_level
 
     # NOTE: stderr/stdout may be closed if we are running from dvc.daemon.
     # On Linux we directly call cli.main after double forking and closing
@@ -166,9 +171,8 @@ def main(argv=None):  # noqa: C901
         logging.disable(logging.INFO)
 
     args = None
-    disable_other_loggers()
 
-    outerLogLevel = logger.level
+    outer_log_level = logger.level
     try:
         args = parse_args(argv)
 
@@ -186,10 +190,9 @@ def main(argv=None):  # noqa: C901
         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))
@@ -227,7 +230,7 @@ def main(argv=None):  # noqa: C901
         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
 
@@ -239,9 +242,9 @@ def main(argv=None):  # noqa: C901
 
         return ret
     finally:
-        logger.setLevel(outerLogLevel)
+        logger.setLevel(outer_log_level)
 
-        from dvc.external_repo import clean_repos
+        from dvc.repo.open_repo import clean_repos
 
         # Remove cached repos in the end of the call, these are anonymous
         # so won't be reused by any other subsequent run anyway.

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

@@ -2,6 +2,7 @@
 import argparse
 import logging
 import os
+from functools import lru_cache
 
 from dvc import __version__
 from dvc.commands import (
@@ -40,7 +41,6 @@ from dvc.commands import (
     remove,
     repro,
     root,
-    run,
     stage,
     unprotect,
     update,
@@ -61,7 +61,6 @@ COMMANDS = [
     remove,
     move,
     unprotect,
-    run,
     repro,
     data_sync,
     gc,
@@ -138,7 +137,6 @@ 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."
@@ -146,10 +144,12 @@ 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
 
 
+@lru_cache(maxsize=1)
 def get_main_parser():
     parent_parser = get_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

@@ -2,68 +2,61 @@ 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:
+            logger.exception("")
+            return 1
         except DvcException:
             logger.exception("")
             return 1
@@ -80,36 +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",
@@ -129,8 +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'. "
@@ -140,8 +114,13 @@ def add_parser(subparsers, parent_parser):
         ),
         metavar="<number>",
     )
-
-    _add_annotating_args(parser)
+    parser.add_argument(
+        "-f",
+        "--force",
+        action="store_true",
+        default=False,
+        help="Override local file or folder if exists.",
+    )
     parser.add_argument(
         "targets", nargs="+", help="Input files/directories to add."
     ).complete = completion.FILE

dvc/commands/cache.py

@@ -1,6 +1,8 @@
 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
@@ -9,13 +11,21 @@ from dvc.ui import ui
 class CmdCacheDir(CmdConfig):
     def run(self):
         if self.args.value is None and not self.args.unset:
+            from dvc.config import ConfigError
+
             if self.args.level:
                 conf = self.config.read(level=self.args.level)
             else:
                 # Use merged config with default values
                 conf = self.config
-            self._check(conf, False, "cache", "dir")
-            ui.write(conf["cache"]["dir"])
+            try:
+                self._check(conf, False, "cache", "dir")
+                path = conf["cache"]["dir"]
+            except ConfigError:
+                if not self.config.dvc_dir or self.args.level:
+                    raise
+                path = os.path.join(self.config.dvc_dir, "cache")
+            ui.write(path)
             return 0
         with self.config.edit(level=self.args.level) as conf:
             if self.args.unset:
@@ -27,6 +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
 
@@ -45,8 +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]
     )
@@ -77,3 +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/check_ignore.py

@@ -39,7 +39,7 @@ class CmdCheckIgnore(CmdBase):
                 target = input()
             except (KeyboardInterrupt, EOFError):
                 break
-            if target == "":
+            if not target:
                 break
             if not self._check_one_file(target):
                 ret = 0

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

@@ -9,13 +9,16 @@ from dvc.ui import ui
 logger = logging.getLogger(__name__)
 
 
+SUPPORTED_SHELLS = ["bash", "zsh"]
+
+
 class CmdCompletion(CmdBaseNoRepo):
     def run(self):
         import shtab
 
         shell = self.args.shell
         parser = self.args.parser
-        script = shtab.complete(parser, shell=shell, preamble=PREAMBLE)
+        script = shtab.complete(parser, shell=shell, preamble=PREAMBLE)  # nosec B604
         ui.write(script, force=True)
         return 0
 
@@ -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/config.py

@@ -32,7 +32,7 @@ class CmdConfig(CmdBaseNoRepo):
 
         super().__init__(args)
 
-        self.config = Config(validate=False)
+        self.config = Config.from_cwd(validate=False)
 
     def run(self):
         if self.args.show_origin and (self.args.value or self.args.unset):

dvc/commands/dag.py

@@ -9,11 +9,11 @@ if TYPE_CHECKING:
     from networkx import DiGraph
 
 
-def _show_ascii(G: "DiGraph"):
+def _show_ascii(graph: "DiGraph"):
     from dvc.dagascii import draw
     from dvc.repo.graph import get_pipelines
 
-    pipelines = get_pipelines(G)
+    pipelines = get_pipelines(graph)
 
     ret = []
     for pipeline in pipelines:
@@ -31,7 +31,7 @@ def _quote_label(node):
     return label
 
 
-def _show_dot(G: "DiGraph"):
+def _show_dot(graph: "DiGraph"):
     import io
 
     import networkx as nx
@@ -39,15 +39,15 @@ def _show_dot(G: "DiGraph"):
 
     dot_file = io.StringIO()
 
-    nx.relabel_nodes(G, _quote_label, copy=False)
-    write_dot(G.reverse(), dot_file)
+    nx.relabel_nodes(graph, _quote_label, copy=False)
+    write_dot(graph.reverse(), dot_file)
     return dot_file.getvalue()
 
 
-def _show_mermaid(G, markdown: bool = False):
+def _show_mermaid(graph, markdown: bool = False):
     from dvc.repo.graph import get_pipelines
 
-    pipelines = get_pipelines(G)
+    pipelines = get_pipelines(graph)
 
     graph = "flowchart TD"
 
@@ -100,43 +100,42 @@ def _transform(index, outs):
     def _relabel(node) -> str:
         return node.addressing if isinstance(node, Stage) else str(node)
 
-    G = index.outs_graph if outs else index.graph
-    return nx.relabel_nodes(G, _relabel, copy=True)
+    graph = index.outs_graph if outs else index.graph
+    return nx.relabel_nodes(graph, _relabel, copy=True)
 
 
-def _filter(G, targets, full):
+def _filter(graph, targets, full):
     import networkx as nx
 
     if not targets:
-        return G
+        return graph
 
-    H = G.copy()
+    new_graph = graph.copy()
     if not full:
         descendants = set()
         for target in targets:
-            descendants.update(nx.descendants(G, target))
+            descendants.update(nx.descendants(graph, target))
             descendants.add(target)
-        H.remove_nodes_from(set(G.nodes()) - descendants)
+        new_graph.remove_nodes_from(set(graph.nodes()) - descendants)
 
-    undirected = H.to_undirected()
+    undirected = new_graph.to_undirected()
     connected = set()
     for target in targets:
         connected.update(nx.node_connected_component(undirected, target))
 
-    H.remove_nodes_from(set(H.nodes()) - connected)
-
-    return H
+    new_graph.remove_nodes_from(set(new_graph.nodes()) - connected)
+    return new_graph
 
 
 def _build(repo, target=None, full=False, outs=False):
     targets = _collect_targets(repo, target, outs)
-    G = _transform(repo.index, outs)
-    return _filter(G, targets, full)
+    graph = _transform(repo.index, outs)
+    return _filter(graph, targets, full)
 
 
 class CmdDAG(CmdBase):
     def run(self):
-        G = _build(
+        graph = _build(
             self.repo,
             target=self.args.target,
             full=self.args.full,
@@ -144,12 +143,12 @@ class CmdDAG(CmdBase):
         )
 
         if self.args.dot:
-            ui.write(_show_dot(G))
+            ui.write(_show_dot(graph))
         elif self.args.mermaid or self.args.markdown:
-            ui.write(_show_mermaid(G, self.args.markdown))
+            ui.write(_show_mermaid(graph, self.args.markdown))
         else:
             with ui.pager():
-                ui.write(_show_ascii(G))
+                ui.write(_show_ascii(graph))
 
         return 0
 
@@ -177,7 +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
 
@@ -21,12 +18,14 @@ logger = logging.getLogger(__name__)
 
 class CmdDataStatus(CmdBase):
     COLORS = {
+        "not_in_remote": "red",
         "not_in_cache": "red",
         "committed": "green",
         "uncommitted": "yellow",
         "untracked": "cyan",
     }
     LABELS = {
+        "not_in_remote": "Not in remote",
         "not_in_cache": "Not in cache",
         "committed": "DVC committed changes",
         "uncommitted": "DVC uncommitted changes",
@@ -34,6 +33,7 @@ class CmdDataStatus(CmdBase):
         "unchanged": "DVC unchanged files",
     }
     HINTS = {
+        "not_in_remote": ('use "dvc push <file>..." to upload files',),
         "not_in_cache": ('use "dvc fetch <file>..." to download files',),
         "committed": ("git commit the corresponding dvc files to update the repo",),
         "uncommitted": (
@@ -90,9 +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
 
@@ -114,6 +112,8 @@ class CmdDataStatus(CmdBase):
             status = self.repo.data_status(
                 granular=self.args.granular,
                 untracked_files=self.args.untracked_files,
+                not_in_remote=self.args.not_in_remote,
+                remote_refresh=self.args.remote_refresh,
             )
 
         if not self.args.unchanged:
@@ -127,63 +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",
@@ -212,13 +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",
@@ -239,52 +175,16 @@ def add_parser(subparsers, parent_parser):
         nargs="?",
         help="Show untracked files.",
     )
+    data_status_parser.add_argument(
+        "--not-in-remote",
+        action="store_true",
+        default=False,
+        help="Show files not in remote.",
+    )
+    data_status_parser.add_argument(
+        "--no-remote-refresh",
+        dest="remote_refresh",
+        action="store_false",
+        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

@@ -40,6 +40,7 @@ class CmdDataPull(CmdDataBase):
                 recursive=self.args.recursive,
                 run_cache=self.args.run_cache,
                 glob=self.args.glob,
+                allow_missing=self.args.allow_missing,
             )
             self.log_summary(stats)
         except (CheckoutError, DvcException) as exc:
@@ -90,6 +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:
@@ -197,6 +200,12 @@ def add_parser(subparsers, _parent_parser):
         default=False,
         help=argparse.SUPPRESS,
     )
+    pull_parser.add_argument(
+        "--allow-missing",
+        action="store_true",
+        default=False,
+        help="Ignore errors if some of the files or directories are missing.",
+    )
     pull_parser.set_defaults(func=CmdDataPull)
 
     # Push
@@ -315,6 +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
@@ -396,7 +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

@@ -4,10 +4,9 @@ from dvc.cli.utils import append_doc_link, fix_subparsers, hide_subparsers_from_
 from dvc.commands.experiments import (
     apply,
     branch,
+    clean,
     diff,
     exec_run,
-    gc,
-    init,
     ls,
     pull,
     push,
@@ -21,10 +20,9 @@ from dvc.commands.experiments import (
 SUB_COMMANDS = [
     apply,
     branch,
+    clean,
     diff,
     exec_run,
-    gc,
-    init,
     ls,
     pull,
     push,
@@ -80,6 +78,7 @@ def add_rev_selection_flags(
     experiments_subcmd_parser.add_argument(
         "--rev",
         type=str,
+        action="append",
         default=None,
         help=msg,
         metavar="<commit>",

dvc/commands/experiments/apply.py

@@ -4,13 +4,21 @@ import logging
 from dvc.cli import completion
 from dvc.cli.command import CmdBase
 from dvc.cli.utils import append_doc_link
+from dvc.ui import ui
 
 logger = logging.getLogger(__name__)
 
 
 class CmdExperimentsApply(CmdBase):
     def run(self):
-        self.repo.experiments.apply(self.args.experiment, force=self.args.force)
+        if not self.args.force:
+            ui.write(
+                "The --no-force option is deprecated and will be removed in a future"
+                " DVC release. To revert the result of 'exp apply', run:\n"
+                "\n\tgit reset --hard\n"
+                "\tgit stash apply refs/exps/apply/stash\n"
+            )
+        self.repo.experiments.apply(self.args.experiment)
 
         return 0
 

dvc/commands/experiments/branch.py

@@ -26,5 +26,13 @@ def add_parser(experiments_subparsers, parent_parser):
     experiments_branch_parser.add_argument(
         "experiment", help="Experiment to be promoted."
     )
-    experiments_branch_parser.add_argument("branch", help="Name for the new Git branch")
+    experiments_branch_parser.add_argument(
+        "branch",
+        nargs="?",
+        default=None,
+        help=(
+            "Optional name for the new Git branch. "
+            "Defaults to '{experiment-name}-branch'."
+        ),
+    )
     experiments_branch_parser.set_defaults(func=CmdExperimentsBranch)

dvc/commands/experiments/clean.py

@@ -0,0 +1,25 @@
+import argparse
+import logging
+
+from dvc.cli.command import CmdBase
+from dvc.cli.utils import append_doc_link
+
+logger = logging.getLogger(__name__)
+
+
+class CmdExperimentsClean(CmdBase):
+    def run(self):
+        self.repo.experiments.clean()
+        return 0
+
+
+def add_parser(experiments_subparsers, parent_parser):
+    EXPERIMENTS_CLEAN_HELP = "Cleanup experiments temporary internal files."
+    experiments_clean_parser = experiments_subparsers.add_parser(
+        "clean",
+        parents=[parent_parser],
+        description=append_doc_link(EXPERIMENTS_CLEAN_HELP, "exp/clean"),
+        help=EXPERIMENTS_CLEAN_HELP,
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+    )
+    experiments_clean_parser.set_defaults(func=CmdExperimentsClean)

dvc/commands/experiments/diff.py

@@ -52,7 +52,7 @@ class CmdExperimentsDiff(CmdBase):
 
 
 def add_parser(experiments_subparsers, parent_parser):
-    EXPERIMENTS_DIFF_HELP = "Show changes between experiments in the DVC repository."
+    EXPERIMENTS_DIFF_HELP = "Show changes between experiments."
 
