build(deps): bump github/codeql-action from 4.37.4 to 4.38.2

Bumps [github/codeql-action](https://github.com/github/codeql-action) from 4.37.4 to 4.38.2.
- [Release notes](https://github.com/github/codeql-action/releases)
- [Changelog](https://github.com/github/codeql-action/blob/main/CHANGELOG.md)
- [Commits](https://github.com/github/codeql-action/compare/v4.37.4...v4.38.2)

---
updated-dependencies:
- dependency-name: github/codeql-action
  dependency-version: 4.38.2
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <[email protected]>

.github/workflows/benchmarks.yaml

@@ -14,11 +14,11 @@ jobs:
     name: run benchmarks
     runs-on: ubuntu-latest
     steps:
-      - uses: actions/setup-python@v6
+      - uses: actions/setup-python@v7
         with:
             python-version: "3.14"
 
-      - uses: actions/checkout@v6
+      - uses: actions/checkout@v7
         with:
           ref: ${{ github.event.pull_request.base.sha }}
           fetch-depth: 0
@@ -32,7 +32,7 @@ jobs:
       - name: run benchmarks on base branch
         run: pytest --benchmark-autosave dvc/testing/benchmarks/ -k 'test_init or test_help'
 
-      - uses: actions/checkout@v6
+      - uses: actions/checkout@v7
         with:
           fetch-depth: 0
           clean: false

.github/workflows/build.yaml

@@ -19,11 +19,11 @@ jobs:
     name: Build & Verify package
     runs-on: ubuntu-latest
     steps:
-    - uses: actions/checkout@v6
+    - uses: actions/checkout@v7
       with:
         fetch-depth: 0
 
-    - uses: actions/setup-python@v6
+    - uses: actions/setup-python@v7
       with:
         python-version: "3.14"
 
@@ -41,7 +41,7 @@ jobs:
     - name: Check dist
       run: uv tool run twine check --strict dist/*
 
-    - uses: actions/upload-artifact@v6
+    - uses: actions/upload-artifact@v7
       with:
         name: Packages
         path: dist/
@@ -61,7 +61,7 @@ jobs:
       id-token: write
 
     steps:
-    - uses: actions/download-artifact@v7
+    - uses: actions/download-artifact@v8
       with:
         name: Packages
         path: dist
@@ -86,7 +86,7 @@ jobs:
       id-token: write
 
     steps:
-    - uses: actions/download-artifact@v7
+    - uses: actions/download-artifact@v8
       with:
         name: Packages
         path: dist

.github/workflows/codeql.yml

@@ -28,13 +28,13 @@ jobs:
 
     steps:
     - name: Checkout repository
-      uses: actions/checkout@v6
+      uses: actions/checkout@v7
 
     - name: Initialize CodeQL
-      uses: github/codeql-action/init@v4
+      uses: github/codeql-action/init@v4.38.2
       with:
         languages: ${{ matrix.language }}
         queries: security-extended
 
     - name: Perform CodeQL Analysis
-      uses: github/codeql-action/analyze@v4
+      uses: github/codeql-action/analyze@v4.38.2

.github/workflows/plugin_tests.yaml

@@ -23,18 +23,18 @@ jobs:
         plugin: ["dvc-s3"]
 
     steps:
-    - uses: actions/checkout@v6
+    - uses: actions/checkout@v7
       with:
         path: dvc
 
-    - uses: actions/checkout@v6
+    - uses: actions/checkout@v7
       with:
         repository: treeverse/${{ matrix.plugin }}
         ref: main
         path: ${{ matrix.plugin }}
 
     - name: Set up Python
-      uses: actions/setup-python@v6
+      uses: actions/setup-python@v7
       with:
         python-version: ${{ matrix.pyv }}
 

.github/workflows/tests.yaml

@@ -26,11 +26,11 @@ jobs:
       matrix:
         os: [ubuntu-latest, macos-latest, windows-latest]
     steps:
-    - uses: actions/checkout@v6
+    - uses: actions/checkout@v7
       with:
         fetch-depth: 0
     - name: Set up Python
-      uses: actions/setup-python@v6
+      uses: actions/setup-python@v7
       with:
         python-version: '3.14'
     - uses: astral-sh/setup-uv@v7
@@ -42,7 +42,7 @@ jobs:
         UV_LINK_MODE: ${{ runner.os == 'Windows' && 'symlink' || 'hardlink' }}
 
