chore: replace some old code (python 2) (#1020)

* add test for tree write update

* rename test

* move test to new file to avoid skip

* remove python 2 related comment

* update description

* delete new test

* remove comment

* remove comment

* remove lzma helper

* remove "isstr" helper

* fix annotation

Author: lobis

.github/workflows/build-test.yml

@@ -44,7 +44,7 @@ jobs:
           mamba install root
           conda list
 
-      - name: Install sshd for sshfs tests
+      - name: Install sshd for fsspec ssh tests
         if: runner.os != 'macOS'  &&  runner.os != 'Windows'
         run: |
           sudo apt-get install -y openssh-server

src/uproot/_util.py

@@ -56,13 +56,6 @@ def isnum(x):
     )
 
 
-def isstr(x):
-    """
-    Returns True if and only if ``x`` is a string (including Python 2 unicode).
-    """
-    return isinstance(x, str)
-
-
 def ensure_str(x):
     """
     Ensures that ``x`` is a string (decoding with 'surrogateescape' if necessary).
@@ -166,7 +159,7 @@ def regularize_filter(filter):
         return no_filter
     elif callable(filter):
         return filter
-    elif isstr(filter):
+    elif isinstance(filter, str):
         m = _regularize_filter_regex.match(filter)
         if m is not None:
             regex, flags = m.groups()
@@ -207,7 +200,7 @@ def regularize_rename(rename):
     elif callable(rename):
         return rename
 
-    elif isstr(rename):
+    elif isinstance(rename, str):
         m = _regularize_filter_regex_rename.match(rename)
         if m is not None:
             regex, trans, flags = m.groups()
@@ -222,7 +215,7 @@ def regularize_rename(rename):
     elif isinstance(rename, Iterable) and not isinstance(rename, bytes):
         rules = []
         for x in rename:
-            if isstr(x):
+            if isinstance(x, str):
                 m = _regularize_filter_regex_rename.match(x)
                 if m is not None:
                     regex, trans, flags = m.groups()
@@ -387,7 +380,7 @@ def file_path_to_source_class(file_path, options):
             return out, file_path
 
     if (
-        not isstr(file_path)
+        not isinstance(file_path, str)
         and hasattr(file_path, "read")
         and hasattr(file_path, "seek")
     ):
@@ -593,7 +586,7 @@ def memory_size(data, error_message=None):
     Regularizes strings like '## kB' and plain integer number of bytes to
     an integer number of bytes.
     """
-    if isstr(data):
+    if isinstance(data, str):
         m = re.match(
             r"^\s*([+-]?(\d+(\.\d*)?|\.\d+)(e[+-]?\d+)?)\s*([kmgtpezy]?b)\s*$",
             data,
@@ -1001,11 +994,11 @@ def _regularize_files_inner(files, parse_colon, counter, HasBranches, steps_allo
 
     maybe_steps = None
 
-    if isstr(files2) and not isstr(files):
+    if isinstance(files2, str) and not isinstance(files, str):
         parse_colon = False
         files = files2
 
-    if isstr(files):
+    if isinstance(files, str):
         if parse_colon:
             file_path, object_path = file_object_path_split(files)
         else:
@@ -1098,7 +1091,7 @@ def regularize_files(files, steps_allowed):
     for file_path, object_path, maybe_steps in _regularize_files_inner(
         files, True, counter, HasBranches, steps_allowed
     ):
-        if isstr(file_path):
+        if isinstance(file_path, str):
             key = (counter[0], file_path, object_path)
             if key not in seen:
                 out.append((file_path, object_path))
@@ -1135,7 +1128,7 @@ def regularize_object_path(
             object_cache=None,
             array_cache=None,
             custom_classes=custom_classes,
-            **options,  # NOTE: a comma after **options breaks Python 2
+            **options,
         ).root_directory
         if object_path is None:
             trees = file.keys(filter_classname="TTree", cycle=False)

src/uproot/behaviors/TBranch.py

@@ -63,7 +63,7 @@ def iterate(
     report=False,
     custom_classes=None,
     allow_missing=False,
-    **options,  # NOTE: a comma after **options breaks Python 2
+    **options,
 ):
     """
     Args:
@@ -243,7 +243,7 @@ def concatenate(
     how=None,
     custom_classes=None,
     allow_missing=False,
-    **options,  # NOTE: a comma after **options breaks Python 2
+    **options,
 ):
     """
     Args:
@@ -1617,7 +1617,7 @@ def __getitem__(self, where):
 