     experiments_diff_parser = experiments_subparsers.add_parser(
         "diff",
@@ -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/exec_run.py

@@ -19,6 +19,8 @@ class CmdExecutorRun(CmdBaseNoRepo):
             queue=None,
             log_level=logger.getEffectiveLevel(),
             infofile=self.args.infofile,
+            copy_paths=self.args.copy_paths,
+            message=self.args.message,
         )
         return 0
 
@@ -36,4 +38,21 @@ def add_parser(experiments_subparsers, parent_parser):
         help="Path to executor info file",
         default=None,
     )
+    exec_run_parser.add_argument(
+        "-C",
+        "--copy-paths",
+        action="append",
+        default=[],
+        help=(
+            "List of ignored or untracked paths to copy into the temp directory."
+            " Only used if `--temp` or `--queue` is specified."
+        ),
+    )
+    exec_run_parser.add_argument(
+        "-M",
+        "--message",
+        type=str,
+        default=None,
+        help="Custom commit message to use when committing the experiment.",
+    )
     exec_run_parser.set_defaults(func=CmdExecutorRun)

dvc/commands/experiments/gc.py

@@ -1,145 +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 specifiedGit 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 +12,36 @@ 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,
         )
 
         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):
+                tag_base = "refs/tags/"
+                branch_base = "refs/heads/"
+                name = baseline[:7]
+                if baseline.startswith(tag_base):
+                    name = baseline[len(tag_base) :]
+                elif baseline.startswith(branch_base):
+                    name = baseline[len(branch_base) :]
+                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 +58,18 @@ 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

@@ -1,46 +1,74 @@
 import argparse
 import logging
+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
+
+        def join_exps(exps):
+            return humanize.join([f"[bold]{e}[/]" for e in exps])
+
+        if diverged_exps := result.get("diverged"):
+            exps = join_exps(diverged_exps)
+            ui.error_write(
+                f"[yellow]Local experiment {exps} has diverged "
+                "from remote experiment with the same name.\n"
+                "To override the remote experiment re-run with '--force'.",
+                styled=True,
+            )
+        if uptodate_exps := result.get("up_to_date"):
+            exps = join_exps(uptodate_exps)
+            verb = "are" if len(uptodate_exps) > 1 else "is"
+            ui.write(
+                f"Experiment {exps} {verb} up to date on Git remote {remote!r}.",
+                styled=True,
+            )
+        if pushed_exps := result.get("success"):
+            exps = join_exps(pushed_exps)
+            ui.write(f"Pushed experiment {exps} to Git remote {remote!r}.", styled=True)
+        if not uptodate_exps and not pushed_exps:
+            ui.write("No experiments to push.")
+
+        if uploaded := result.get("uploaded"):
+            stats = {"uploaded": uploaded}
+            ui.write(humanize.get_summary(stats.items()))
+
+        if project_url := result.get("url"):
+            ui.rich_print(
+                "View your experiments at", project_url, style="yellow", soft_wrap=True
             )
 
     def run(self):
-        self.raise_error_if_all_disabled()
+        from dvc.repo.experiments.push import UploadError
 
-        pushed_exps = self.repo.experiments.push(
-            self.args.git_remote,
-            self.args.experiment,
-            all_commits=self.args.all_commits,
-            rev=self.args.rev,
-            num=self.args.num,
-            force=self.args.force,
-            push_cache=self.args.push_cache,
-            dvc_remote=self.args.dvc_remote,
-            jobs=self.args.jobs,
-            run_cache=self.args.run_cache,
-        )
-
-        if pushed_exps:
-            ui.write(
-                f"Pushed experiment '{pushed_exps}'"
-                f"to Git remote '{self.args.git_remote}'."
+        try:
+            result = self.repo.experiments.push(
+                self.args.git_remote,
+                self.args.experiment,
+                all_commits=self.args.all_commits,
+                rev=self.args.rev,
+                num=self.args.num,
+                force=self.args.force,
+                push_cache=self.args.push_cache,
+                dvc_remote=self.args.dvc_remote,
+                jobs=self.args.jobs,
+                run_cache=self.args.run_cache,
             )
-        else:
-            ui.write("No experiments to push.")
+        except UploadError as e:
+            self.log_result(e.result, self.args.git_remote)
+            raise
+
+        self.log_result(result, self.args.git_remote)
         if not self.args.push_cache:
             ui.write(
                 "To push cached outputs",
@@ -64,7 +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/queue_worker.py

@@ -9,7 +9,9 @@ class CmdQueueWorker(CmdBase):
     """Run the exp queue worker."""
 
     def run(self):
-        self.repo.experiments.celery_queue.worker.start(self.args.name)
+        self.repo.experiments.celery_queue.worker.start(
+            self.args.name, fsapp_clean=self.args.clean
+        )
         return 0
 
 
@@ -22,4 +24,9 @@ def add_parser(experiments_subparsers, parent_parser):
         add_help=False,
     )
     parser.add_argument("name", help="Celery worker name.")
+    parser.add_argument(
+        "--clean",
+        action="store_true",
+        help="Automatically cleanup celery broker on shutdown.",
+    )
     parser.set_defaults(func=CmdQueueWorker)

dvc/commands/experiments/remove.py

@@ -24,9 +24,11 @@ class CmdExperimentsRemove(CmdBase):
             )
 
     def run(self):
+        from dvc.utils import humanize
+
         self.check_arguments()
 
-        removed_list = self.repo.experiments.remove(
+        removed = self.repo.experiments.remove(
             exp_names=self.args.experiment,
             all_commits=self.args.all_commits,
             rev=self.args.rev,
@@ -34,8 +36,10 @@ class CmdExperimentsRemove(CmdBase):
             queue=self.args.queue,
             git_remote=self.args.git_remote,
         )
-        removed = ",".join(removed_list)
-        ui.write(f"Removed experiments: {removed}")
+        if removed:
+            ui.write(f"Removed experiments: {humanize.join(map(repr, removed))}")
+        else:
+            ui.write("No experiments to remove.")
 
         return 0
 

dvc/commands/experiments/run.py

@@ -1,49 +1,28 @@
 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
 
 
@@ -57,22 +36,6 @@ 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)
 
 
@@ -136,3 +99,20 @@ def _add_run_common(parser):
         # )
         # metavar="<name>",
     )
+    parser.add_argument(
+        "-C",
+        "--copy-paths",
+        action="append",
+        default=[],
+        help=(
+            "List of ignored or untracked paths to copy into the temp directory."
+            " Only used if `--temp` or `--queue` is specified."
+        ),
+    )
+    parser.add_argument(
+        "-m",
+        "--message",
+        type=str,
+        default=None,
+        help="Custom commit message to use when committing the experiment.",
+    )

dvc/commands/experiments/save.py

@@ -16,6 +16,7 @@ class CmdExperimentsSave(CmdBase):
                 name=self.args.name,
                 force=self.args.force,
                 include_untracked=self.args.include_untracked,
+                message=self.args.message,
             )
         except DvcException:
             logger.exception("failed to save experiment")
@@ -35,7 +36,7 @@ class CmdExperimentsSave(CmdBase):
 
 
 def add_parser(experiments_subparsers, parent_parser):
-    EXPERIMENTS_SAVE_HELP = "Save current workspace as a dvc experiment."
+    EXPERIMENTS_SAVE_HELP = "Save current workspace as an experiment."
     save_parser = experiments_subparsers.add_parser(
         "save",
         parents=[parent_parser],
@@ -52,7 +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.",
@@ -75,4 +75,11 @@ def add_parser(experiments_subparsers, parent_parser):
         help="List of untracked paths to include in the experiment.",
         metavar="<path>",
     )
+    save_parser.add_argument(
+        "-M",
+        "--message",
+        type=str,
+        default=None,
+        help="Custom commit message to use when committing the experiment.",
+    )
     save_parser.set_defaults(func=CmdExperimentsSave)

dvc/commands/experiments/show.py

@@ -1,20 +1,16 @@
 import argparse
 import logging
-import os
 import re
-from collections import Counter, OrderedDict, defaultdict
 from datetime import date, datetime
-from typing import TYPE_CHECKING, Any, Dict, Set
+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
-from dvc.exceptions import DvcException, InvalidArgumentError
+from dvc.exceptions import DvcException
 from dvc.ui import ui
-from dvc.utils.flatten import flatten
 from dvc.utils.serialize import encode_exception
 
 if TYPE_CHECKING:
@@ -28,263 +24,13 @@ FILL_VALUE_ERRORED = "!"
 logger = logging.getLogger(__name__)
 
 
-def _update_names(names, items):
-    for name, item in items:
-        item = item.get("data", {})
-        if isinstance(item, dict):
-            item = flatten(item)
-            names[name].update({key: None for key in item})
-
-
-def _collect_names(all_experiments, **kwargs):
-    metric_names: Dict[str, Dict[str, Any]] = defaultdict(dict)
-    param_names: Dict[str, Dict[str, Any]] = defaultdict(dict)
-    deps_names: Set[str] = set()
-
-    for _, experiments in all_experiments.items():
-        for exp_data in experiments.values():
-            exp = exp_data.get("data", {})
-            _update_names(metric_names, exp.get("metrics", {}).items())
-            _update_names(param_names, exp.get("params", {}).items())
-            for dep_name in exp.get("deps", {}):
-                deps_names.add(dep_name)
-
-    return metric_names, param_names, sorted(deps_names)
-
-
 experiment_types = {
-    "checkpoint_tip": "│ ╓",
-    "checkpoint_commit": "│ ╟",
-    "checkpoint_base": "├─╨",
     "branch_commit": "├──",
     "branch_base": "└──",
     "baseline": "",
 }
 
