Add unspecified-encoding checker #3826 (#4753)

* Add unspecified-encoding checker #3826
This adds an unspecified-encoding checker that adds a warning
whenever open() is called without an explicit encoding argument.
This closes #3826

* Update tests to conform to unspecified-encoding
With addition of the unspecified-encoding checker calls of open()
need an encoding argument.
Where necessary this argument has been added, or the message has been
disabled.
This also includes small linting changes to a small number
of tests. Their test-data has been updated to reflect new line numbers.

* Update scripts to conform to unspecified-encoding
With addition of the unspecified-encoding checker calls of open()
need an encoding argument.
Where necessary this argument has been added.

* Update pylint to conform to unspecified-encoding
With addition of the unspecified-encoding checker calls of open()
need an encoding argument.
Where necessary this argument has been added.

Co-authored-by: Marc Mueller <30130371+cdce8p@users.noreply.github.com>
Co-authored-by: Marc Mueller <30130371+cdce8p@users.noreply.github.com>

38 files changed, 221 insertions, 84 deletions

diff --git a/CONTRIBUTORS.txt b/CONTRIBUTORS.txt
index f1d54dbd5..c0865d0c6 100644
--- a/CONTRIBUTORS.txt
+++ b/CONTRIBUTORS.txt
@@ -521,3 +521,5 @@ contributors:
 * Yilei Yang: contributor
 
 * Marcin Kurczewski (rr-): contributor
+
+* Daniel van Noord (DanielNoord): contributor
diff --git a/ChangeLog b/ChangeLog
index 778e86393..52ce722f3 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -9,6 +9,10 @@ Release date: TBA
 ..
   Put new features here and also in 'doc/whatsnew/2.10.rst'
 
+* Added ``unspecified-encoding``: Emitted when open() is called without specifying an encoding
+
+  Closes #3826
+
 
 What's New in Pylint 2.9.6?
 ===========================
diff --git a/doc/whatsnew/2.10.rst b/doc/whatsnew/2.10.rst
index 3e17dc781..a2c4e1fbf 100644
--- a/doc/whatsnew/2.10.rst
+++ b/doc/whatsnew/2.10.rst
@@ -12,6 +12,9 @@ Summary -- Release highlights
 New checkers
 ============
 
+* Added ``unspecified-encoding``: Emitted when open() is called without specifying an encoding
+
+  Closes #3826
 
 
 Other Changes
diff --git a/pylint/checkers/similar.py b/pylint/checkers/similar.py
index 8ddbf6857..8d3155c9c 100644
--- a/pylint/checkers/similar.py
+++ b/pylint/checkers/similar.py
@@ -546,7 +546,7 @@ def Run(argv=None):
         min_lines, ignore_comments, ignore_docstrings, ignore_imports, ignore_signatures
     )
     for filename in args:
-        with open(filename) as stream:
+        with open(filename, encoding="utf-8") as stream:
             sim.append_stream(filename, stream)
     sim.run()
     sys.exit(0)
diff --git a/pylint/checkers/spelling.py b/pylint/checkers/spelling.py
index e2b6df189..1b0a7f653 100644
--- a/pylint/checkers/spelling.py
+++ b/pylint/checkers/spelling.py
@@ -318,7 +318,7 @@ class SpellingChecker(BaseTokenChecker):
                 dict_name, self.config.spelling_private_dict_file
             )
             self.private_dict_file = open(  # pylint: disable=consider-using-with
-                self.config.spelling_private_dict_file, "a"
+                self.config.spelling_private_dict_file, "a", encoding="utf-8"
             )
         else:
             self.spelling_dict = enchant.Dict(dict_name)
diff --git a/pylint/checkers/stdlib.py b/pylint/checkers/stdlib.py
index c0e789ad8..1fba67afd 100644
--- a/pylint/checkers/stdlib.py
+++ b/pylint/checkers/stdlib.py
@@ -30,6 +30,7 @@
 # Copyright (c) 2021 Marc Mueller <30130371+cdce8p@users.noreply.github.com>
 # Copyright (c) 2021 Matus Valo <matusvalo@users.noreply.github.com>
 # Copyright (c) 2021 victor <16359131+jiajunsu@users.noreply.github.com>
+# Copyright (c) 2021 Daniel van Noord <13665637+DanielNoord@users.noreply.github.com>
 