         if uproot._util.isint(where):
             return self.branches[where]
-        elif uproot._util.isstr(where):
+        elif isinstance(where, str):
             where = uproot._util.ensure_str(where)
         else:
             raise TypeError(f"where must be an integer or a string, not {where!r}")
@@ -2666,11 +2666,11 @@ def _regularize_executors(decompression_executor, interpretation_executor, file)
 def _regularize_array_cache(array_cache, file):
     if isinstance(array_cache, MutableMapping):
         return array_cache
-    elif uproot._util.isstr(array_cache) and array_cache == "inherit":
+    elif isinstance(array_cache, str) and array_cache == "inherit":
         return file._array_cache
     elif array_cache is None:
         return None
-    elif uproot._util.isint(array_cache) or uproot._util.isstr(array_cache):
+    elif uproot._util.isint(array_cache) or isinstance(array_cache, str):
         return uproot.cache.LRUArrayCache(array_cache)
     else:
         raise TypeError("array_cache must be None, a MutableMapping, or a memory size")
@@ -2895,7 +2895,7 @@ def _regularize_expressions(
                     branchname,
                 )
 
-    elif uproot._util.isstr(expressions):
+    elif isinstance(expressions, str):
         _regularize_expression(
             hasbranches,
             expressions,
@@ -2917,7 +2917,7 @@ def _regularize_expressions(
         else:
             items = []
             for expression in expressions:
-                if uproot._util.isstr(expression):
+                if isinstance(expression, str):
                     items.append((expression, None))
                 elif isinstance(expression, tuple) and len(expression) == 2:
                     items.append(expression)
@@ -2972,7 +2972,7 @@ def _regularize_expressions(
 
     if cut is None:
         pass
-    elif uproot._util.isstr(cut):
+    elif isinstance(cut, str):
         _regularize_expression(
             hasbranches,
             cut,

src/uproot/compression.py

@@ -5,6 +5,7 @@
 as functions for compressing and decompressing a :doc:`uproot.source.chunk.Chunk`.
 """
 
+from __future__ import annotations
 
 import struct
 import threading
@@ -88,7 +89,7 @@ class _DecompressZLIB:
     _2byte = b"ZL"
     _method = b"\x08"
 
-    def decompress(self, data, uncompressed_bytes=None):
+    def decompress(self, data: bytes, uncompressed_bytes=None) -> bytes:
         import zlib
 
         return zlib.decompress(data)
@@ -125,7 +126,7 @@ def level(self, value):
             raise ValueError("Compression level must be between 0 and 9 (inclusive)")
         self._level = int(value)
 
-    def compress(self, data):
+    def compress(self, data: bytes) -> bytes:
         import zlib
 
         return zlib.compress(data, self._level)
@@ -136,8 +137,9 @@ class _DecompressLZMA:
     _2byte = b"XZ"
     _method = b"\x00"
 
-    def decompress(self, data, uncompressed_bytes=None):
-        lzma = uproot.extras.lzma()
+    def decompress(self, data: bytes, uncompressed_bytes=None) -> bytes:
+        import lzma
+
         return lzma.decompress(data)
 
 
@@ -150,8 +152,6 @@ class LZMA(Compression, _DecompressLZMA):
     Represents the LZMA compression algorithm.
 
     Uproot uses ``lzma`` from the Python 3 standard library.
-
-    In Python 2, ``backports.lzma`` must be installed.
     """
 
     def __init__(self, level):
@@ -174,8 +174,9 @@ def level(self, value):
             raise ValueError("Compression level must be between 0 and 9 (inclusive)")
         self._level = int(value)
 
-    def compress(self, data):
-        lzma = uproot.extras.lzma()
+    def compress(self, data: bytes) -> bytes:
+        import lzma
+
         return lzma.compress(data, preset=self._level)
 
 
@@ -184,7 +185,7 @@ class _DecompressLZ4:
     _2byte = b"L4"
     _method = b"\x01"
 