 
-def _collect_rows(  # noqa: C901
-    base_rev,
-    experiments,
-    all_headers,
-    metric_headers,
-    param_headers,
-    metric_names,
-    param_names,
-    deps_names,
-    precision=DEFAULT_PRECISION,
-    sort_by=None,
-    sort_order=None,
-    fill_value=FILL_VALUE,
-    iso=False,
-):
-    from dvc.scm import Git
-
-    if sort_by:
-        sort_path, sort_name, sort_type = _sort_column(
-            sort_by, metric_names, param_names
-        )
-        reverse = sort_order == "desc"
-        experiments = _sort_exp(experiments, sort_path, sort_name, sort_type, reverse)
-
-    new_checkpoint = True
-    for i, (rev, results) in enumerate(experiments.items()):
-        fill_value = FILL_VALUE_ERRORED if results.get("error") else fill_value
-        row_dict = {k: fill_value for k in all_headers}
-
-        exp = results.get("data", {})
-
-        state = exp.get("status")
-        if state == "Success":
-            state = fill_value
-
-        is_baseline = rev == "baseline"
-
-        if is_baseline:
-            name_rev = base_rev[:7] if Git.is_sha(base_rev) else base_rev
-        else:
-            name_rev = rev[:7]
-
-        tip = exp.get("checkpoint_tip")
-        parent_rev = exp.get("checkpoint_parent", "")
-        parent_exp = experiments.get(parent_rev, {}).get("data", {})
-        parent_tip = parent_exp.get("checkpoint_tip")
-
-        parent = ""
-        if is_baseline:
-            typ = "baseline"
-        elif tip:
-            if tip == parent_tip:
-                typ = "checkpoint_tip" if new_checkpoint else "checkpoint_commit"
-            elif parent_rev == base_rev:
-                typ = "checkpoint_base"
-            else:
-                typ = "checkpoint_commit"
-                parent = parent_rev[:7]
-        elif i < len(experiments) - 1:
-            typ = "branch_commit"
-        else:
-            typ = "branch_base"
-
-        if not is_baseline:
-            new_checkpoint = not (tip and tip == parent_tip)
-
-        row_dict["Experiment"] = exp.get("name", "")
-        row_dict["rev"] = name_rev
-        row_dict["typ"] = typ
-        row_dict["Created"] = format_time(exp.get("timestamp"), fill_value, iso)
-        row_dict["parent"] = parent
-        row_dict["State"] = state
-        row_dict["Executor"] = exp.get("executor", fill_value)
-
-        _extend_row(
-            row_dict,
-            metric_names,
-            metric_headers,
-            exp.get("metrics", {}).items(),
-            precision,
-            fill_value=fill_value,
-        )
-        _extend_row(
-            row_dict,
-            param_names,
-            param_headers,
-            exp.get("params", {}).items(),
-            precision,
-            fill_value=fill_value,
-        )
-        for dep in deps_names:
-            hash_info = exp.get("deps", {}).get(dep, {}).get("hash")
-            if hash_info is not None:
-                hash_info = hash_info[:7]
-            row_dict[dep] = hash_info
-        yield list(row_dict.values())
-
-
-def _sort_column(sort_by, metric_names, param_names):
-    path, _, sort_name = sort_by.rpartition(":")
-    matches = set()
-
-    if path:
-        if path in metric_names and sort_name in metric_names[path]:
-            matches.add((path, sort_name, "metrics"))
-        if path in param_names and sort_name in param_names[path]:
-            matches.add((path, sort_name, "params"))
-    else:
-        for path in metric_names:
-            if sort_name in metric_names[path]:
-                matches.add((path, sort_name, "metrics"))
-        for path in param_names:
-            if sort_name in param_names[path]:
-                matches.add((path, sort_name, "params"))
-
-    if len(matches) == 1:
-        return matches.pop()
-    if len(matches) > 1:
-        raise InvalidArgumentError(
-            "Ambiguous sort column '{}' matched '{}'".format(
-                sort_by,
-                ", ".join([f"{path}:{name}" for path, name, _ in matches]),
-            )
-        )
-    raise InvalidArgumentError(f"Unknown sort column '{sort_by}'")
-
-
-def _sort_exp(experiments, sort_path, sort_name, typ, reverse):
-    def _sort(item):
-        rev, exp = item
-        exp_data = exp.get("data", {})
-        tip = exp_data.get("checkpoint_tip")
-        if tip and tip != rev:
-            # Sort checkpoint experiments by tip commit
-            return _sort((tip, experiments[tip]))
-        data = exp_data.get(typ, {}).get(sort_path, {}).get("data", {})
-        val = flatten(data).get(sort_name)
-        return val is None, val
-
-    ret = OrderedDict()
-    if "baseline" in experiments:
-        ret["baseline"] = experiments.pop("baseline")
-
-    ret.update(sorted(experiments.items(), key=_sort, reverse=reverse))
-    return ret
-
-
-def format_time(datetime_obj, fill_value=FILL_VALUE, iso=False):
-    if datetime_obj is None:
-        return fill_value
-
-    if iso:
-        return datetime_obj.isoformat()
-
-    if datetime_obj.date() == date.today():
-        fmt = "%I:%M %p"
-    else:
-        fmt = "%b %d, %Y"
-    return datetime_obj.strftime(fmt)
-
-
-def _extend_row(row, names, headers, items, precision, fill_value=FILL_VALUE):
-    from dvc.compare import _format_field, with_value
-
-    for fname, data in items:
-        item = data.get("data", {})
-        item = flatten(item) if isinstance(item, dict) else {fname: item}
-        for name in names[fname]:
-            value = with_value(
-                item.get(name),
-                FILL_VALUE_ERRORED if data.get("error", None) else fill_value,
-            )
-            # wrap field data in ui.rich_text, otherwise rich may
-            # interpret unescaped braces from list/dict types as rich
-            # markup tags
-            value = ui.rich_text(str(_format_field(value, precision)))
-            if name in headers:
-                row[name] = value
-            else:
-                row[f"{fname}:{name}"] = value
-
-
-def experiments_table(
-    all_experiments,
-    headers,
-    metric_headers,
-    metric_names,
-    param_headers,
-    param_names,
-    deps_names,
-    sort_by=None,
-    sort_order=None,
-    precision=DEFAULT_PRECISION,
-    fill_value=FILL_VALUE,
-    iso=False,
-) -> "TabularData":
-    from funcy import lconcat
-
-    from dvc.compare import TabularData
-
-    all_headers = lconcat(headers, metric_headers, param_headers, deps_names)
-    td = TabularData(all_headers, fill_value=fill_value)
-    for base_rev, experiments in all_experiments.items():
-        rows = _collect_rows(
-            base_rev,
-            experiments,
-            all_headers,
-            metric_headers,
-            param_headers,
-            metric_names,
-            param_names,
-            deps_names,
-            sort_by=sort_by,
-            sort_order=sort_order,
-            precision=precision,
-            fill_value=fill_value,
-            iso=iso,
-        )
-        td.extend(rows)
-
-    return td
-
-
 def prepare_exp_id(kwargs) -> "RichText":
     exp_name = kwargs["Experiment"]
     rev = kwargs["rev"]
@@ -309,49 +55,15 @@ def baseline_styler(typ):
 
 
 def show_experiments(  # noqa: C901
-    all_experiments,
+    td: "TabularData",
+    headers: Dict[str, Iterable[str]],
     keep=None,
     drop=None,
     pager=True,
     csv=False,
     markdown=False,
-    pcp=False,
     **kwargs,
 ):
-    from funcy.seqs import flatten as flatten_list
-
-    metric_names, param_names, deps_names = _collect_names(all_experiments)
-
-    headers_list = [
-        "Experiment",
-        "rev",
-        "typ",
-        "Created",
-        "parent",
-        "State",
-        "Executor",
-    ]
-
-    names = {**metric_names, **param_names}
-    counter = Counter(flatten_list([list(a.keys()) for a in names.values()]))
-    counter.update(headers_list)
-    metric_headers = _normalize_headers(metric_names, counter)
-    param_headers = _normalize_headers(param_names, counter)
-
-    td = experiments_table(
-        all_experiments,
-        headers_list,
-        metric_headers,
-        metric_names,
-        param_headers,
-        param_names,
-        deps_names,
-        kwargs.get("sort_by"),
-        kwargs.get("sort_order"),
-        kwargs.get("precision"),
-        kwargs.get("fill_value"),
-        kwargs.get("iso"),
-    )
     if keep:
         for col in td.keys():  # noqa: SIM118
             if re.match(keep, col):
@@ -368,11 +80,6 @@ def show_experiments(  # noqa: C901
         td.column("Experiment")[:] = map(prepare_exp_id, td.as_dict(merge_headers))
         td.drop(*merge_headers[1:])
 
-    headers = {
-        "metrics": metric_headers,
-        "params": param_headers,
-        "deps": deps_names,
-    }
     styles = {
         "Experiment": {"no_wrap": True, "header_style": "black on grey93"},
         "Created": {"header_style": "black on grey93"},
@@ -397,51 +104,23 @@ def show_experiments(  # noqa: C901
         }
     )
 
-    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):
@@ -460,8 +139,10 @@ def _format_json(item):
 
 class CmdExperimentsShow(CmdBase):
     def run(self):
+        from dvc.repo.experiments.show import tabulate
+
         try:
-            all_experiments = self.repo.experiments.show(
+            exps = self.repo.experiments.show(
                 all_branches=self.args.all_branches,
                 all_tags=self.args.all_tags,
                 all_commits=self.args.all_commits,
@@ -472,36 +153,40 @@ class CmdExperimentsShow(CmdBase):
                 sha_only=self.args.sha,
                 param_deps=self.args.param_deps,
                 fetch_running=self.args.fetch_running,
+                force=self.args.force,
             )
         except DvcException:
             logger.exception("failed to show experiments")
             return 1
 
         if self.args.json:
-            ui.write_json(all_experiments, default=_format_json)
+            ui.write_json([exp.dumpd() for exp in exps], default=_format_json)
         else:
             precision = (
                 self.args.precision or None if self.args.csv else DEFAULT_PRECISION
             )
             fill_value = "" if self.args.csv else FILL_VALUE
             iso = self.args.csv
+            td, headers = tabulate(
+                exps,
+                precision=precision,
+                fill_value=fill_value,
+                iso=iso,
+                sort_by=self.args.sort_by,
+                sort_order=self.args.sort_order,
+            )
 
             show_experiments(
-                all_experiments,
+                td,
+                headers,
                 keep=self.args.keep,
                 drop=self.args.drop,
                 sort_by=self.args.sort_by,
                 sort_order=self.args.sort_order,
-                precision=precision,
-                fill_value=fill_value,
-                iso=iso,
                 pager=not self.args.no_pager,
                 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
 
@@ -594,21 +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",
@@ -623,31 +305,16 @@ 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",
         action="store_false",
         help=argparse.SUPPRESS,
     )
+    experiments_show_parser.add_argument(
+        "-f",
+        "--force",
+        action="store_true",
+        help="Force re-collection of experiments instead of loading from exp cache.",
+    )
     experiments_show_parser.set_defaults(func=CmdExperimentsShow)

dvc/commands/gc.py

@@ -10,7 +10,7 @@ logger = logging.getLogger(__name__)
 
 
 class CmdGC(CmdBase):
-    def run(self):  # noqa: C901
+    def run(self):  # noqa: C901, PLR0912
         from dvc.repo.gc import _validate_args
 
         _validate_args(
@@ -22,6 +22,8 @@ class CmdGC(CmdBase):
             workspace=self.args.workspace,
             rev=self.args.rev,
             num=self.args.num,
+            cloud=self.args.cloud,
+            not_in_remote=self.args.not_in_remote,
         )
 
         if self.args.rev:
@@ -46,6 +48,10 @@ class CmdGC(CmdBase):
 
         if self.args.all_experiments:
             msg += " and all experiments"
+
+        if self.args.not_in_remote:
+            msg += " that are present in the DVC remote"
+
         if self.args.repos:
             msg += " of the current and the following repos:"
 
@@ -74,6 +80,7 @@ class CmdGC(CmdBase):
             workspace=self.args.workspace,
             rev=self.args.rev,
             num=self.args.num,
+            not_in_remote=self.args.not_in_remote,
         )
         return 0
 
@@ -157,12 +164,18 @@ def add_parser(subparsers, parent_parser):
         default=False,
         help="Keep data files for all experiments.",
     )
+    gc_parser.add_argument(
+        "--not-in-remote",
+        action="store_true",
+        default=False,
+        help="Keep data files that are not present in the remote.",
+    )
     gc_parser.add_argument(
         "-c",
         "--cloud",
         action="store_true",
         default=False,
-        help="Collect garbage in remote repository.",
+        help="Collect garbage in remote storage in addition to local cache.",
     )
     gc_parser.add_argument(
         "-r",

dvc/commands/get.py

@@ -3,9 +3,8 @@ 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
-from dvc.scm import CloneError
 
 logger = logging.getLogger(__name__)
 
@@ -15,12 +14,8 @@ class CmdGet(CmdBaseNoRepo):
         from dvc.api import get_url
         from dvc.ui import ui
 
-        try:
-            url = get_url(self.args.path, repo=self.args.url, rev=self.args.rev)
-            ui.write(url, force=True)
-        except DvcException:
-            logger.exception("failed to show URL")
-            return 1
+        url = get_url(self.args.path, repo=self.args.url, rev=self.args.rev)
+        ui.write(url, force=True)
 
         return 0
 
@@ -32,6 +27,7 @@ class CmdGet(CmdBaseNoRepo):
 
     def _get_file_from_repo(self):
         from dvc.repo import Repo
+        from dvc.scm import CloneError
 
         try:
             Repo.get(
@@ -40,6 +36,10 @@ class CmdGet(CmdBaseNoRepo):
                 out=self.args.out,
                 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:
@@ -98,4 +98,34 @@ def add_parser(subparsers, parent_parser):
         ),
         metavar="<number>",
     )
+    get_parser.add_argument(
+        "-f",
+        "--force",
+        action="store_true",
+        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/get_url.py