     - name: Cache mypy
-      uses: actions/cache@v5
+      uses: actions/cache@v6
       with:
         path: .mypy_cache
         key: mypy-${{ runner.os }}-${{ env.pythonLocation }}-${{ hashFiles('pyproject.toml') }}
@@ -130,11 +130,11 @@ jobs:
           pyv: "3.14"
           pytestargs: "--splitting-algorithm=least_duration --splits 4 --group 4"
     steps:
-    - uses: actions/checkout@v6
+    - uses: actions/checkout@v7
       with:
         fetch-depth: 0
     - name: Set up Python
-      uses: actions/setup-python@v6
+      uses: actions/setup-python@v7
       with:
         python-version: ${{ matrix.pyv }}
     - uses: astral-sh/setup-uv@v7
@@ -155,7 +155,7 @@ jobs:
         pytest ${{ matrix.pytestargs }} -n=logical --dist=worksteal --timeout=300 --durations=0
         --cov --cov-report=xml --cov-report=term  --durations-path=./.github/.test_durations
     - name: upload coverage report
-      uses: codecov/codecov-action@v5
+      uses: codecov/codecov-action@v7
   check:
     if: always()
     needs: [tests]
@@ -170,7 +170,7 @@ jobs:
     runs-on: ubuntu-latest
     steps:
     - name: Slack Notification
-      uses: rtCamp/action-slack-notify@v2.3.3
+      uses: rtCamp/action-slack-notify@v2.4.0
       env:
         SLACK_WEBHOOK: ${{ secrets.SLACK_WEBHOOK }}
         SLACK_COLOR: ${{ job.status }}

.pre-commit-config.yaml

@@ -22,13 +22,13 @@ repos:
       - id: sort-simple-yaml
       - id: trailing-whitespace
   - repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: 'v0.14.10'
+    rev: 'v0.15.20'
     hooks:
       - id: ruff-check
         args: [--fix, --exit-non-zero-on-fix]
       - id: ruff-format
   - repo: https://github.com/codespell-project/codespell
-    rev: v2.4.1
+    rev: v2.4.2
     hooks:
       - id: codespell
         additional_dependencies: ["tomli"]

dvc/__pyinstaller/hook-celery.py

@@ -8,7 +8,7 @@ from PyInstaller.utils.hooks import collect_submodules, is_module_or_submodule
 _EXCLUDES = ("celery.bin", "celery.contrib")
 hiddenimports = collect_submodules(
     "celery",
-    filter=lambda name: not any(
-        is_module_or_submodule(name, module) for module in _EXCLUDES
+    filter=lambda name: (
+        not any(is_module_or_submodule(name, module) for module in _EXCLUDES)
     ),
 )

dvc/__pyinstaller/hook-dvc.py

@@ -8,7 +8,6 @@ datas += copy_metadata("gcsfs")
 datas += copy_metadata("pyarrow")
 datas += copy_metadata("pydrive2")
 datas += copy_metadata("s3fs", recursive=True)
-datas += copy_metadata("boto3")
 datas += copy_metadata("ossfs")
 datas += copy_metadata("sshfs")
 datas += copy_metadata("webdav4")

dvc/analytics.py

@@ -147,7 +147,7 @@ def _git_remote_path_hash(scm: Optional["Base"]) -> Optional[str]:
             path = _parse_git_remote_path(remote_url)
             h = hashlib.md5(path.encode("utf-8"), usedforsecurity=False)  # for FIPS
             return h.hexdigest()
-    except Exception:  # noqa: BLE001
+    except Exception:
         logger.debug("Failed to get git remote path", exc_info=True)
     return None
 

dvc/commands/gc.py

@@ -40,9 +40,16 @@ class CmdGC(CmdBase):
             msg += " and all git commits"
         else:
             if self.args.all_branches and self.args.all_tags:
-                msg += " and all git branches and tags"
+                if self.args.num:
+                    msg += f" and last {self.args.num} commits from all git branches"
+                    msg += " and all git tags"
+                else:
+                    msg += " and all git branches and tags"
             elif self.args.all_branches:
-                msg += " and all git branches"
+                if self.args.num:
+                    msg += f" and last {self.args.num} commits from all git branches"
+                else:
+                    msg += " and all git branches"
             elif self.args.all_tags:
                 msg += " and all git tags"
             if self.args.commit_date:
@@ -126,8 +133,8 @@ def add_parser(subparsers, parent_parser):
         metavar="<num>",
         help=(
             "Keep data files used in the last `num` commits "
-            "starting from the `--rev` <commit>. "
-            "Only used if `--rev` is also provided. "
+            "starting from each selected revision root. "
+            "Can be used with `--rev` and `--all-branches`. "
             "Defaults to `1`."
         ),
     )

