drop python 3.8 (#509)

.github/workflows/tests.yml

@@ -21,9 +21,9 @@ jobs:
       fail-fast: false
       matrix:
         os: [ubuntu-20.04, windows-latest, macos-latest]
-        pyv: ['3.8', '3.9', '3.10', '3.11', '3.12']
+        pyv: ['3.9', '3.10', '3.11', '3.12']
         include:
-        - {os: ubuntu-latest, pyv: 'pypy3.8'}
+        - {os: ubuntu-latest, pyv: 'pypy3.9'}
         - {os: macos-14, pyv: '3.11'}
 
     steps:

pyproject.toml

@@ -12,14 +12,13 @@ license = {text = "Apache-2.0"}
 authors = [{ name = "Iterative", email = "[email protected]" }]
 classifiers = [
     "Programming Language :: Python :: 3",
-    "Programming Language :: Python :: 3.8",
     "Programming Language :: Python :: 3.9",
     "Programming Language :: Python :: 3.10",
     "Programming Language :: Python :: 3.11",
     "Programming Language :: Python :: 3.12",
     "Development Status :: 4 - Beta",
 ]
-requires-python = ">=3.8"
+requires-python = ">=3.9"
 dynamic = ["version"]
 dependencies = [
     "funcy>=1.14",

src/dvc_data/callbacks.py

@@ -3,7 +3,7 @@
 import re
 import sys
 from threading import RLock
-from typing import Any, BinaryIO, ClassVar, Dict, Optional, Type, Union
+from typing import Any, BinaryIO, ClassVar, Optional, Union
 
 import fsspec
 from tqdm import tqdm
@@ -39,7 +39,7 @@ class Tqdm(tqdm):
         " [{elapsed}<{remaining}, {rate_fmt:>11}]"
     )
     BAR_FMT_NOTOTAL = "{desc}{bar:b}|{postfix[info]}{n_fmt} [{elapsed}, {rate_fmt:>11}]"
-    BYTES_DEFAULTS: ClassVar[Dict[str, Any]] = {
+    BYTES_DEFAULTS: ClassVar[dict[str, Any]] = {
         "unit": "B",
         "unit_scale": True,
         "unit_divisor": 1024,
@@ -146,7 +146,7 @@ def __init__(
         size: Optional[int] = None,
         value: int = 0,
         progress_bar: Optional["tqdm"] = None,
-        tqdm_cls: Optional[Type["tqdm"]] = None,
+        tqdm_cls: Optional[type["tqdm"]] = None,
         **tqdm_kwargs,
     ):
         tqdm_kwargs.pop("total", None)

src/dvc_data/cli.py

@@ -10,7 +10,6 @@
 from itertools import accumulate
 from pathlib import Path
 from posixpath import relpath
-from typing import List
 
 import click
 import typer
@@ -580,7 +579,7 @@ def checkout(
     path: Path = typer.Argument(..., resolve_path=True),
     relink: bool = False,
     force: bool = False,
-    type: List[LinkEnum] = typer.Option(["copy"]),  # noqa: A002
+    type: list[LinkEnum] = typer.Option(["copy"]),  # noqa: A002
 ):
     odb = get_odb(type=[t.value for t in type])
     oid = from_shortoid(odb, oid)

src/dvc_data/fs.py

@@ -5,7 +5,7 @@
 import posixpath
 import typing
 from collections import deque
-from typing import Any, BinaryIO, NamedTuple, Optional, Tuple
+from typing import Any, BinaryIO, NamedTuple, Optional
 
 from fsspec import AbstractFileSystem
 from fsspec.callbacks import DEFAULT_CALLBACK
@@ -43,7 +43,7 @@ def join(cls, *parts: str) -> str:
         return posixpath.join(*parts)
 
     @classmethod
-    def parts(cls, path: str) -> Tuple[str, ...]:
+    def parts(cls, path: str) -> tuple[str, ...]:
         ret = []
         while True:
             path, part = posixpath.split(path)
@@ -77,10 +77,10 @@ def relpath(self, path: str, start: Optional[str] = None) -> str:
             start = "."
         return posixpath.relpath(self.abspath(path), start=self.abspath(start))
 