@@ -14,7 +14,12 @@ class CmdGetUrl(CmdBaseNoRepo):
         from dvc.repo import Repo
 
         try:
-            Repo.get_url(self.args.url, out=self.args.out, jobs=self.args.jobs)
+            Repo.get_url(
+                self.args.url,
+                out=self.args.out,
+                jobs=self.args.jobs,
+                force=self.args.force,
+            )
             return 0
         except DvcException:
             logger.exception("failed to get '%s'", self.args.url)
@@ -46,4 +51,11 @@ def add_parser(subparsers, parent_parser):
         ),
         metavar="<number>",
     )
+    get_parser.add_argument(
+        "-f",
+        "--force",
+        action="store_true",
+        default=False,
+        help="Override local file or folder if exists.",
+    )
     get_parser.set_defaults(func=CmdGetUrl)

dvc/commands/imp.py

@@ -3,29 +3,28 @@ 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
-from dvc.scm import CloneError
 
 logger = logging.getLogger(__name__)
 
 
 class CmdImport(CmdBase):
     def run(self):
+        from dvc.scm import CloneError
+
         try:
             self.repo.imp(
                 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)
@@ -41,8 +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."
@@ -74,11 +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",
@@ -105,6 +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

@@ -15,16 +15,12 @@ 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,
             )
         except DvcException:
@@ -40,8 +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(
@@ -69,11 +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",
@@ -112,11 +101,17 @@ def add_parser(subparsers, parent_parser):
         ),
         metavar="<number>",
     )
+    import_parser.add_argument(
+        "-f",
+        "--force",
+        action="store_true",
+        default=False,
+        help="Override local file or folder if exists.",
+    )
     import_parser.add_argument(
         "--version-aware",
         action="store_true",
         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

@@ -302,7 +302,7 @@ class CmdMachineSsh(CmdBase):
         return 0
 
 
-def add_parser(subparsers, parent_parser):
+def add_parser(subparsers, parent_parser):  # noqa: PLR0915
     from dvc.commands.config import parent_config_parser
 
     machine_HELP = "Set up and manage cloud machines."

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

@@ -1,9 +1,9 @@
 import argparse
-import json
 import logging
 import os
+from typing import TYPE_CHECKING, Dict, List, Optional
 
-from funcy import first
+from funcy import compact, first, get_in
 
 from dvc.cli import completion
 from dvc.cli.command import CmdBase
@@ -12,14 +12,51 @@ from dvc.exceptions import DvcException
 from dvc.ui import ui
 from dvc.utils import format_link
 
+if TYPE_CHECKING:
+    from dvc.render.match import RendererWithErrors
+
+
 logger = logging.getLogger(__name__)
 
 
-def _show_json(renderers, split=False):
+def _show_json(
+    renderers_with_errors: List["RendererWithErrors"],
+    split=False,
+    errors: Optional[Dict[str, Exception]] = None,
+):
     from dvc.render.convert import to_json
+    from dvc.utils.serialize import encode_exception
 
-    result = {renderer.name: to_json(renderer, split) for renderer in renderers}
-    ui.write_json(result)
+    all_errors: List[Dict] = []
+    data = {}
+
+    for renderer, src_errors, def_errors in renderers_with_errors:
+        name = renderer.name
+        data[name] = to_json(renderer, split)
+        all_errors.extend(
+            {
+                "name": name,
+                "rev": rev,
+                "source": source,
+                **encode_exception(e),
+            }
+            for rev, per_rev_src_errors in src_errors.items()
+            for source, e in per_rev_src_errors.items()
+        )
+        all_errors.extend(
+            {
+                "name": name,
+                "rev": rev,
+                **encode_exception(e),
+            }
+            for rev, e in def_errors.items()
+        )
+
+    # these errors are not tied to any renderers
+    errors = errors or {}
+    all_errors.extend({"rev": rev, **encode_exception(e)} for rev, e in errors.items())
+
+    ui.write_json(compact({"errors": all_errors, "data": data}), highlight=False)
 
 
 def _adjust_vega_renderers(renderers):
@@ -76,10 +113,6 @@ class CmdPlots(CmdBase):
         props = {p: getattr(self.args, p) for p in PLOT_PROPS}
         return {k: v for k, v in props.items() if v is not None}
 
-    def _config_files(self):
-        if self.args.from_config:
-            return {self.args.from_config}
-
     def _html_template_path(self):
         html_template_path = self.args.html_template
         if not html_template_path:
@@ -89,9 +122,9 @@ class CmdPlots(CmdBase):
             if html_template_path and not os.path.isabs(html_template_path):
                 assert self.repo.dvc_dir
                 html_template_path = os.path.join(self.repo.dvc_dir, html_template_path)
-        return html_template_path  # noqa: RET504
+        return html_template_path
 
-    def run(self) -> int:  # noqa: C901
+    def run(self) -> int:  # noqa: C901, PLR0911, PLR0912
         from pathlib import Path
 
         from dvc.render.match import match_defs_renderers
@@ -114,10 +147,9 @@ class CmdPlots(CmdBase):
             plots_data = self._func(
                 targets=self.args.targets,
                 props=self._props(),
-                config_files=self._config_files(),
             )
 
-            if not plots_data:
+            if not plots_data and not self.args.json:
                 ui.error_write(
                     "No plots were loaded, visualization file will not be created."
                 )
@@ -127,21 +159,27 @@ class CmdPlots(CmdBase):
             )
 
             renderers_out = out if self.args.json else os.path.join(out, "static")
-
-            renderers = match_defs_renderers(
+            renderers_with_errors = match_defs_renderers(
                 data=plots_data,
                 out=renderers_out,
                 templates_dir=self.repo.plots.templates_dir,
             )
             if self.args.json:
-                _show_json(renderers, self.args.split)
+                errors = compact(
+                    {
+                        rev: get_in(data, ["definitions", "error"])
+                        for rev, data in plots_data.items()
+                    }
+                )
+                _show_json(renderers_with_errors, self.args.split, errors=errors)
                 return 0
 
+            renderers = [r.renderer for r in renderers_with_errors]
             _adjust_vega_renderers(renderers)
             if self.args.show_vega:
                 renderer = first(filter(lambda r: r.TYPE == "vega", renderers))
                 if renderer:
-                    ui.write_json(json.loads(renderer.get_filled_template()))
+                    ui.write_json(renderer.get_filled_template(as_string=False))
                 return 0
 
             output_file: Path = (Path.cwd() / out).resolve() / "index.html"
@@ -150,7 +188,7 @@ class CmdPlots(CmdBase):
                 render_html(
                     renderers=renderers,
                     output_file=output_file,
-                    template_path=self._html_template_path(),
+                    html_template=self._html_template_path(),
                 )
 
                 ui.write(output_file.as_uri())
@@ -392,7 +430,6 @@ def _add_ui_arguments(parser):
     )
     parser.add_argument(
         "--json",
-        "--show-json",
         action="store_true",
         default=False,
         help=argparse.SUPPRESS,
@@ -412,9 +449,3 @@ def _add_ui_arguments(parser):
         help="Custom HTML template for VEGA visualization.",
         metavar="<path>",
     )
-    parser.add_argument(
-        "--from-config",
-        default=None,
-        metavar="<path>",
-        help=argparse.SUPPRESS,
-    )

dvc/commands/queue/status.py

@@ -1,10 +1,8 @@
 import argparse
 import logging
-from typing import Dict, List, Optional
 
 from dvc.cli.command import CmdBase
 from dvc.cli.utils import append_doc_link
-from dvc.commands.experiments.show import format_time
 from dvc.compare import TabularData
 from dvc.ui import ui
 
@@ -15,9 +13,9 @@ class CmdQueueStatus(CmdBase):
     """Show queue task and worker status."""
 
     def run(self) -> int:
-        result: List[
-            Dict[str, Optional[str]]
-        ] = self.repo.experiments.celery_queue.status()
+        from dvc.repo.experiments.show import format_time
+
+        result = self.repo.experiments.celery_queue.status()
         if result:
             all_headers = ["Task", "Name", "Created", "Status"]
             td = TabularData(all_headers)

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
@@ -38,6 +32,8 @@ class CmdRepro(CmdBase):
             "recursive": self.args.recursive,
             "force_downstream": self.args.force_downstream,
             "pull": self.args.pull,
+            "allow_missing": self.args.allow_missing,
+            "on_error": self.args.on_error,
         }
 
     @property
@@ -104,13 +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",
@@ -135,6 +124,12 @@ and then the stage name name.
             "from the run-cache."
         ),
     )
+    repro_parser.add_argument(
+        "--allow-missing",
+        action="store_true",
+        default=False,
+        help=("Skip stages with missing data but no other changes."),
+    )
     repro_parser.add_argument(
         "--dry",
         action="store_true",
@@ -143,6 +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/commands/update.py

@@ -31,8 +31,7 @@ class CmdUpdate(CmdBase):
 def add_parser(subparsers, parent_parser):
     UPDATE_HELP = (
         "Update data artifact imported (via dvc import or dvc import-url) "
-        "from an external DVC repository or URL, or update data from a "
-        "DVC worktree remote."
+        "from an external DVC repository or URL."
     )
     update_parser = subparsers.add_parser(
         "update",

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
 
 
@@ -35,7 +34,7 @@ def with_value(value, default):
 
 
 class TabularData(MutableSequence[Sequence["CellT"]]):
-    def __init__(self, columns: Sequence[str], fill_value: str = ""):
+    def __init__(self, columns: Sequence[str], fill_value: Optional[str] = ""):
         self._columns: Dict[str, Column] = {name: Column() for name in columns}
         self._keys: List[str] = list(columns)
         self._fill_value = fill_value
@@ -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)
 
@@ -209,7 +189,7 @@ class TabularData(MutableSequence[Sequence["CellT"]]):
         keys = self.keys() if cols is None else set(cols)
         return [{k: self._columns[k][i] for k in keys} for i in range(len(self))]
 
-    def dropna(  # noqa: C901
+    def dropna(  # noqa: C901, PLR0912
         self,
         axis: str = "rows",
         how="any",
@@ -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
@@ -371,7 +351,7 @@ def diff_table(
     return td
 
 
-def show_diff(
+def show_diff(  # noqa: PLR0913
     diff,
     title: str,
     old: bool = True,

dvc/config.py

@@ -79,9 +79,6 @@ class Config(dict):
         ConfigError: thrown if config has an invalid format.
     """
 
-    APPNAME = "dvc"
-    APPAUTHOR = "iterative"
-
     SYSTEM_LEVELS = ("system", "global")
     REPO_LEVELS = ("repo", "local")
     # In the order they shadow each other
@@ -96,6 +93,8 @@ class Config(dict):
         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
 
@@ -103,28 +102,34 @@ class Config(dict):
         self.wfs = LocalFileSystem()
         self.fs = fs or self.wfs
 
-        if not dvc_dir:
-            try:
-                from dvc.repo import Repo
+        if dvc_dir:
+            self.dvc_dir = self.fs.path.abspath(dvc_dir)
 
-                self.dvc_dir = Repo.find_dvc_dir()
-            except NotDvcRepoError:
-                self.dvc_dir = None
-        else:
-            self.dvc_dir = self.fs.path.abspath(self.fs.path.realpath(dvc_dir))
+        self.load(
+            validate=validate, config=config, remote=remote, remote_config=remote_config
+        )
 
-        self.load(validate=validate, config=config)
+    @classmethod
+    def from_cwd(cls, fs: Optional["FileSystem"] = None, **kwargs):
+        from dvc.repo import Repo
+
+        try:
+            dvc_dir = Repo.find_dvc_dir(fs=fs)
+        except NotDvcRepoError:
+            dvc_dir = None
+
+        return cls(dvc_dir=dvc_dir, fs=fs, **kwargs)
 
     @classmethod
     def get_dir(cls, level):
-        from appdirs import site_config_dir, user_config_dir
+        from dvc.dirs import global_config_dir, system_config_dir
 
         assert level in ("global", "system")
 
         if level == "global":
-            return user_config_dir(cls.APPNAME, cls.APPAUTHOR)
+            return global_config_dir()
         if level == "system":
-            return site_config_dir(cls.APPNAME, cls.APPAUTHOR)
+            return system_config_dir()
 
     @cached_property
     def files(self) -> Dict[str, str]:
@@ -153,7 +158,13 @@ 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 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:
@@ -168,34 +179,50 @@ class Config(dict):
             conf = self.validate(conf)
 
         self.clear()
-        self.update(conf)
 