 # Licensed under the GPL: https://www.gnu.org/licenses/old-licenses/gpl-2.0.html
 # For details: https://github.com/PyCQA/pylint/blob/main/LICENSE
@@ -44,12 +45,13 @@ import astroid
 from pylint.checkers import BaseChecker, DeprecatedMixin, utils
 from pylint.interfaces import IAstroidChecker
 
-OPEN_FILES = {"open", "file"}
+OPEN_FILES_MODE = ("open", "file")
+OPEN_FILES_ENCODING = ("open",)
 UNITTEST_CASE = "unittest.case"
 THREADING_THREAD = "threading.Thread"
 COPY_COPY = "copy.copy"
 OS_ENVIRON = "os._Environ"
-ENV_GETTERS = {"os.getenv"}
+ENV_GETTERS = ("os.getenv",)
 SUBPROCESS_POPEN = "subprocess.Popen"
 SUBPROCESS_RUN = "subprocess.run"
 OPEN_MODULE = "_io"
@@ -425,6 +427,13 @@ class StdlibChecker(DeprecatedMixin, BaseChecker):
             "deprecated-decorator",
             "The decorator is marked as deprecated and will be removed in the future.",
         ),
+        "W1514": (
+            "Using open without explicitly specifying an encoding",
+            "unspecified-encoding",
+            "It is better to specify an encoding when opening documents. "
+            "Using the system default implicitly can create problems on other operating systems. "
+            "See https://www.python.org/dev/peps/pep-0597/",
+        ),
     }
 
     def __init__(self, linter=None):
@@ -485,6 +494,7 @@ class StdlibChecker(DeprecatedMixin, BaseChecker):
         "subprocess-popen-preexec-fn",
         "subprocess-run-check",
         "deprecated-class",
+        "unspecified-encoding",
     )
     def visit_call(self, node):
         """Visit a Call node."""
@@ -494,8 +504,18 @@ class StdlibChecker(DeprecatedMixin, BaseChecker):
                 if inferred is astroid.Uninferable:
                     continue
                 if inferred.root().name == OPEN_MODULE:
-                    if getattr(node.func, "name", None) in OPEN_FILES:
+                    if (
+                        isinstance(node.func, astroid.Name)
+                        and node.func.name in OPEN_FILES_MODE
+                    ):
                         self._check_open_mode(node)
+                    if (
+                        isinstance(node.func, astroid.Name)
+                        and node.func.name in OPEN_FILES_ENCODING
+                        or isinstance(node.func, astroid.Attribute)
+                        and node.func.attrname in OPEN_FILES_ENCODING
+                    ):
+                        self._check_open_encoded(node)
                 elif inferred.root().name == UNITTEST_CASE:
                     self._check_redundant_assert(node, inferred)
                 elif isinstance(inferred, astroid.ClassDef):
@@ -573,6 +593,34 @@ class StdlibChecker(DeprecatedMixin, BaseChecker):
             ):
                 self.add_message("bad-open-mode", node=node, args=mode_arg.value)
 
+    def _check_open_encoded(self, node: astroid.Call) -> None:
+        """Check that the encoded argument of an open call is valid."""
+        mode_arg = None
+        try:
+            mode_arg = utils.get_argument_from_call(node, position=1, keyword="mode")
+        except utils.NoSuchArgumentError:
+            pass
+
+        if mode_arg:
+            mode_arg = utils.safe_infer(mode_arg)
+        if not mode_arg or "b" not in mode_arg.value:
+            encoding_arg = None
+            try:
+                encoding_arg = utils.get_argument_from_call(
+                    node, position=None, keyword="encoding"
+                )
+            except utils.NoSuchArgumentError:
+                self.add_message("unspecified-encoding", node=node)
+
+            if encoding_arg:
+                encoding_arg = utils.safe_infer(encoding_arg)
+
+                if (
+                    isinstance(encoding_arg, astroid.Const)
+                    and encoding_arg.value is None
+                ):
+                    self.add_message("unspecified-encoding", node=node)
+
     def _check_env_function(self, node, infer):
         env_name_kwarg = "key"
         env_value_kwarg = "default"
diff --git a/pylint/config/find_default_config_files.py b/pylint/config/find_default_config_files.py
index d4faa259e..2a8937e52 100644
--- a/pylint/config/find_default_config_files.py
+++ b/pylint/config/find_default_config_files.py
@@ -9,7 +9,7 @@ from toml.decoder import TomlDecodeError
 