-    def relparts(self, path: str, start: Optional[str] = None) -> Tuple[str, ...]:
+    def relparts(self, path: str, start: Optional[str] = None) -> tuple[str, ...]:
         return self.parts(self.relpath(path, start=start))
 
-    def _get_key(self, path: str) -> Tuple[str, ...]:
+    def _get_key(self, path: str) -> tuple[str, ...]:
         path = self.abspath(path)
         if path == self.root_marker:
             return ()
@@ -125,7 +125,7 @@ def _cache_remote_file(
         fs: "FileSystem",
         path: "AnyFSPath",
         hash_info: Optional["HashInfo"],
-    ) -> Tuple["FileSystem", "AnyFSPath"]:
+    ) -> tuple["FileSystem", "AnyFSPath"]:
         from dvc_objects.fs.local import LocalFileSystem
 
         odb: "HashFileDB" = cache_storage.odb

src/dvc_data/hashfile/__init__.py

@@ -1,6 +1,7 @@
 """DVC data."""
 import logging
-from typing import TYPE_CHECKING, Iterator, Union, cast
+from collections.abc import Iterator
+from typing import TYPE_CHECKING, Union, cast
 
 from .tree import Tree
 

src/dvc_data/hashfile/_ignore.py

@@ -1,4 +1,5 @@
-from typing import TYPE_CHECKING, Any, Iterator
+from collections.abc import Iterator
+from typing import TYPE_CHECKING, Any
 
 from typing_extensions import Protocol
 

src/dvc_data/hashfile/build.py

@@ -1,7 +1,7 @@
 import hashlib
 import logging
 import os
-from typing import TYPE_CHECKING, Any, Dict, Optional, Tuple, cast
+from typing import TYPE_CHECKING, Any, Optional, cast
 
 from fsspec.callbacks import DEFAULT_CALLBACK, Callback
 
@@ -45,7 +45,7 @@ def _upload_file(
     odb: "HashFileDB",
     upload_odb: "HashFileDB",
     callback: Optional[Callback] = None,
-) -> Tuple[Meta, HashFile]:
+) -> tuple[Meta, HashFile]:
     from dvc_objects.fs.utils import tmp_fname
 
     from .hash import HashStreamFile
@@ -131,7 +131,7 @@ def _build_tree(
 
         # NOTE: we know for sure that root starts with path, so we can use
         # faster string manipulation instead of a more robust relparts()
-        rel_key: Tuple[Optional[Any], ...] = ()
+        rel_key: tuple[Optional[Any], ...] = ()
         if root != path:
             rel_key = tuple(root[len(path) + 1 :].split(fs.sep))
 
@@ -160,7 +160,7 @@ def _build_tree(
     return tree_meta, tree
 
 
-_url_cache: Dict[str, str] = {}
+_url_cache: dict[str, str] = {}
 
 
 def _make_staging_url(fs: "FileSystem", odb: "HashFileDB", path: Optional[str]):
@@ -232,7 +232,7 @@ def build(
     upload: bool = False,
     dry_run: bool = False,
     **kwargs,
-) -> Tuple["HashFileDB", "Meta", "HashFile"]:
+) -> tuple["HashFileDB", "Meta", "HashFile"]:
     """Stage (prepare) objects from the given path for addition to an ODB.
 
     Returns at tuple of (object_store, object) where addition to the ODB can

src/dvc_data/hashfile/checkout.py

@@ -1,6 +1,6 @@
 import logging
 from itertools import chain
-from typing import TYPE_CHECKING, List, Optional
+from typing import TYPE_CHECKING, Optional
 
 from dvc_objects.fs.generic import test_links, transfer
 from fsspec.callbacks import DEFAULT_CALLBACK
@@ -24,7 +24,7 @@ def __init__(self, path: str) -> None:
 
 
 class CheckoutError(Exception):
-    def __init__(self, paths: List[str]) -> None:
+    def __init__(self, paths: list[str]) -> None:
         self.paths = paths
         super().__init__("Checkout failed")
 