dvc/commands/ls/__init__.py

@@ -109,8 +109,6 @@ def _build_tree_structure(
 
 
 def show_tree(entries, with_color=False, with_size=False, with_hash=False):
-    import tabulate
-
     rows = _build_tree_structure(
         entries,
         with_color=with_color,
@@ -119,13 +117,7 @@ def show_tree(entries, with_color=False, with_size=False, with_hash=False):
     )
 
     colalign = ("right",) if with_size else None
-
-    _orig = tabulate.PRESERVE_WHITESPACE
-    tabulate.PRESERVE_WHITESPACE = True
-    try:
-        ui.table(rows, colalign=colalign)
-    finally:
-        tabulate.PRESERVE_WHITESPACE = _orig
+    ui.table(rows, colalign=colalign, preserve_whitespace=True)
 
 
 class CmdList(CmdBaseNoRepo):

dvc/config.py

@@ -41,18 +41,6 @@ class RemoteNotFoundError(RemoteConfigError):
     pass
 
 
-class MachineConfigError(ConfigError):
-    pass
-
-
-class NoMachineError(MachineConfigError):
-    pass
-
-
-class MachineNotFoundError(MachineConfigError):
-    pass
-
-
 @memoize
 def get_compiled_schema():
     from voluptuous import Schema

dvc/ignore.py

@@ -5,7 +5,15 @@ from collections.abc import Iterable, Iterator
 from itertools import chain, groupby, takewhile
 from typing import TYPE_CHECKING, Any, Literal, NamedTuple, Optional, Union, overload
 
-from pathspec.patterns import GitWildMatchPattern
+try:
+    from pathspec.patterns.gitignore.spec import (  # type: ignore[import-not-found]
+        GitIgnoreSpecPattern,
+    )
+except ImportError:  # pathspec<1
+    from pathspec.patterns import (
+        GitWildMatchPattern as GitIgnoreSpecPattern,
+    )
+
 from pathspec.util import normalize_file
 from pygtrie import Trie
 
@@ -34,7 +42,14 @@ class DvcIgnorePatterns(DvcIgnore):
     def __init__(
         self, pattern_list: Iterable[Union[PatternInfo, str]], dirname: str, sep: str
     ) -> None:
-        from pathspec.patterns.gitwildmatch import _DIR_MARK
+        try:
+            from pathspec.patterns.gitignore.spec import (  # type: ignore[import-not-found]
+                _DIR_MARK,
+            )
+        except ImportError:  # pathspec<1
+            from pathspec.patterns.gitwildmatch import (  # type: ignore[attr-defined, no-redef]
+                _DIR_MARK,
+            )
 
         pattern_infos = [
             pattern if isinstance(pattern, PatternInfo) else PatternInfo(pattern, "")
@@ -48,7 +63,7 @@ class DvcIgnorePatterns(DvcIgnore):
 
         regex_pattern_list: list[tuple[str, bool, bool, PatternInfo]] = []
         for count, pattern_info in enumerate(pattern_infos):
-            regex, ignore = GitWildMatchPattern.pattern_to_regex(pattern_info.patterns)
+            regex, ignore = GitIgnoreSpecPattern.pattern_to_regex(pattern_info.patterns)
             if regex is not None and ignore is not None:
                 self.pattern_list.append(pattern_info)
                 regex = regex.replace(f"<{_DIR_MARK}>", f"<{_DIR_MARK}{count}>")

dvc/parsing/interpolate.py

@@ -37,13 +37,13 @@ KEYCRE = re.compile(
 def get_parser():
     from pyparsing import CharsNotIn, ParserElement, Suppress, ZeroOrMore
 
-    ParserElement.enablePackrat()
+    ParserElement.enable_packrat()
 
     word = CharsNotIn(f"{PERIOD}{LBRACK}{RBRACK}")
     idx = Suppress(LBRACK) + word + Suppress(RBRACK)
     attr = Suppress(PERIOD) + word
     parser = word + ZeroOrMore(attr ^ idx)
-    parser.setParseAction(PERIOD.join)
+    parser.set_parse_action(PERIOD.join)
 
     return parser
 
@@ -178,7 +178,7 @@ def parse_expr(s: str):
     from pyparsing import ParseException
 
     try:
-        result = get_parser().parseString(s, parseAll=True)
+        result = get_parser().parse_string(s, parse_all=True)
     except ParseException as exc:
         format_and_raise_parse_error(exc)
         raise AssertionError("unreachable")  # noqa: B904

