[pre-commit.ci] pre-commit autoupdate (#10981)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Saugat Pachhai (सौगात) <[email protected]>

.github/workflows/build.yaml

@@ -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/tests.yaml

@@ -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@v6
   check:
     if: always()
     needs: [tests]

.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: 'v0.15.0'
+    rev: 'v0.15.8'
     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/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/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, parse_all=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/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(

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/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"

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")

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/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)