src/dvc_data/hashfile/db/__init__.py

@@ -3,7 +3,7 @@
 import os
 from contextlib import suppress
 from copy import copy
-from typing import TYPE_CHECKING, Callable, ClassVar, List, Optional, Union
+from typing import TYPE_CHECKING, Callable, ClassVar, Optional, Union
 
 from dvc_objects.db import ObjectDB
 from dvc_objects.errors import ObjectFormatError
@@ -49,7 +49,7 @@ def get_index(odb) -> "ObjectDBIndexBase":
 
 class HashFileDB(ObjectDB):
     DEFAULT_VERIFY = False
-    DEFAULT_CACHE_TYPES: ClassVar[List[str]] = ["copy"]
+    DEFAULT_CACHE_TYPES: ClassVar[list[str]] = ["copy"]
     CACHE_MODE: Optional[int] = None
 
     def __init__(self, fs: "FileSystem", path: str, read_only: bool = False, **config):
@@ -72,9 +72,9 @@ def get(self, oid: str) -> HashFile:
 
     def add(
         self,
-        path: Union["AnyFSPath", List["AnyFSPath"]],
+        path: Union["AnyFSPath", list["AnyFSPath"]],
         fs: "FileSystem",
-        oid: Union[str, List[str]],
+        oid: Union[str, list[str]],
         hardlink: bool = False,
         callback: "Callback" = DEFAULT_CALLBACK,
         check_exists: bool = True,

src/dvc_data/hashfile/db/index.py

@@ -1,7 +1,8 @@
 import logging
 import os
 from abc import ABC, abstractmethod
-from typing import TYPE_CHECKING, Iterable, Iterator, Set
+from collections.abc import Iterable, Iterator
+from typing import TYPE_CHECKING
 
 from dvc_objects.errors import ObjectDBError
 
@@ -44,7 +45,7 @@ def update(self, dir_hashes: Iterable[str], file_hashes: Iterable[str]) -> None:
         pass
 
     @abstractmethod
-    def intersection(self, hashes: Set[str]) -> Iterator[str]:
+    def intersection(self, hashes: set[str]) -> Iterator[str]:
         pass
 
 
@@ -73,7 +74,7 @@ def clear(self) -> None:
     def update(self, dir_hashes: Iterable[str], file_hashes: Iterable[str]) -> None:
         pass
 
-    def intersection(self, hashes: Set[str]) -> Iterator[str]:
+    def intersection(self, hashes: set[str]) -> Iterator[str]:
         yield from []
 
 
@@ -131,6 +132,6 @@ def update(self, dir_hashes: Iterable[str], file_hashes: Iterable[str]) -> None:
         except Timeout as exc:
             raise ObjectDBError("Failed to update ODB index") from exc
 
-    def intersection(self, hashes: Set[str]) -> Iterator[str]:
+    def intersection(self, hashes: set[str]) -> Iterator[str]:
         """Iterate over values from `hashes` which exist in the index."""
         yield from hashes.intersection(self.index.keys())

src/dvc_data/hashfile/db/local.py

@@ -2,7 +2,7 @@
 import os
 import stat
 from functools import partial
-from typing import ClassVar, List
+from typing import ClassVar
 
 from dvc_objects.db import noop, wrap_iter
 from dvc_objects.errors import ObjectDBError, ObjectFormatError
@@ -19,7 +19,7 @@
 
 
 class LocalHashFileDB(HashFileDB):
-    DEFAULT_CACHE_TYPES: ClassVar[List[str]] = ["reflink", "copy"]
+    DEFAULT_CACHE_TYPES: ClassVar[list[str]] = ["reflink", "copy"]
     CACHE_MODE = 0o444
     UNPACKED_DIR_SUFFIX = ".unpacked"
 

src/dvc_data/hashfile/db/migrate.py

@@ -1,5 +1,5 @@
 from functools import partial, wraps
-from typing import TYPE_CHECKING, Any, Callable, Dict, List, NamedTuple, Tuple
+from typing import TYPE_CHECKING, Any, Callable, NamedTuple
 
 from dvc_objects.executors import ThreadPoolExecutor
 from fsspec.callbacks import DEFAULT_CALLBACK
@@ -14,8 +14,8 @@
 class PreparedMigration(NamedTuple):
     src: "HashFileDB"
     dest: "HashFileDB"
-    paths: List[str]
-    oids: List[str]
+    paths: list[str]
+    oids: list[str]
 