dvc/repo/experiments/__init__.py

@@ -188,7 +188,7 @@ class Experiments:
         rev_names = self.get_exact_name(revs)
         for rev in revs:
             name = rev_names[rev]
-            names.append(name if name else rev[:7])
+            names.append(name or rev[:7])
         ui.write("\nRan experiment(s): {}".format(", ".join(names)))
         if tmp_dir:
             ui.write(

dvc/repo/experiments/collect.py

@@ -58,7 +58,7 @@ def collect_rev(
         if not (rev == "workspace" or param_deps or data.contains_error):
             cache.put(data, force=True)
         return ExpState(rev=rev, data=data)
-    except Exception as exc:  # noqa: BLE001
+    except Exception as exc:
         logger.debug("", exc_info=True)
         error = SerializableError(str(exc), type(exc).__name__)
         return ExpState(rev=rev, error=error)
@@ -328,7 +328,7 @@ def collect(
                 name=baseline_names.get(baseline_rev),
                 data=baseline_data.data,
                 error=baseline_data.error,
-                experiments=experiments if experiments else None,
+                experiments=experiments or None,
             )
         )
     return result

dvc/repo/experiments/executor/base.py

@@ -757,7 +757,7 @@ class BaseExecutor(ABC):
             logger.debug("Commit to current experiment branch '%s'", branch)
         else:
             baseline_rev = scm.get_ref(EXEC_BASELINE)
-            name = exp_name if exp_name else f"exp-{exp_hash[:5]}"
+            name = exp_name or f"exp-{exp_hash[:5]}"
             ref_info = ExpRefInfo(baseline_rev, name)
             branch = str(ref_info)
             old_ref = None

dvc/repo/experiments/queue/base.py

@@ -90,7 +90,7 @@ class BaseStashQueue(ABC):
         """Construct a queue.
 
         Arguments:
-            scm: Git SCM instance for this queue.
+            repo: DVC repo instance for this queue.
             ref: Git stash ref for this queue.
             failed_ref: Failed run Git stash ref for this queue.
         """

dvc/repo/experiments/queue/tasks.py

@@ -84,8 +84,8 @@ def cleanup_exp(executor: TempDirExecutor, infofile: str) -> None:
     """Cleanup after an experiment.
 
     Arguments:
-        tmp_dir: Temp directory to be removed.
-        entry_dict: Serialized QueueEntry for this experiment.
+        executor: Executor to clean up.
+        infofile: Path to the executor info file.
     """
     executor.cleanup(infofile)
 

dvc/repo/experiments/refs.py

@@ -34,7 +34,7 @@ class ExpRefInfo:
 
     def __init__(self, baseline_sha: str, name: Optional[str] = None):
         self.baseline_sha = baseline_sha
-        self.name: str = name if name else ""
+        self.name: str = name or ""
 
     def __str__(self):
         return "/".join(self.parts)

dvc/repo/experiments/stash.py

@@ -67,7 +67,7 @@ class ExpStash(Stash):
         branch: Optional[str] = None,
     ) -> str:
         msg = cls.MESSAGE_FORMAT.format(
-            rev=rev, baseline_rev=baseline_rev, name=name if name else ""
+            rev=rev, baseline_rev=baseline_rev, name=name or ""
         )
         branch_msg = f":{branch}" if branch else ""
         return f"{msg}{branch_msg}"
@@ -124,9 +124,7 @@ class ApplyStash(Stash):
 
     @classmethod
     def format_message(cls, head_rev: str, rev: str, name: Optional[str] = None) -> str:
-        return cls.MESSAGE_FORMAT.format(
-            head_rev=head_rev, rev=rev, name=name if name else ""
-        )
+        return cls.MESSAGE_FORMAT.format(head_rev=head_rev, rev=rev, name=name or "")
 