-    def decompress(self, data, uncompressed_bytes=None):
+    def decompress(self, data: bytes, uncompressed_bytes=None) -> bytes:
         lz4_block = uproot.extras.lz4_block()
         if uncompressed_bytes is None:
             raise ValueError(
@@ -224,7 +225,7 @@ def level(self, value):
             raise ValueError("Compression level must be between 0 and 12 (inclusive)")
         self._level = int(value)
 
-    def compress(self, data):
+    def compress(self, data: bytes) -> bytes:
         lz4_block = uproot.extras.lz4_block()
         return lz4_block.compress(data, compression=self._level, store_size=False)
 
@@ -245,7 +246,7 @@ def decompressor(self):
             self._decompressor.obj = zstandard.ZstdDecompressor()
         return self._decompressor.obj
 
-    def decompress(self, data, uncompressed_bytes=None):
+    def decompress(self, data: bytes, uncompressed_bytes=None) -> bytes:
         return self.decompressor.decompress(data)
 
 
@@ -288,7 +289,7 @@ def compressor(self):
             self._compressor = zstandard.ZstdCompressor(level=self._level)
         return self._compressor
 
-    def compress(self, data):
+    def compress(self, data: bytes) -> bytes:
         return self.compressor.compress(data)
 
 
@@ -475,7 +476,7 @@ def hook_after_block(**kwargs):
 _4byte = struct.Struct("<I")  # compressed sizes are 3-byte little endian!
 
 
-def compress(data, compression):
+def compress(data: bytes, compression: Compression) -> bytes:
     """
     Args:
         data (bytes, memoryview, or NumPy array): Data to compress.

src/uproot/containers.py

@@ -117,7 +117,7 @@ def _read_nested(
 
 
 def _tostring(value):
-    if uproot._util.isstr(value):
+    if isinstance(value, str):
         return repr(value)
     else:
         return str(value)

src/uproot/extras.py

@@ -8,7 +8,6 @@
 error messages containing instructions on how to install the library.
 """
 
-
 import atexit
 import importlib.metadata as importlib_metadata
 import os
@@ -160,16 +159,6 @@ def xrootd_version():
             return None
 
 
-def lzma():
-    """
-    Imports and returns ``lzma`` (which is part of the Python 3 standard
-    library, but not Python 2).
-    """
-    import lzma
-
-    return lzma
-
-
 def lz4_block():
     """
     Imports and returns ``lz4``.

src/uproot/interpretation/library.py

@@ -867,7 +867,7 @@ def group(self, arrays, expression_context, how):
         elif how is dict:
             return {_rename(name, c): arrays[name] for name, c in expression_context}
 
-        elif uproot._util.isstr(how) or how is None:
+        elif isinstance(how, str) or how is None:
             arrays, names = _pandas_only_series(pandas, arrays, expression_context)
             return _pandas_memory_efficient(pandas, arrays, names)
 

src/uproot/model.py

@@ -591,8 +591,8 @@ def base(self, *cls):
 
         The ``cls`` arguments may be Python classes or C++ classname strings to match.
         """
-        cpp_names = [classname_regularize(x) for x in cls if uproot._util.isstr(x)]
-        py_types = tuple(x for x in cls if not uproot._util.isstr(x))
+        cpp_names = [classname_regularize(x) for x in cls if isinstance(x, str)]
+        py_types = tuple(x for x in cls if not isinstance(x, str))
 
         out = []
         for x in getattr(self, "_bases", []):
@@ -610,8 +610,8 @@ def is_instance(self, *cls):
 
         The ``cls`` arguments may be Python classes or C++ classname strings to match.
         """
-        cpp_names = [classname_regularize(x) for x in cls if uproot._util.isstr(x)]
-        py_types = tuple(x for x in cls if not uproot._util.isstr(x))
+        cpp_names = [classname_regularize(x) for x in cls if isinstance(x, str)]
+        py_types = tuple(x for x in cls if not isinstance(x, str))
 
         if isinstance(self, py_types) or any(self.classname == n for n in cpp_names):
             return True

src/uproot/models/TTree.py

@@ -949,7 +949,7 @@ def num_entries(paths):
 
     if isinstance(paths, dict):
         paths = list(paths.items())
-    elif not uproot._util.isstr(paths):
+    elif not isinstance(paths, str):
         paths = [(uproot._util.file_object_path_split(path)) for path in paths]
     else:
         paths = [uproot._util.file_object_path_split(paths)]