 
 def migrate(
@@ -65,7 +65,7 @@ def _hash_task(
     path: str,
     callback: "Callback" = DEFAULT_CALLBACK,
     **kwargs,
-) -> Tuple[str, str]:
+) -> tuple[str, str]:
     from dvc_data.hashfile.hash import hash_file
 
     func = _wrap_hash_file(callback, hash_file)
@@ -79,7 +79,7 @@ def _hash_task(
 def _wrap_hash_file(callback: "Callback", fn: Callable):
     @wraps(fn)
     def func(path: str, *args, **kwargs):
-        kw: Dict[str, Any] = dict(kwargs)
+        kw: dict[str, Any] = dict(kwargs)
         with callback.branched(path, path) as child:
             res = fn(path, *args, callback=child, **kw)
             callback.relative_update()

src/dvc_data/hashfile/db/reference.py

@@ -1,5 +1,5 @@
 import logging
-from typing import TYPE_CHECKING, Callable, Dict, List, Optional, Union
+from typing import TYPE_CHECKING, Callable, Optional, Union
 
 from dvc_data.hashfile.obj import HashFile
 
@@ -15,7 +15,7 @@
 class ReferenceHashFileDB(HashFileDB):
     def __init__(self, fs: "FileSystem", path: str, **config):
         super().__init__(fs, path, **config)
-        self._obj_cache: Dict["str", "HashFile"] = {}
+        self._obj_cache: dict["str", "HashFile"] = {}
 
     def __hash__(self):
         return hash((self.fs.protocol, self.path, *self._obj_cache.keys()))
@@ -31,9 +31,9 @@ def get(self, oid: str):
 
     def add(
         self,
-        path: Union["AnyFSPath", List["AnyFSPath"]],
+        path: Union["AnyFSPath", list["AnyFSPath"]],
         fs: "FileSystem",
-        oid: Union[str, List[str]],
+        oid: Union[str, list[str]],
         hardlink: bool = False,
         callback: Optional["Callback"] = None,
         check_exists: bool = True,

src/dvc_data/hashfile/diff.py

@@ -1,5 +1,5 @@
 import reprlib
-from typing import TYPE_CHECKING, Dict, List, Optional, Tuple
+from typing import TYPE_CHECKING, Optional
 
 from attrs import asdict, define, field
 
@@ -19,7 +19,7 @@
 @define(hash=True, order=True)
 class TreeEntry:
     in_cache: bool = field(default=False, eq=False)
-    key: Tuple[str, ...] = ()
+    key: tuple[str, ...] = ()
     meta: Optional["Meta"] = field(default=None, eq=False)
     oid: Optional["HashInfo"] = None
 
@@ -55,16 +55,16 @@ def __bool__(self):
 
 @define
 class DiffResult:
-    added: List[Change] = field(factory=list, repr=reprlib.repr)
-    modified: List[Change] = field(factory=list, repr=reprlib.repr)
-    deleted: List[Change] = field(factory=list, repr=reprlib.repr)
-    unchanged: List[Change] = field(factory=list, repr=reprlib.repr)
+    added: list[Change] = field(factory=list, repr=reprlib.repr)
+    modified: list[Change] = field(factory=list, repr=reprlib.repr)
+    deleted: list[Change] = field(factory=list, repr=reprlib.repr)
+    unchanged: list[Change] = field(factory=list, repr=reprlib.repr)
 
     def __bool__(self):
         return bool(self.added or self.modified or self.deleted)
 
     @property
-    def stats(self) -> Dict[str, int]:
+    def stats(self) -> dict[str, int]:
         return {k: len(v) for k, v in asdict(self).items() if k != "unchanged"}
 
 

src/dvc_data/hashfile/gc.py

@@ -1,4 +1,5 @@
-from typing import TYPE_CHECKING, Iterable, Optional
+from collections.abc import Iterable
+from typing import TYPE_CHECKING, Optional
 
 if TYPE_CHECKING:
     from .db import HashFileDB