     @contextmanager
     def preserve_workspace(

dvc/repo/experiments/utils.py

@@ -200,7 +200,7 @@ def exp_commits(
 ) -> Iterable[str]:
     """Iterate over all experiment commits."""
     shas: set[str] = set()
-    refs = ref_infos if ref_infos else exp_refs(scm)
+    refs = ref_infos or exp_refs(scm)
     for ref_info in refs:
         shas.update(scm.branch_revs(str(ref_info), ref_info.baseline_sha))
     yield from shas

dvc/repo/gc.py

@@ -28,8 +28,10 @@ def _validate_args(**kwargs):
             "`--all-experiments`, `--all-commits`, `--date` or `--rev` "
             "needs to be set."
         )
-    if kwargs.get("num") and not kwargs.get("rev"):
-        raise InvalidArgumentError("`--num` can only be used alongside `--rev`")
+    if kwargs.get("num") and not (kwargs.get("rev") or kwargs.get("all_branches")):
+        raise InvalidArgumentError(
+            "`--num` can only be used alongside `--rev` or `--all-branches`"
+        )
 
 
 def _used_obj_ids_not_in_remote(
@@ -128,9 +130,11 @@ def gc(  # noqa: C901, PLR0912, PLR0913
         used_obj_ids = set()
         used_obj_ids.update(*odb_to_obj_ids.values())
 
+    seen_odbs = set()
     for scheme, odb in self.cache.by_scheme():
-        if not odb:
+        if not odb or odb in seen_odbs:
             continue
+        seen_odbs.add(odb)
         num_removed = ogc(odb, used_obj_ids, jobs=jobs, dry=dry)
         if num_removed:
             logger.info("Removed %d objects from %s cache.", num_removed, scheme)

dvc/repo/metrics/show.py

@@ -123,7 +123,7 @@ def try_expand_paths(fs: "FileSystem", paths: Iterable[str]) -> Iterator[str]:
             if fs.isdir(path):
                 yield from fs.find(path)
                 continue
-        except Exception as e:  # noqa: BLE001
+        except Exception as e:
             logger.debug(
                 "failed to expand %r: %s",
                 path,

dvc/repo/scm_context.py

@@ -10,6 +10,8 @@ from dvc.utils import relpath
 from dvc.utils.collections import ensure_list
 
 if TYPE_CHECKING:
+    from typing_extensions import Self
+
     from dvc.repo import Repo
     from dvc.scm import Base
 
@@ -96,7 +98,7 @@ class SCMContext:
     @contextmanager
     def __call__(
         self, autostage: Optional[bool] = None, quiet: Optional[bool] = None
-    ) -> Iterator["SCMContext"]:
+    ) -> Iterator["Self"]:
         try:
             yield self
         except Exception:
@@ -131,9 +133,10 @@ class SCMContext:
 
         self.files_to_track = set()
 
-    def __enter__(self) -> "SCMContext":
+    def __enter__(self) -> "Self":
         self._cm = self()
-        return self._cm.__enter__()
+        self._cm.__enter__()
+        return self
 
     def __exit__(self, *exc_args) -> None:
         assert self._cm

dvc/scm.py

@@ -234,7 +234,7 @@ def iter_revs(
         results.extend(scm.list_all_commits())
     else:
         if all_branches:
-            results.extend(scm.list_branches())
+            results.extend(_get_n_commits(scm, scm.list_branches(), num))
 
         if all_tags:
             results.extend(scm.list_tags())
@@ -258,7 +258,10 @@ def iter_revs(
         results.extend(exp_commits(scm))
 
     rev_resolver = partial(resolve_rev, scm)
-    return group_by(rev_resolver, results)
+    grouped = group_by(rev_resolver, results)
+    for rev, names in grouped.items():
+        grouped[rev] = list(dict.fromkeys(names))
+    return grouped
 
 
 def lfs_prefetch(fs: "FileSystem", paths: list[str]):

dvc/testing/benchmarks/fixtures.py

@@ -9,7 +9,7 @@ from typing import Optional
 import pytest
 from dulwich.porcelain import clone
 from funcy import first
-from packaging import version
+from packaging import specifiers, version
 
 from dvc.types import StrPath
 
@@ -109,11 +109,20 @@ def make_dvc_bin(
             else:
                 pkg = "dvc"
             packages = [f"{pkg} @ git+file://{dvc_repo}@{dvc_rev}"]
-            try:
-                if version.Version(dvc_rev) < version.Version("3.50.3"):
-                    packages.append("pygit2==1.14.1")
-            except version.InvalidVersion:
-                pass
+
+            version_constraints = [
+                ("<3.50.3", ["pygit2==1.14.1"]),
+                ("<3.44.0", ["dulwich<1.0.0"]),
+                ("<3.67.0", ["pathspec<1"]),
+            ]
+            for spec, pkgs in version_constraints:
+                try:
+                    _dvc_version = version.Version(dvc_rev)
+                except version.InvalidVersion:
+                    continue
+                if _dvc_version in specifiers.SpecifierSet(spec):
+                    packages.extend(pkgs)
+
             venv.install(*packages)
 
             dvc_venvs[dvc_rev] = venv

dvc/testing/path_info.py

@@ -2,7 +2,7 @@ import os
 import pathlib
 import posixpath
 import sys
-from typing import Callable, ClassVar
+from typing import Callable, ClassVar, Union
 from urllib.parse import urlparse
 
 from dvc.utils import relpath
@@ -79,15 +79,17 @@ class PathInfo(pathlib.PurePath, _BasePath):
             and self._cparts[:n] == other._cparts  # type: ignore[attr-defined]
         )
 