-        # Add resolved default cache.dir
-        if not self["cache"].get("dir") and self.dvc_dir:
-            self["cache"]["dir"] = os.path.join(self.dvc_dir, "cache")
+        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):
         # NOTE: this might be a Gitfs, which doesn't see things outside of
         # 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

dvc/config_schema.py

@@ -1,8 +1,22 @@
+import logging
 import os
 from urllib.parse import urlparse
 
-from funcy import walk_values
-from voluptuous import All, Any, Coerce, Invalid, Lower, Optional, Range, Schema
+from funcy import once, walk_values
+from voluptuous import (
+    REMOVE_EXTRA,
+    All,
+    Any,
+    Coerce,
+    Exclusive,
+    Invalid,
+    Lower,
+    Optional,
+    Range,
+    Schema,
+)
+
+logger = logging.getLogger(__name__)
 
 Bool = All(
     Lower,
@@ -30,7 +44,7 @@ def supported_cache_type(types):
     return types
 
 
-def Choices(*choices):
+def Choices(*choices):  # noqa: N802
     """Checks that value belongs to the specified set of values
 
     Args:
@@ -40,7 +54,7 @@ def Choices(*choices):
     return Any(*choices, msg="expected one of {}".format(", ".join(choices)))
 
 
-def ByUrl(mapping):
+def ByUrl(mapping):  # noqa: N802
     schemas = walk_values(Schema, mapping)
 
     def validate(data):
@@ -49,7 +63,9 @@ def ByUrl(mapping):
 
         parsed = urlparse(data["url"])
         # Windows absolute paths should really have scheme == "" (local)
-        if os.name == "nt" and len(parsed.scheme) == 1 and parsed.netloc == "":
+        if os.name == "nt" and len(parsed.scheme) == 1 and not parsed.netloc:
+            return schemas[""](data)
+        if not parsed.netloc:
             return schemas[""](data)
         if parsed.scheme not in schemas:
             raise Invalid(f"Unsupported URL type {parsed.scheme}://")
@@ -60,7 +76,28 @@ def ByUrl(mapping):
 
 
 class RelPath(str):
-    pass
+    __slots__ = ()
+
+
+class FeatureSchema(Schema):
+    def __init__(self, schema, required=False):
+        super().__init__(schema, required=required, extra=REMOVE_EXTRA)
+
+    @staticmethod
+    @once
+    def _log_deprecated(keys):
+        # only run this once per session
+        message = "%s config option%s unsupported"
+        paths = ", ".join(f"'feature.{key}'" for key in keys)
+        pluralize = " is" if len(keys) == 1 else "s are"
+        logger.warning(message, paths, pluralize)
+
+    def __call__(self, data):
+        ret = super().__call__(data)
+        extra_keys = data.keys() - ret.keys()
+        if extra_keys:
+            self._log_deprecated(sorted(extra_keys))
+        return ret
 
 
 REMOTE_COMMON = {
@@ -114,16 +151,17 @@ SCHEMA = {
         Optional("autostage", default=False): Bool,
         Optional("experiments"): Bool,  # obsoleted
         Optional("check_update", default=True): Bool,
+        "site_cache_dir": str,
         "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,
@@ -178,6 +216,7 @@ SCHEMA = {
                     "timeout": Coerce(int),
                     "gss_auth": Bool,
                     "allow_agent": Bool,
+                    "max_sessions": Coerce(int),
                     Optional("verify", default=False): Bool,
                     **REMOTE_COMMON,
                 },
@@ -237,12 +276,12 @@ SCHEMA = {
         )
     },
     "state": {
-        "dir": str,
+        "dir": str,  # obsoleted
         "row_limit": All(Coerce(int), Range(1)),  # obsoleted
         "row_cleanup_quota": All(Coerce(int), Range(0, 100)),  # obsoleted
     },
     "index": {
-        "dir": str,
+        "dir": str,  # obsoleted
     },
     "machine": {
         str: {
@@ -260,11 +299,12 @@ SCHEMA = {
         },
     },
     # section for experimental features
-    "feature": {
-        Optional("machine", default=False): Bool,
-        # enabled by default. It's of no use, kept for backward compatibility.
-        Optional("parametrization", default=True): Bool,
-    },
+    # only specified keys are validated, others get logged and then ignored/removed
+    "feature": FeatureSchema(
+        {
+            Optional("machine", default=False): Bool,
+        },
+    ),
     "plots": {
         "html_template": str,
         Optional("auto_open", default=False): Bool,
@@ -285,7 +325,14 @@ 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": {
+        "token": str,
+        "url": str,
+        "repo_url": str,
+        Optional("offline", default=False): Bool,
+    },
 }

dvc/daemon.py

@@ -5,7 +5,7 @@ import logging
 import os
 import platform
 import sys
-from subprocess import Popen
+from subprocess import Popen  # nosec B404
 from typing import List
 
 from dvc.env import DVC_DAEMON
@@ -14,19 +14,28 @@ from dvc.utils import fix_env, is_binary
 logger = logging.getLogger(__name__)
 
 
-def _popen(cmd, **kwargs):
+def _suppress_resource_warning(popen: Popen):
+    """Sets the returncode to avoid ResourceWarning when popen is garbage collected."""
+    # only use for daemon processes.
+    # See https://bugs.python.org/issue38890.
+    popen.returncode = 0
+
+
+def _popen(cmd, **kwargs) -> Popen:
     prefix = [sys.executable]
     if not is_binary():
         main_entrypoint = os.path.join(
             os.path.abspath(os.path.dirname(__file__)), "__main__.py"
         )
         prefix += [main_entrypoint]
-    return Popen(prefix + cmd, close_fds=True, shell=False, **kwargs)
+    return Popen(
+        prefix + cmd, close_fds=True, shell=False, **kwargs  # nosec B603  # noqa: S603
+    )
 
 
 def _spawn_windows(cmd, env):
     if sys.platform == "win32":
-        from subprocess import (
+        from subprocess import (  # nosec B404
             CREATE_NEW_PROCESS_GROUP,
             CREATE_NO_WINDOW,
             STARTF_USESHOWWINDOW,
@@ -40,12 +49,22 @@ def _spawn_windows(cmd, env):
         startupinfo = STARTUPINFO()
         startupinfo.dwFlags |= STARTF_USESHOWWINDOW
 
-        _popen(cmd, env=env, creationflags=creationflags, startupinfo=startupinfo)
+        popen = _popen(
+            cmd, env=env, creationflags=creationflags, startupinfo=startupinfo
+        )
+        _suppress_resource_warning(popen)
 
 
 def _spawn_posix(cmd, env):
     from dvc.cli import main
 
+    # `fork` will copy buffers, so we need to flush them before forking.
+    # Otherwise, we will get duplicated outputs.
+    if sys.stdout and not sys.stdout.closed:
+        sys.stdout.flush()
+    if sys.stderr and not sys.stderr.closed:
+        sys.stderr.flush()
+
     # NOTE: using os._exit instead of sys.exit, because dvc built
     # with PyInstaller has trouble with SystemExit exception and throws
     # errors such as "[26338] Failed to execute script __main__"
@@ -72,6 +91,7 @@ def _spawn_posix(cmd, env):
     sys.stdin.close()
     sys.stdout.close()
     sys.stderr.close()
+    os.closerange(0, 3)
 
     if platform.system() == "Darwin":
         # workaround for MacOS bug

dvc/dagascii.py

@@ -80,7 +80,7 @@ class AsciiCanvas:
 
         self.canvas[y][x] = char
 
-    def line(self, x0, y0, x1, y1, char):  # noqa: C901
+    def line(self, x0, y0, x1, y1, char):  # noqa: C901, PLR0912
         """Create a line on ASCII canvas.
 
         Args:
@@ -209,12 +209,36 @@ 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
     # everything to the positive plane before we actually draw it.
-    Xs = []  # pylint: disable=invalid-name
-    Ys = []  # pylint: disable=invalid-name
+    Xs = []  # noqa: N806, pylint: disable=invalid-name
+    Ys = []  # noqa: N806, pylint: disable=invalid-name
 
     sug = _build_sugiyama_layout(vertices, edges)
 

dvc/data_cloud.py

@@ -1,40 +1,69 @@
 """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__)
 
 
 class Remote:
-    def __init__(self, name: str, path: str, fs: "FileSystem", **config):
+    def __init__(self, name: str, path: str, fs: "FileSystem", *, index=None, **config):
         self.path = path
         self.fs = fs
         self.name = name
+        self.index = index
 
         self.worktree: bool = config.pop("worktree", False)
         self.config = config
 
     @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, **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:
@@ -62,7 +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")
@@ -74,19 +103,23 @@ class DataCloud:
                     config["version_aware"] = True
 
             fs = cls(**config)
-            config["tmp_dir"] = self.repo.index_db_dir
-            return Remote(name, fs_path, fs, **config)
+            config["tmp_dir"] = self.repo.site_cache_dir
+            if self.repo.data_index is not None:
+                index = self.repo.data_index.view(("remote", name))
+            else:
+                index = None
+            return Remote(name, fs_path, fs, index=index, **config)
 
         if bool(self.repo.config["remote"]):
             error_msg = (
-                "no remote specified. Setup default remote with\n"
+                f"no remote specified in {self.repo}. Setup default remote with\n"
                 "    dvc remote default <remote name>\n"
                 "or use:\n"
-                "    dvc {} -r <remote name>".format(command)
+                f"    dvc {command} -r <remote name>"
             )
         else:
             error_msg = (
-                "no remote specified. Create a default remote with\n"
+                f"no remote specified in {self.repo}. Create a default remote with\n"
                 "    dvc remote add -d <remote name> <remote url>"
             )
 
@@ -96,12 +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"):
@@ -145,16 +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.odb.local,
-            odb,
-            objs,
-            jobs=jobs,
-            dest_index=get_index(odb),
-            cache_odb=self.repo.odb.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,
@@ -172,17 +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.odb.local,
-            objs,
-            jobs=jobs,
-            src_index=get_index(odb),
-            cache_odb=self.repo.odb.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,
@@ -201,17 +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.odb.local,
+            cache,
             odb,
             objs,
             jobs=jobs,
             dest_index=get_index(odb),
-            cache_odb=self.repo.odb.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, cast
+from typing import Any, Dict
 
-import dpath.util
-from voluptuous import Any
+import dpath
 
 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):
@@ -87,17 +86,17 @@ class ParamsDependency(Dependency):
         if flatten:
             for param in self.params:
                 try:
-                    ret[param] = dpath.util.get(config, param, separator=".")
+                    ret[param] = dpath.get(config, param, separator=".")
                 except KeyError:
                     continue
             return ret
 
-        from dpath.util import merge
+        from dpath import merge
 
         for param in self.params:
             merge(
                 ret,
-                dpath.util.search(config, param, separator="."),
+                dpath.search(config, param, separator="."),
                 separator=".",
             )
         return ret
@@ -111,7 +110,8 @@ class ParamsDependency(Dependency):
         from funcy import ldistinct
 
         status: Dict[str, Any] = defaultdict(dict)
-        info = cast(dict, self.hash_info.value) if self.hash_info else {}
+        assert isinstance(self.hash_info.value, dict)
+        info = self.hash_info.value if self.hash_info else {}
         actual = self.read_params()
 
         # NOTE: we want to preserve the order of params as specified in the

dvc/dependency/repo.py

@@ -1,24 +1,14 @@
-import os
-from collections import defaultdict
-from copy import copy
-from typing import TYPE_CHECKING, Dict, Optional, Set, Tuple, Union
+from typing import TYPE_CHECKING, Dict, Optional, Union
 
-from voluptuous import Required
+from voluptuous import Any, Required
 
-from dvc.prompt import confirm
+from dvc.utils import as_posix
 
 from .base import Dependency
 
 if TYPE_CHECKING:
-    from typing import ContextManager
-
-    from dvc.output import Output
-    from dvc.repo import Repo
+    from dvc.fs import DVCFileSystem
     from dvc.stage import Stage
-    from dvc_data.hashfile.hash_info import HashInfo
-    from dvc_data.hashfile.meta import Meta
-    from dvc_data.hashfile.obj import HashFile
-    from dvc_objects.db import ObjectDB
 
 
 class RepoDependency(Dependency):
@@ -26,21 +16,26 @@ 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),
         }
     }
 
     def __init__(self, def_repo: Dict[str, str], stage: "Stage", *args, **kwargs):
         self.def_repo = def_repo
-        self._objs: Dict[str, "HashFile"] = {}
-        self._meta: Dict[str, "Meta"] = {}
         super().__init__(stage, *args, **kwargs)
 
+        self.fs = self._make_fs()
+        self.fs_path = as_posix(self.def_path)
+
     def _parse_path(self, fs, fs_path):  # noqa: ARG002
         return None
 