-    def relative_to(self, other, *args, **kwargs):
+    def relative_to(
+        self, *other: Union[str, "os.PathLike[str]"], **kwargs
+    ) -> "PathInfo":
         # pathlib relative_to raises exception when one path is not a direct
         # descendant of the other when os.path.relpath would return abspath.
         # For DVC PathInfo we only need the relpath behavior.
         # See: https://bugs.python.org/issue40358
         try:
-            path = super().relative_to(other, *args, **kwargs)
+            path = super().relative_to(*other, **kwargs)
         except ValueError:
-            path = relpath(self, other)
+            path = relpath(self, other[0])
         return self.__class__(path)
 
 

dvc/ui/__init__.py

@@ -289,7 +289,7 @@ class Console:
 
         return console.Console(stderr=True)
 
-    def table(
+    def table(  # noqa: PLR0913
         self,
         data: "TableData",
         headers: Optional["Headers"] = None,
@@ -301,6 +301,7 @@ class Console:
         row_styles: Optional[Sequence["Styles"]] = None,
         borders: Union[bool, str] = False,
         colalign: Optional[tuple[str, ...]] = None,
+        preserve_whitespace: bool = False,
     ) -> None:
         from dvc.ui import table as t
 
@@ -329,6 +330,7 @@ class Console:
             pager=pager,
             force=force,
             colalign=colalign,
+            preserve_whitespace=preserve_whitespace,
         )
 
     def status(self, status: str, **kwargs: Any) -> "Status":

dvc/ui/table.py

@@ -30,19 +30,37 @@ def plain_table(
     pager: bool = False,
     force: bool = True,
     colalign: Optional[tuple[str, ...]] = None,
+    preserve_whitespace: bool = False,
 ) -> None:
+    import tabulate as tabulate_mod
     from funcy import nullcontext
+    from packaging.version import Version
     from tabulate import tabulate
 
-    text: str = tabulate(
-        data,
-        headers if headers is not None else (),
-        tablefmt="github" if markdown else "plain",
-        disable_numparse=True,
-        # None will be shown as "" by default, overriding
-        missingval="-",
-        colalign=colalign,
-    )
+    # NOTE: tabulate 0.10.0+ supports preserve_whitespace as a kwarg,
+    # while 0.9.0 only respects the global PRESERVE_WHITESPACE.
+    kwargs: dict = {}
+    _orig = tabulate_mod.PRESERVE_WHITESPACE
+    if preserve_whitespace:
+        if Version(tabulate_mod.__version__) >= Version("0.10"):
+            kwargs["preserve_whitespace"] = True
+        else:
+            tabulate_mod.PRESERVE_WHITESPACE = True
+
+    try:
+        text: str = tabulate(
+            data,
+            headers if headers is not None else (),
+            tablefmt="github" if markdown else "plain",
+            disable_numparse=True,
+            # None will be shown as "" by default, overriding
+            missingval="-",
+            colalign=colalign,
+            **kwargs,
+        )
+    finally:
+        tabulate_mod.PRESERVE_WHITESPACE = _orig
+
     if markdown:
         # NOTE: md table is incomplete without the trailing newline
         text += "\n"

pyproject.toml

@@ -41,13 +41,13 @@ dependencies = [
     "distro>=1.3",
     "dpath<3,>=2.1.0",
     "dulwich",
-    "dvc-data>=3.17.0,<3.18",
+    "dvc-data>=3.18.2,<3.19.0",
     "dvc-http>=2.29.0",
     "dvc-objects",
     "dvc-render>=1.0.1,<2",
     "dvc-studio-client>=0.21,<1",
     "dvc-task>=0.3.0,<1",
-    "flatten_dict<1,>=0.4.1",
+    "flatten-dict<1,>=0.4.1",
     "flufl.lock>=8.1.0,<10",
     "fsspec>=2024.2.0",
     "funcy>=1.14",
@@ -59,12 +59,12 @@ dependencies = [
     "networkx>=2.5",
     "omegaconf",
     "packaging>=19",
-    "pathspec>=0.10.3,<1",
+    "pathspec>=0.10.3,<2",
     "platformdirs<5,>=3.1.1",
     "psutil>=5.8",
     "pydot>=1.2.4",
     "pygtrie>=2.3.2",
-    "pyparsing>=2.4.7",
+    "pyparsing>=3.0.0",
     "requests>=2.22",
     "rich>=12",
     "ruamel.yaml>=0.17.11",

tests/func/experiments/test_experiments.py

@@ -53,7 +53,7 @@ def test_new_simple(tmp_dir, scm, dvc, exp_stage, mocker, name, workspace):
     if workspace:
         assert (tmp_dir / "metrics.yaml").read_text().strip() == "foo: 2"
 
-    exp_name = name if name else ref_info.name
+    exp_name = name or ref_info.name
     assert dvc.experiments.get_exact_name([exp])[exp] == exp_name
     assert resolve_rev(scm, exp_name) == exp
 

tests/func/experiments/test_save.py

@@ -28,7 +28,7 @@ def test_exp_save(tmp_dir, dvc, scm, name):
     assert ref_info
     assert ref_info.baseline_sha == baseline
 
-    exp_name = name if name else ref_info.name
+    exp_name = name or ref_info.name
     assert dvc.experiments.get_exact_name([exp])[exp] == exp_name
     assert resolve_rev(scm, exp_name) == exp
 

tests/func/test_data_status.py

@@ -1244,3 +1244,27 @@ def test_shallow_should_iterate_upto_tracked_directory(tmp_dir, dvc, scm, local_
         "committed": {"added": M.unordered(*files)},
         "git": M.dict(),
     }
+
+
+def test_duplicate_hashes_not_in_remote(tmp_dir, dvc, scm, local_remote):
+    """Test that files with identical content (same hash) are all correctly reported.
+
+    Regression test for https://github.com/iterative/dvc/issues/10959
+    """
+    tmp_dir.dvc_gen(
+        {"foo": "content", "bar": "content", "foobar": "foobar"}, commit="add files"
+    )
+    assert dvc.data_status(not_in_remote=True, granular=True, remote_refresh=True) == {
+        **EMPTY_STATUS,
+        "unchanged": M.unordered("foo", "bar", "foobar"),
+        "not_in_remote": M.unordered("foo", "bar", "foobar"),
+        "git": M.dict(),
+    }
+
+    dvc.push()
+    assert dvc.data_status(not_in_remote=True, granular=True, remote_refresh=True) == {
+        **EMPTY_STATUS,
+        "unchanged": M.unordered("foo", "bar", "foobar"),
+        "not_in_remote": [],
+        "git": M.dict(),
+    }

tests/func/test_gc.py

@@ -92,6 +92,27 @@ def test_gc_multiple_dvc_repos(tmp_dir, scm, dvc, erepo_dir):
     assert len(list(odb.all())) == 2
 
 
+def test_gc_multiple_dvc_repos_all_branches_num(tmp_dir, scm, dvc, erepo_dir):
+    tmp_dir.dvc_gen("main", "main", commit="main")
+
+    erepo_dir.dvc.cache.local.path = dvc.cache.local.path
+    with erepo_dir.chdir():
+        erepo_dir.scm_gen("base", "base", commit="base")
+        with erepo_dir.branch("feature", new=True):
+            erepo_dir.dvc_gen("shared", "feature-v1", commit="feature-v1")
+            erepo_dir.dvc.remove("shared.dvc")
+            erepo_dir.dvc_gen("shared", "feature-v2", commit="feature-v2")
+
+    odb = dvc.cache.local
+    assert len(list(odb.all())) == 3
+
+    dvc.gc(all_branches=True, num=2, repos=[erepo_dir])
+    assert len(list(odb.all())) == 3
+
+    dvc.gc(all_branches=True, repos=[erepo_dir])
+    assert len(list(odb.all())) == 2
+
+
 def test_all_commits(tmp_dir, scm, dvc):
     tmp_dir.dvc_gen("testfile", "uncommitted")
     tmp_dir.dvc_gen("testfile", "committed", commit="committed")
@@ -437,8 +458,16 @@ def test_gc_logging(caplog, dvc, good_and_bad_cache):
     with caplog.at_level(logging.INFO, logger="dvc"):
         dvc.gc(workspace=True)
 