@@ -52,8 +47,8 @@ class RepoDependency(Dependency):
         return f"{self.def_path} ({self.def_repo[self.PARAM_URL]})"
 
     def workspace_status(self):
-        current = self.get_obj(locked=True).hash_info
-        updated = self.get_obj(locked=False).hash_info
+        current = self._make_fs(locked=True).repo.get_rev()
+        updated = self._make_fs(locked=False).repo.get_rev()
 
         if current != updated:
             return {str(self): "update available"}
@@ -64,33 +59,39 @@ class RepoDependency(Dependency):
         return self.workspace_status()
 
     def save(self):
-        pass
+        rev = self.fs.repo.get_rev()
+        if self.def_repo.get(self.PARAM_REV_LOCK) is None:
+            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]}
 
-    def download(self, to: "Output", jobs: Optional[int] = None):
-        from dvc_data.hashfile.checkout import checkout
+        rev = self.def_repo.get(self.PARAM_REV)
+        if rev:
+            repo[self.PARAM_REV] = self.def_repo[self.PARAM_REV]
 
-        for odb, objs in self.get_used_objs().items():
-            self.repo.cloud.pull(objs, jobs=jobs, odb=odb)
+        rev_lock = self.def_repo.get(self.PARAM_REV_LOCK)
+        if rev_lock:
+            repo[self.PARAM_REV_LOCK] = rev_lock
 
-        obj = self.get_obj()
-        checkout(
-            to.fs_path,
-            to.fs,
-            obj,
-            self.repo.odb.local,
-            ignore=None,
-            state=self.repo.state,
-            prompt=confirm,
-        )
+        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:
             self.def_repo[self.PARAM_REV] = rev
-        with self._make_repo(locked=False) as repo:
-            self.def_repo[self.PARAM_REV_LOCK] = repo.get_rev()
+        self.fs = self._make_fs(rev=rev, locked=False)
+        self.def_repo[self.PARAM_REV_LOCK] = self.fs.repo.get_rev()
 
     def changed_checksum(self) -> bool:
         # From current repo point of view what describes RepoDependency is its
@@ -98,124 +99,37 @@ class RepoDependency(Dependency):
         # immutable, hence its impossible for checksum to change.
         return False
 
-    def get_used_objs(self, **kwargs) -> Dict[Optional["ObjectDB"], Set["HashInfo"]]:
-        used, _, _ = self._get_used_and_obj(**kwargs)
-        return used
+    def _make_fs(
+        self, rev: Optional[str] = None, locked: bool = True
+    ) -> "DVCFileSystem":
+        from dvc.config import Config
+        from dvc.fs import DVCFileSystem
 
-    def _get_used_and_obj(
-        self, obj_only: bool = False, **kwargs
-    ) -> Tuple[Dict[Optional["ObjectDB"], Set["HashInfo"]], "Meta", "HashFile"]:
-        from dvc.config import NoRemoteError
-        from dvc.exceptions import NoOutputOrStageError, PathMissingError
-        from dvc.utils import as_posix
-        from dvc_data.hashfile.build import build
-        from dvc_data.hashfile.tree import Tree, TreeError
+        rem = self.def_repo.get("remote")
+        if isinstance(rem, dict):
+            remote = None  # type: ignore[unreachable]
+            remote_config = rem
+        else:
+            remote = rem
+            remote_config = None
 
-        local_odb = self.repo.odb.local
-        locked = kwargs.pop("locked", True)
-        with self._make_repo(locked=locked, cache_dir=local_odb.path) as repo:
-            used_obj_ids = defaultdict(set)
-            rev = repo.get_rev()
-            if locked and self.def_repo.get(self.PARAM_REV_LOCK) is None:
-                self.def_repo[self.PARAM_REV_LOCK] = rev
+        conf = self.def_repo.get("config")
+        if conf:
+            config = Config.load_file(conf)
+        else:
+            config = {}
 
-            if not obj_only:
-                try:
-                    for odb, obj_ids in repo.used_objs(
-                        [os.path.join(repo.root_dir, self.def_path)],
-                        force=True,
-                        jobs=kwargs.get("jobs"),
-                        recursive=True,
-                    ).items():
-                        if odb is None:
-                            odb = repo.cloud.get_remote_odb()
-                            odb.read_only = True
-                        self._check_circular_import(odb, obj_ids)
-                        used_obj_ids[odb].update(obj_ids)
-                except (NoRemoteError, NoOutputOrStageError):
-                    pass  # noqa: S110
+        config["cache"] = self.repo.config["cache"]
+        config["cache"]["dir"] = self.repo.cache.local_cache_dir
 
-            try:
-                object_store, meta, obj = build(
-                    local_odb,
-                    as_posix(self.def_path),
-                    repo.dvcfs,
-                    local_odb.fs.PARAM_CHECKSUM,
-                )
-            except (FileNotFoundError, TreeError) as exc:
-                raise PathMissingError(
-                    self.def_path, self.def_repo[self.PARAM_URL]
-                ) from exc
-            object_store = copy(object_store)
-            object_store.read_only = True
-
-            self._objs[rev] = obj
-            self._meta[rev] = meta
-
-            used_obj_ids[object_store].add(obj.hash_info)
-            if isinstance(obj, Tree):
-                used_obj_ids[object_store].update(oid for _, _, oid in obj)
-            return used_obj_ids, meta, obj
-
-    def _check_circular_import(self, odb: "ObjectDB", obj_ids: Set["HashInfo"]) -> None:
-        from dvc.exceptions import CircularImportError
-        from dvc.fs.dvc import DVCFileSystem
-        from dvc_data.hashfile.db.reference import ReferenceHashFileDB
-        from dvc_data.hashfile.tree import Tree
-
-        if not isinstance(odb, ReferenceHashFileDB):
-            return
-
-        def iter_objs():
-            for hash_info in obj_ids:
-                if hash_info.isdir:
-                    tree = Tree.load(odb, hash_info)
-                    yield from (odb.get(hi.value) for _, _, hi in tree)
-                else:
-                    assert hash_info.value
-                    yield odb.get(hash_info.value)
-
-        checked_urls = set()
-        for obj in iter_objs():
-            if not isinstance(obj.fs, DVCFileSystem):
-                continue
-            if obj.fs.repo_url in checked_urls or obj.fs.repo.root_dir in checked_urls:
-                continue
-            self_url = self.repo.url or self.repo.root_dir
-            if (
-                obj.fs.repo_url is not None
-                and obj.fs.repo_url == self_url
-                or obj.fs.repo.root_dir == self.repo.root_dir
-            ):
-                raise CircularImportError(self, obj.fs.repo_url, self_url)
-            checked_urls.update([obj.fs.repo_url, obj.fs.repo.root_dir])
-
-    def get_obj(self, filter_info=None, **kwargs):
-        locked = kwargs.get("locked", True)
-        rev = self._get_rev(locked=locked)
-        if rev in self._objs:
-            return self._objs[rev]
-        _, _, obj = self._get_used_and_obj(
-            obj_only=True, filter_info=filter_info, **kwargs
+        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,
         )
-        return obj
-
-    def get_meta(self, filter_info=None, **kwargs):
-        locked = kwargs.get("locked", True)
-        rev = self._get_rev(locked=locked)
-        if rev in self._meta:
-            return self._meta[rev]
-        _, meta, _ = self._get_used_and_obj(
-            obj_only=True, filter_info=filter_info, **kwargs
-        )
-        return meta
-
-    def _make_repo(self, locked: bool = True, **kwargs) -> "ContextManager[Repo]":
-        from dvc.external_repo import external_repo
-
-        d = self.def_repo
-        rev = self._get_rev(locked=locked)
-        return external_repo(d[self.PARAM_URL], rev=rev, **kwargs)
 
     def _get_rev(self, locked: bool = True):
         d = self.def_repo

dvc/dirs.py

@@ -0,0 +1,26 @@
+import os
+
+import platformdirs
+
+from . import env
+
+APPNAME = "dvc"
+APPAUTHOR = "iterative"
+
+
+def system_config_dir():
+    return os.getenv(env.DVC_SYSTEM_CONFIG_DIR) or platformdirs.site_config_dir(
+        APPNAME, APPAUTHOR
+    )
+
+
+def global_config_dir():
+    return os.getenv(env.DVC_SYSTEM_CONFIG_DIR) or platformdirs.user_config_dir(
+        APPNAME, APPAUTHOR
+    )
+
+
+def site_cache_dir():
+    return os.getenv(env.DVC_SITE_CACHE_DIR) or platformdirs.site_cache_dir(
+        APPNAME, APPAUTHOR, opinion=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)
@@ -187,11 +180,12 @@ class FileMixin:
 
 class SingleStageFile(FileMixin):
     from dvc.schema import COMPILED_SINGLE_STAGE_SCHEMA as SCHEMA
-    from dvc.stage.loader import SingleStageLoader as LOADER
+    from dvc.stage.loader import SingleStageLoader as LOADER  # noqa: N814
 
     metrics: List[str] = []
     plots: Any = {}
     params: List[str] = []
+    artifacts: Dict[str, Optional[Dict[str, Any]]] = {}
 
     @property
     def stage(self) -> "Stage":
@@ -231,7 +225,7 @@ class ProjectFile(FileMixin):
     """Abstraction for pipelines file, .yaml + .lock combined."""
 
     from dvc.schema import COMPILED_MULTI_STAGE_SCHEMA as SCHEMA
-    from dvc.stage.loader import StageLoader as LOADER
+    from dvc.stage.loader import StageLoader as LOADER  # noqa: N814
 
     @property
     def _lockfile(self):
@@ -323,6 +317,10 @@ class ProjectFile(FileMixin):
     def params(self) -> List[str]:
         return self.contents.get("params", [])
 
+    @property
+    def artifacts(self) -> Dict[str, Optional[Dict[str, Any]]]:
+        return self.contents.get("artifacts", {})
+
     def remove(self, force=False):
         if not force:
             logger.warning("Cannot remove pipeline file.")
@@ -353,36 +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 = {k: v for k, v in d.items()}
-
-    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
@@ -396,24 +366,14 @@ 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)
-            else:
-                if not data:
-                    data.update(self.latest_version_info)
-                    # order is important, meta should always be at the top
-                    logger.info("Generating lock file '%s'", self.relpath)
+            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)
 
             data["stages"] = data.get("stages", {})
             modified = data["stages"].get(stage.name, {}) != stage_data.get(
@@ -432,8 +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

@@ -1,11 +1,17 @@
-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_EXP_GIT_REMOTE = "DVC_EXP_GIT_REMOTE"
+DVC_EXP_NAME = "DVC_EXP_NAME"
+DVC_GLOBAL_CONFIG_DIR = "DVC_GLOBAL_CONFIG_DIR"
+DVC_IGNORE_ISATTY = "DVC_IGNORE_ISATTY"
+DVC_NO_ANALYTICS = "DVC_NO_ANALYTICS"
 DVC_PAGER = "DVC_PAGER"
 DVC_ROOT = "DVC_ROOT"
-DVCLIVE_RESUME = "DVCLIVE_RESUME"
-DVC_IGNORE_ISATTY = "DVC_IGNORE_ISATTY"
-DVC_EXP_BASELINE_REV = "DVC_EXP_BASELINE_REV"
-DVC_EXP_NAME = "DVC_EXP_NAME"
-DVC_EXP_GIT_REMOTE = "DVC_EXP_GIT_REMOTE"
-DVC_EXP_AUTO_PUSH = "DVC_EXP_AUTO_PUSH"
-DVC_NO_ANALYTICS = "DVC_NO_ANALYTICS"
+DVC_SHOW_TRACEBACK = "DVC_SHOW_TRACEBACK"
+DVC_SITE_CACHE_DIR = "DVC_SITE_CACHE_DIR"
+DVC_STUDIO_OFFLINE = "DVC_STUDIO_OFFLINE"
+DVC_STUDIO_REPO_URL = "DVC_STUDIO_REPO_URL"
+DVC_STUDIO_TOKEN = "DVC_STUDIO_TOKEN"  # noqa: S105 # nosec B105
+DVC_STUDIO_URL = "DVC_STUDIO_URL"
+DVC_SYSTEM_CONFIG_DIR = "DVC_SYSTEM_CONFIG_DIR"

dvc/exceptions.py

@@ -1,4 +1,5 @@
 """Exceptions raised by the dvc."""
+import errno
 from typing import Dict, List
 
 from dvc.utils import format_link
@@ -16,6 +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
@@ -31,14 +36,19 @@ class OutputDuplicationError(DvcException):
 
         assert isinstance(output, str)
         assert all(hasattr(stage, "relpath") for stage in stages)
+        msg = ""
+        stage_names = [s.addressing for s in stages]
+        stages_str = " ".join(stage_names)
         if len(stages) == 1:
-            msg = "output '{}' is already specified in {}.".format(
-                output, first(stages)
-            )
+            stage_name = first(stages)
+            msg = f"output '{output}' is already specified in {stage_name}."
         else:
             msg = "output '{}' is already specified in stages:\n{}".format(
-                output, "\n".join(f"\t- {s.addressing}" for s in stages)
+                output, "\n".join(f"\t- {s}" for s in stage_names)
             )
+        msg += (
+            f"\nUse `dvc remove {stages_str}` to stop tracking the overlapping output."
+        )
         super().__init__(msg)
         self.stages = stages
         self.output = output
@@ -57,12 +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.
@@ -74,11 +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):
@@ -150,9 +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."
         )
 
 
@@ -161,9 +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):
@@ -176,11 +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
@@ -201,6 +200,18 @@ class ETagMismatchError(DvcException):
         )
 
 
+class FileExistsLocallyError(FileExistsError, DvcException):
+    def __init__(self, path, hint=None):
+        import os.path
+
+        self.path = path
+        hint = "" if hint is None else f". {hint}"
+        path_typ = "directory" if os.path.isdir(path) else "file"
+        msg = f"The {path_typ} '{path}' already exists locally{hint}"
+        super().__init__(msg)
+        self.errno = errno.EEXIST
+
+
 class FileMissingError(DvcException):
     def __init__(self, path, hint=None):
         self.path = path
@@ -286,22 +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."""
 
 
@@ -333,14 +329,6 @@ class CacheLinkError(DvcException):
         self.fs_paths = fs_paths
 
 
-class CircularImportError(DvcException):
-    def __init__(self, dep, a, b):
-        super().__init__(
-            f"'{dep}' contains invalid circular import. "
-            f"DVC repo '{a}' already imports from '{b}'."
-        )
-
-
 class PrettyDvcException(DvcException):
     def __pretty_exc__(self, **kwargs):
         """Print prettier exception message."""

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
@@ -6,7 +7,6 @@ from dvc.config import ConfigError as RepoConfigError
 from dvc.config_schema import SCHEMA, Invalid
 
 # pylint: disable=unused-import
-from dvc_objects.fs import utils  # noqa: F401
 from dvc_objects.fs import (  # noqa: F401
     LocalFileSystem,
     MemoryFileSystem,
@@ -17,8 +17,9 @@ from dvc_objects.fs import (  # noqa: F401
     localfs,
     registry,
     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
 
@@ -105,23 +133,19 @@ 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)
     try:
-        remote_conf = SCHEMA["remote"][str](remote_conf)
+        remote_conf = SCHEMA["remote"][str](remote_conf)  # type: ignore[index]
     except Invalid as exc:
         raise RepoConfigError(str(exc)) from None
 

dvc/fs/callbacks.py

@@ -1,13 +1,10 @@
 # pylint: disable=unused-import
 from contextlib import ExitStack
-from typing import TYPE_CHECKING, Any, BinaryIO, Dict, Optional, Union, cast
+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,
@@ -54,7 +104,8 @@ class RichCallback(Callback):
             disable=self.disable,
             console=ui.error_console,
         )
-        return cast(RichTransferProgress, self._stack.enter_context(progress))
+        self._stack.enter_context(progress)
+        return progress
 
     @cached_property
     def task(self) -> "TaskID":

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
 
@@ -251,33 +261,39 @@ class _DVCFileSystem(AbstractFileSystem):  # pylint:disable=abstract-method
         try:
             return self.repo.fs.open(fs_path, mode=mode)
         except FileNotFoundError:
-            _, dvc_fs, subkey = self._get_subrepo_info(key)
+            repo, dvc_fs, subkey = self._get_subrepo_info(key)
             if not dvc_fs:
                 raise
 
         dvc_path = _get_dvc_path(dvc_fs, subkey)
-        return dvc_fs.open(dvc_path, mode=mode)
+        kw = {}
+        if kwargs.get("cache_remote_stream", False):
+            kw["cache_odb"] = repo.cache.local
+        return dvc_fs.open(dvc_path, mode=mode, **kw)
 
     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
+    def ls(  # pylint: disable=arguments-differ # noqa: C901
         self, path, detail=True, dvc_only=False, **kwargs
     ):
         key = self._get_key_from_relative(path)
         repo, dvc_fs, subkey = self._get_subrepo_info(key)
 
+        dvc_exists = False
         dvc_infos = {}
         if dvc_fs:
-            with suppress(FileNotFoundError):
-                dvc_path = _get_dvc_path(dvc_fs, subkey)
+            dvc_path = _get_dvc_path(dvc_fs, subkey)
+            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)
 
+        fs_exists = False
         fs_infos = {}
         ignore_subrepos = kwargs.get("ignore_subrepos", True)
         if not dvc_only:
@@ -289,19 +305,28 @@ class _DVCFileSystem(AbstractFileSystem):  # pylint:disable=abstract-method
                 ):
                     fs_infos[fs.path.name(info["name"])] = info
             except (FileNotFoundError, NotADirectoryError):
-                pass  # noqa: S110
+                pass
+
+            fs_exists = bool(fs_infos) or fs.exists(fs_path)
 
         dvcfiles = kwargs.get("dvcfiles", False)
 
         infos = []
         paths = []
         names = set(dvc_infos.keys()) | set(fs_infos.keys())
+
+        if not names and (dvc_exists or fs_exists):
+            # broken symlink or TreeError
+            raise FileNotFoundError(errno.ENOENT, os.strerror(errno.ENOENT), path)
+
         for name in names:
             if not dvcfiles and _is_dvc_file(name):
                 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)
@@ -316,7 +341,9 @@ class _DVCFileSystem(AbstractFileSystem):  # pylint:disable=abstract-method
         ignore_subrepos = kwargs.get("ignore_subrepos", True)
         return self._info(key, path, ignore_subrepos=ignore_subrepos)
 
-    def _info(self, key, path, ignore_subrepos=True, check_ignored=True):  # noqa: C901
+    def _info(  # noqa: C901, PLR0912
+        self, key, path, ignore_subrepos=True, check_ignored=True
+    ):
         repo, dvc_fs, subkey = self._get_subrepo_info(key)
 
         dvc_info = None
@@ -325,7 +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
@@ -355,10 +382,26 @@ 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
 
+    def get_file(self, rpath, lpath, **kwargs):  # pylint: disable=arguments-differ
+        key = self._get_key_from_relative(rpath)
+        fs_path = self._from_key(key)
+        try:
+            return self.repo.fs.get_file(fs_path, lpath, **kwargs)
+        except FileNotFoundError:
+            _, dvc_fs, subkey = self._get_subrepo_info(key)
+            if not dvc_fs:
+                raise
+
+        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"
@@ -369,7 +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:
@@ -392,3 +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

@@ -6,20 +6,18 @@ 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",
     "dvc_objects",
     "dvc_render",
     "dvc_task",
-    "dvclive",
     "scmrepo",
 )
 package = "" if PKG is None else f" ({PKG})"
@@ -33,6 +31,7 @@ def get_dvc_info():
         f"Platform: Python {platform.python_version()} on {platform.platform()}",
         f"Subprojects:{_get_subprojects()}",
         f"Supports:{_get_supported_remotes()}",
+        f"Config:{_get_config_dirs()}",
     ]
 
     try:
@@ -41,20 +40,21 @@ def get_dvc_info():
             # can't auto-create it, as it might cause issues if the user
             # later decides to enable shared cache mode with
             # `dvc config cache.shared group`.
-            if os.path.exists(repo.odb.local.path):
+            if os.path.exists(repo.cache.local.path):
                 info.append(f"Cache types: {_get_linktype_support_info(repo)}")
-                fs_type = get_fs_type(repo.odb.local.path)
+                fs_type = _get_fs_type(repo.cache.local.path)
                 info.append(f"Cache directory: {fs_type}")
             else:
                 info.append("Cache types: " + error_link("no-dvc-cache"))
 
-            info.append(f"Caches: {_get_caches(repo.odb)}")
+            info.append(f"Caches: {_get_caches(repo.cache)}")
             info.append(f"Remotes: {_get_remotes(repo.config)}")
 
             root_directory = repo.root_dir
-            fs_root = get_fs_type(os.path.abspath(root_directory))
+            fs_root = _get_fs_type(os.path.abspath(root_directory))
             info.append(f"Workspace directory: {fs_root}")
             info.append(f"Repo: {_get_dvc_repo_info(repo)}")
+            info.append(f"Repo.site_cache_dir: {repo.site_cache_dir}")
     except NotDvcRepoError:
         pass
     except SCMError:
@@ -67,7 +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
@@ -84,7 +84,7 @@ def _get_remotes(config):
 
 
 def _get_linktype_support_info(repo):
-    odb = repo.odb.local
+    odb = repo.cache.local
 
     links = generic.test_links(
         ["reflink", "hardlink", "symlink"],
@@ -136,10 +136,21 @@ def _get_supported_remotes():
     return "\n\t" + ",\n\t".join(supported_remotes)
 
 
-def get_fs_type(path):
+def _get_config_dirs():
+    from dvc.config import Config
+
+    dirs = [
+        f"Global: {Config.get_dir('global')}",
+        f"System: {Config.get_dir('system')}",
+    ]
+
+    return "\n\t" + "\n\t".join(dirs)
+
+
+def _get_fs_type(path):
     partition = {}
     for part in psutil.disk_partitions(all=True):
-        if part.fstype != "":
+        if part.fstype:
             try:
                 mountpoint = pathlib.Path(part.mountpoint).resolve()
                 partition[mountpoint] = part.fstype + " on " + part.device
@@ -155,11 +166,11 @@ def get_fs_type(path):
     return ("unknown", "none")
 
 
-def _get_dvc_repo_info(self):
-    if self.config.get("core", {}).get("no_scm", False):
+def _get_dvc_repo_info(repo):
+    if repo.config.get("core", {}).get("no_scm", False):
         return "dvc (no_scm)"
 
-    if self.root_dir != self.scm.root_dir:
+    if repo.root_dir != repo.scm.root_dir:
         return "dvc (subdir), git"
 
     return "dvc, git"

dvc/lock.py

@@ -187,7 +187,9 @@ class HardlinkLock(flufl.lock.Lock, LockBase):
 
         if self._tmp_dir is not None:
             # Under Windows file path length is limited so we hash it
-            hasher = hashlib.md5(self._claimfile.encode())  # noqa: S324
+            hasher = hashlib.md5(  # nosec B324, B303  # noqa: S324
+                self._claimfile.encode()
+            )
             filename = hasher.hexdigest()
             self._claimfile = os.path.join(self._tmp_dir, filename + ".lock")
 

dvc/logger.py

@@ -1,77 +1,70 @@
 """Manages logging configuration for DVC repo."""
 
+import logging
 import logging.config
 import logging.handlers
+import os
+import sys
 
 import colorama
 
+from dvc.env import DVC_SHOW_TRACEBACK
 from dvc.progress import Tqdm
 
-FOOTER = (
-    "\n{yellow}Having any troubles?{nc}"
-    " Hit us up at {blue}https://dvc.org/support{nc},"
-    " we are always happy to help!"
-).format(
-    blue=colorama.Fore.BLUE,
-    nc=colorama.Fore.RESET,
-    yellow=colorama.Fore.YELLOW,
-)
 
-
-def addLoggingLevel(levelName, levelNum, methodName=None):
+def add_logging_level(level_name, level_num, method_name=None):
     """
     Adds a new logging level to the `logging` module and the
     currently configured logging class.
 
-    Uses the existing numeric levelNum if already defined.
+    Uses the existing numeric level_num if already defined.
 