-    assert "Removed 3 objects from repo cache." in caplog.text
-    assert "No unused 'local' cache to remove." in caplog.text
+    # Check that 3 objects were removed from either repo or local cache
+    # (they point to the same ODB, so only one will be logged due to deduplication)
+    has_repo_log = "Removed 3 objects from repo cache." in caplog.text
+    has_local_log = "Removed 3 objects from local cache." in caplog.text
+
+    # Exactly one of repo or local should have the log, not both (deduplication)
+    assert has_repo_log ^ has_local_log, (
+        "Should have exactly one log for repo/local cache"
+    )
+
     assert "No unused 'legacy' cache to remove." in caplog.text
 
 
@@ -450,3 +479,21 @@ def test_gc_skip_failed(tmp_dir, dvc):
         dvc.gc(force=True, workspace=True)
 
     dvc.gc(force=True, workspace=True, skip_failed=True)
+
+
+def test_gc_dry_logs_paths(caplog, tmp_dir, dvc):
+    """Test that dry run logs paths of objects to be removed."""
+    odb = dvc.cache.local
+
+    # Add some unused objects directly to cache
+    unused_hashes = ["test123", "test456", "test789"]
+    for hash_val in unused_hashes:
+        odb.add_bytes(hash_val, hash_val.encode("utf8"))
+
+    with caplog.at_level(logging.INFO, logger="dvc_data.hashfile.gc"):
+        dvc.gc(workspace=True, dry=True)
+
+    # Verify that paths are logged by dvc-data layer
+    for hash_val in unused_hashes:
+        expected_path = odb.oid_to_path(hash_val)
+        assert f"Removing {expected_path}" in caplog.text

tests/unit/command/test_diff.py

@@ -196,7 +196,7 @@ def test_diff_show_markdown_and_hash(mocker, show_hash, dvc):
     cmd = args.func(args)
 
     diff = {}
-    show_hash = show_hash if show_hash else False
+    show_hash = show_hash or False
     mock_show_markdown = mocker.patch("dvc.commands.diff._show_markdown")
     mocker.patch("dvc.repo.Repo.diff", return_value=diff.copy())
 

tests/unit/command/test_gc.py

@@ -65,6 +65,30 @@ def test_(dvc, scm, mocker):
     with pytest.raises(InvalidArgumentError):
         cmd.run()
 
+    cli_args = parse_args(["gc", "--all-branches", "--num", "2", "--force"])
+    cmd = cli_args.func(cli_args)
+
+    assert cmd.run() == 0
+
+    m.assert_called_with(
+        workspace=False,
+        all_tags=False,
+        all_branches=True,
+        all_commits=False,
+        all_experiments=False,
+        commit_date=None,
+        cloud=False,
+        remote=None,
+        force=True,
+        jobs=None,
+        repos=None,
+        rev=None,
+        num=2,
+        not_in_remote=False,
+        dry=False,
+        skip_failed=False,
+    )
+
     cli_args = parse_args(["gc", "--cloud", "--not-in-remote"])
     cmd = cli_args.func(cli_args)
     with pytest.raises(InvalidArgumentError):

tests/unit/command/test_help.py

@@ -47,7 +47,7 @@ def test_help(caplog, capsys, command_tuples):
     out, err = capsys.readouterr()
 
     # validate metavars are all in lowercase
-    usage = "\n".join(takewhile(lambda o: bool(o), out.splitlines()))
+    usage = "\n".join(takewhile(bool, out.splitlines()))
 
     message = (
         "metavars not lowercased, you are likely missing formatter_class=XXX "

tests/unit/scm/test_scm.py

@@ -47,6 +47,13 @@ def test_iter_revs(tmp_dir, scm, mocker):
     assert gen == {rev_old: ["tag"]}
     gen = iter_revs(scm, all_branches=True)
     assert gen == {rev_old: [old], rev_new: ["new"], rev_other: ["other"]}
+    gen = iter_revs(scm, all_branches=True, num=2)
+    assert gen == {
+        rev_old: [old, rev_old],
+        rev_root: [rev_root],
+        rev_new: ["new"],
+        rev_other: ["other"],
+    }
     gen = iter_revs(scm, all_tags=True)
     assert gen == {rev_old: ["tag"]}
     gen = iter_revs(scm, all_commits=True)

tests/unit/test_logger.py

@@ -113,7 +113,7 @@ class TestColorFormatter:
         with caplog.at_level(logging.DEBUG, logger="dvc"):
             try:
                 raise Exception("description")
-            except Exception:  # noqa: BLE001
+            except Exception:
                 stack_trace = traceback.format_exc()
                 logger.debug("", exc_info=True)