     Based on https://stackoverflow.com/questions/2183233
     """
-    if methodName is None:
-        methodName = levelName.lower()
+    if method_name is None:
+        method_name = level_name.lower()
 
     # If the level name is already defined as a top-level `logging`
     # constant, then adopt the existing numeric level.
-    if hasattr(logging, levelName):
-        existingLevelNum = getattr(logging, levelName)
-        assert isinstance(existingLevelNum, int)
-        levelNum = existingLevelNum
+    if hasattr(logging, level_name):
+        existing_level_num = getattr(logging, level_name)
+        assert isinstance(existing_level_num, int)
+        level_num = existing_level_num
 
-    def logForLevel(self, message, *args, **kwargs):
-        if self.isEnabledFor(levelNum):
+    def log_for_level(self, message, *args, **kwargs):
+        if self.isEnabledFor(level_num):
             # pylint: disable=protected-access
-            self._log(levelNum, message, args, **kwargs)
+            self._log(level_num, message, args, **kwargs)
 
-    def logToRoot(message, *args, **kwargs):
-        logging.log(levelNum, message, *args, **kwargs)
+    def log_to_root(message, *args, **kwargs):
+        logging.log(level_num, message, *args, **kwargs)
 
     # getLevelName resolves the numeric log level if already defined,
     # otherwise returns a string
-    if not isinstance(logging.getLevelName(levelName), int):
-        logging.addLevelName(levelNum, levelName)
+    if not isinstance(logging.getLevelName(level_name), int):
+        logging.addLevelName(level_num, level_name)
 
-    if not hasattr(logging, levelName):
-        setattr(logging, levelName, levelNum)
+    if not hasattr(logging, level_name):
+        setattr(logging, level_name, level_num)
 
-    if not hasattr(logging.getLoggerClass(), methodName):
-        setattr(logging.getLoggerClass(), methodName, logForLevel)
+    if not hasattr(logging.getLoggerClass(), method_name):
+        setattr(logging.getLoggerClass(), method_name, log_for_level)
 
-    if not hasattr(logging, methodName):
-        setattr(logging, methodName, logToRoot)
+    if not hasattr(logging, method_name):
+        setattr(logging, method_name, log_to_root)
 
 
 class LoggingException(Exception):
     def __init__(self, record):
-        msg = f"failed to log {str(record)}"
+        msg = f"failed to log {record!s}"
         super().__init__(msg)
 
 
-def excludeFilter(level):
-    class ExcludeLevelFilter(logging.Filter):
-        def filter(self, record):  # noqa: A003
-            return record.levelno < level
+def exclude_filter(level: int):
+    def filter_fn(record: "logging.LogRecord") -> bool:
+        return record.levelno < level
 
-    return ExcludeLevelFilter
+    return filter_fn
 
 
 class ColorFormatter(logging.Formatter):
@@ -86,7 +79,8 @@ class ColorFormatter(logging.Formatter):
     exception cause, just the message and the traceback.
     """
 
-    color_code = {
+    reset = colorama.Fore.RESET
+    color_codes = {
         "TRACE": colorama.Fore.GREEN,
         "DEBUG": colorama.Fore.BLUE,
         "WARNING": colorama.Fore.YELLOW,
@@ -94,51 +88,53 @@ class ColorFormatter(logging.Formatter):
         "CRITICAL": colorama.Fore.RED,
     }
 
-    def format(self, record):  # noqa: A003
+    def __init__(self, log_colors: bool = True, show_traceback: bool = False) -> None:
+        super().__init__()
+        self.log_colors = log_colors
+        self.show_traceback = show_traceback
+
+    def format(self, record) -> str:  # noqa: A003, C901
         record.message = record.getMessage()
         msg = self.formatMessage(record)
 
-        if record.levelname == "INFO":
+        if record.levelno == logging.INFO:
             return msg
 
-        if record.exc_info:
-            if getattr(record, "tb_only", False):
-                cause = ""
-            else:
-                cause = ": ".join(_iter_causes(record.exc_info[1]))
+        ei = record.exc_info
+        if ei:
+            cause = ""
+            if not getattr(record, "tb_only", False):
+                cause = ": ".join(_iter_causes(ei[1]))
+            sep = " - " if msg and cause else ""
+            msg = msg + sep + cause
 
-            msg = "{message}{separator}{cause}".format(
-                message=msg or "",
-                separator=" - " if msg and cause else "",
-                cause=cause,
-            )
+        asctime = ""
+        verbose = _is_verbose()
+        if verbose:
+            asctime = self.formatTime(record, self.datefmt)
+        if verbose or self.show_traceback:
+            if ei and not record.exc_text:
+                record.exc_text = self.formatException(ei)
+            if record.exc_text:
+                if msg[-1:] != "\n":
+                    msg = msg + "\n"
+                msg = msg + record.exc_text + "\n"
+            if record.stack_info:
+                if msg[-1:] != "\n":
+                    msg = msg + "\n"
+                msg = msg + self.formatStack(record.stack_info) + "\n"
 
-            if _is_verbose():
-                msg += _stack_trace(record.exc_info)
-
-        return "{asctime}{color}{levelname}{nc}: {msg}".format(
-            asctime=self.formatTime(record, self.datefmt),
-            color=self.color_code[record.levelname],
-            nc=colorama.Fore.RESET,
-            levelname=record.levelname,
-            msg=msg,
-        )
-
-    def formatTime(self, record, datefmt=None):
-        # only show if current level is set to DEBUG
-        # also, skip INFO as it is used for UI
-        if not _is_verbose() or record.levelno == logging.INFO:
-            return ""
-
-        return "{green}{date}{nc} ".format(
-            green=colorama.Fore.GREEN,
-            date=super().formatTime(record, datefmt),
-            nc=colorama.Fore.RESET,
-        )
+        level = record.levelname
+        if self.log_colors:
+            color = self.color_codes[level]
+            if asctime:
+                asctime = color + asctime + self.reset
+            level = color + level + self.reset
+        return asctime + (" " if asctime else "") + level + ": " + msg
 
 
 class LoggerHandler(logging.StreamHandler):
-    def handleError(self, record):
+    def handleError(self, record):  # noqa: N802
         super().handleError(record)
         raise LoggingException(record)
 
@@ -155,15 +151,16 @@ class LoggerHandler(logging.StreamHandler):
                         self.emit_pretty_exception(exc, verbose=_is_verbose())
                         if not _is_verbose():
                             return
-                    except Exception:  # noqa, pylint: disable=broad-except
-                        pass  # noqa: S110
+                    # pylint: disable-next=broad-except
+                    except Exception:  # noqa: BLE001, S110  # nosec B110
+                        pass
 
             msg = self.format(record)
             Tqdm.write(msg, file=self.stream, end=getattr(self, "terminator", "\n"))
             self.flush()
         except (BrokenPipeError, RecursionError):
             raise
-        except Exception:  # noqa, pylint: disable=broad-except
+        except Exception:  # noqa: BLE001, pylint: disable=broad-except
             self.handleError(record)
 
 
@@ -179,31 +176,47 @@ def _iter_causes(exc):
         exc = exc.__cause__
 
 
-def _stack_trace(exc_info):
-    import traceback
-
-    return "\n{red}{line}{nc}\n{trace}{red}{line}{nc}".format(
-        red=colorama.Fore.RED,
-        line="-" * 60,
-        trace="".join(traceback.format_exception(*exc_info)),
-        nc=colorama.Fore.RESET,
-    )
-
-
-def disable_other_loggers():
-    logging.captureWarnings(True)
-    loggerDict = logging.root.manager.loggerDict  # pylint: disable=no-member
-    for logger_name, logger in loggerDict.items():
-        if logger_name != "dvc" and not logger_name.startswith("dvc."):
-            logger.disabled = True  # type: ignore[union-attr]
-
-
 def set_loggers_level(level: int = logging.INFO) -> None:
     for name in ["dvc", "dvc_objects", "dvc_data"]:
         logging.getLogger(name).setLevel(level)
 
 
-def setup(level: int = logging.INFO) -> None:
+def setup(level: int = logging.INFO, log_colors: bool = True) -> None:
+    colorama.init()
+
+    formatter = ColorFormatter(log_colors=log_colors and sys.stdout.isatty())
+
+    console_info = LoggerHandler(sys.stdout)
+    console_info.setLevel(logging.INFO)
+    console_info.setFormatter(formatter)
+    console_info.addFilter(exclude_filter(logging.WARNING))
+
+    console_debug = LoggerHandler(sys.stdout)
+    console_debug.setLevel(logging.DEBUG)
+    console_debug.setFormatter(formatter)
+    console_debug.addFilter(exclude_filter(logging.INFO))
+
+    add_logging_level("TRACE", logging.DEBUG - 5)
+
+    console_trace = LoggerHandler(sys.stdout)
+    console_trace.setLevel(logging.TRACE)  # type: ignore[attr-defined]
+    console_trace.setFormatter(formatter)
+    console_trace.addFilter(exclude_filter(logging.DEBUG))
+
+    show_traceback = bool(os.environ.get(DVC_SHOW_TRACEBACK))
+    err_formatter = ColorFormatter(
+        log_colors=log_colors and sys.stderr.isatty(), show_traceback=show_traceback
+    )
+    console_errors = LoggerHandler(sys.stderr)
+    console_errors.setLevel(logging.WARNING)
+    console_errors.setFormatter(err_formatter)
+
+    for name in ["dvc", "dvc_objects", "dvc_data"]:
+        logger = logging.getLogger(name)
+        logger.setLevel(level)
+        for handler in [console_info, console_debug, console_trace, console_errors]:
+            logger.addHandler(handler)
+
     if level >= logging.DEBUG:
         # Unclosed session errors for asyncio/aiohttp are only available
         # on the tracing mode for extensive debug purposes. They are really
@@ -214,80 +227,3 @@ def setup(level: int = logging.INFO) -> None:
         # action.
         logging.getLogger("asyncio").setLevel(logging.CRITICAL)
         logging.getLogger("aiohttp").setLevel(logging.CRITICAL)
-
-    addLoggingLevel("TRACE", logging.DEBUG - 5)
-    logging.config.dictConfig(
-        {
-            "version": 1,
-            "filters": {
-                "exclude_errors": {"()": excludeFilter(logging.WARNING)},
-                "exclude_info": {"()": excludeFilter(logging.INFO)},
-                "exclude_debug": {"()": excludeFilter(logging.DEBUG)},
-            },
-            "formatters": {"color": {"()": ColorFormatter}},
-            "handlers": {
-                "console_info": {
-                    "class": "dvc.logger.LoggerHandler",
-                    "level": "INFO",
-                    "formatter": "color",
-                    "stream": "ext://sys.stdout",
-                    "filters": ["exclude_errors"],
-                },
-                "console_debug": {
-                    "class": "dvc.logger.LoggerHandler",
-                    "level": "DEBUG",
-                    "formatter": "color",
-                    "stream": "ext://sys.stdout",
-                    "filters": ["exclude_info"],
-                },
-                "console_trace": {
-                    "class": "dvc.logger.LoggerHandler",
-                    "level": "TRACE",
-                    "formatter": "color",
-                    "stream": "ext://sys.stdout",
-                    "filters": ["exclude_debug"],
-                },
-                "console_errors": {
-                    "class": "dvc.logger.LoggerHandler",
-                    "level": "WARNING",
-                    "formatter": "color",
-                    "stream": "ext://sys.stderr",
-                },
-            },
-            "loggers": {
-                "dvc": {
-                    "level": level,
-                    "handlers": [
-                        "console_info",
-                        "console_debug",
-                        "console_trace",
-                        "console_errors",
-                    ],
-                },
-                "dvc_objects": {
-                    "level": level,
-                    "handlers": [
-                        "console_info",
-                        "console_debug",
-                        "console_trace",
-                        "console_errors",
-                    ],
-                },
-                "dvc_data": {
-                    "level": level,
-                    "handlers": [
-                        "console_info",
-                        "console_debug",
-                        "console_trace",
-                        "console_errors",
-                    ],
-                },
-            },
-            "disable_existing_loggers": False,
-        }
-    )
-
-    # NOTE: colorama init needs to go after logging config to avoid potential
-    # conflicts with libs that modify logging (like tensorflow/abseil)
-    # https://github.com/iterative/dvc/issues/8387
-    colorama.init()

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/odbmgr.py

@@ -1,71 +0,0 @@
-from dvc.fs import GitFileSystem, Schemes
-from dvc_data.hashfile.db import get_odb
-
-
-def _get_odb(repo, settings, fs=None):
-    from dvc.fs import get_cloud_fs
-
-    if not settings:
-        return None
-
-    cls, config, fs_path = get_cloud_fs(repo, **settings)
-    config["tmp_dir"] = repo.tmp_dir
-    fs = fs or cls(**config)
-    return get_odb(fs, fs_path, state=repo.state, **config)
-
-
-class ODBManager:
-    CACHE_DIR = "cache"
-    CLOUD_SCHEMES = [
-        Schemes.S3,
-        Schemes.GS,
-        Schemes.SSH,
-        Schemes.HDFS,
-        Schemes.WEBHDFS,
-    ]
-
-    def __init__(self, repo):
-        self._repo = repo
-        self.config = config = repo.config["cache"]
-        self._odb = {}
-
-        local = config.get("local")
-
-        if local:
-            settings = {"name": local}
-        elif "dir" not in config:
-            settings = None
-        else:
-            from dvc.config_schema import LOCAL_COMMON
-
-            settings = {"url": config["dir"]}
-            for opt in LOCAL_COMMON:
-                if opt in config:
-                    settings[str(opt)] = config.get(opt)
-
-        kwargs = {}
-        if not isinstance(repo.fs, GitFileSystem):
-            kwargs["fs"] = repo.fs
-
-        odb = _get_odb(repo, settings, **kwargs)
-        self._odb["repo"] = odb
-        self._odb[Schemes.LOCAL] = 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)
-
-    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()