 
 def _toml_has_config(path):
-    with open(path) as toml_handle:
+    with open(path, encoding="utf-8") as toml_handle:
         try:
             content = toml.load(toml_handle)
         except TomlDecodeError as error:
diff --git a/pylint/config/option_manager_mixin.py b/pylint/config/option_manager_mixin.py
index 86785f36f..5aa2323f0 100644
--- a/pylint/config/option_manager_mixin.py
+++ b/pylint/config/option_manager_mixin.py
@@ -267,7 +267,7 @@ class OptionsManagerMixIn:
             parser = self.cfgfile_parser
 
             if config_file.endswith(".toml"):
-                with open(config_file) as fp:
+                with open(config_file, encoding="utf-8") as fp:
                     content = toml.load(fp)
 
                 try:
diff --git a/pylint/lint/pylinter.py b/pylint/lint/pylinter.py
index c2d536211..25ae29c02 100644
--- a/pylint/lint/pylinter.py
+++ b/pylint/lint/pylinter.py
@@ -571,7 +571,9 @@ class PyLinter(
                     (reporter_output,) = reporter_output
 
                     # pylint: disable=consider-using-with
-                    output_file = stack.enter_context(open(reporter_output, "w"))
+                    output_file = stack.enter_context(
+                        open(reporter_output, "w", encoding="utf-8")
+                    )
 
                     reporter.set_output(output_file)
                     output_files.append(output_file)
@@ -617,8 +619,9 @@ class PyLinter(
                 except KeyError:
                     meth = self._bw_options_methods[optname]
                     warnings.warn(
-                        "%s is deprecated, replace it by %s"
-                        % (optname, optname.split("-")[0]),
+                        "{} is deprecated, replace it by {}".format(
+                            optname, optname.split("-")[0]
+                        ),
                         DeprecationWarning,
                     )
                 value = utils._check_csv(value)
diff --git a/pylint/lint/run.py b/pylint/lint/run.py
index 08806e6c8..66c9b3622 100644
--- a/pylint/lint/run.py
+++ b/pylint/lint/run.py
@@ -373,7 +373,7 @@ group are mutually exclusive.",
 
         if self._output:
             try:
-                with open(self._output, "w") as output:
+                with open(self._output, "w", encoding="utf-8") as output:
                     linter.reporter.set_output(output)
                     linter.check(args)
                     score_value = linter.generate_reports()
diff --git a/pylint/pyreverse/utils.py b/pylint/pyreverse/utils.py
index fea9bda79..06f8d3b7e 100644
--- a/pylint/pyreverse/utils.py
+++ b/pylint/pyreverse/utils.py
@@ -35,7 +35,7 @@ def get_default_options():
     if home:
         rcfile = os.path.join(home, RCFILE)
         try:
-            with open(rcfile) as file_handle:
+            with open(rcfile, encoding="utf-8") as file_handle:
                 options = file_handle.read().split()
         except OSError:
             pass  # ignore if no config file found
diff --git a/pylint/pyreverse/writer.py b/pylint/pyreverse/writer.py
index db54c3f0b..f45783c49 100644
--- a/pylint/pyreverse/writer.py
+++ b/pylint/pyreverse/writer.py
@@ -187,7 +187,9 @@ class VCGWriter(DiagramWriter):
 
     def set_printer(self, file_name, basename):
         """initialize VCGWriter for a UML graph"""
-        self.graph_file = open(file_name, "w+")  # pylint: disable=consider-using-with
+        self.graph_file = open(  # pylint: disable=consider-using-with
+            file_name, "w+", encoding="utf-8"
+        )
         self.printer = VCGPrinter(self.graph_file)
         self.printer.open_graph(
             title=basename,
diff --git a/pylint/testutils/lint_module_test.py b/pylint/testutils/lint_module_test.py
index c88463821..8f24daf86 100644
--- a/pylint/testutils/lint_module_test.py
+++ b/pylint/testutils/lint_module_test.py
@@ -141,14 +141,14 @@ class LintModuleTest:
     # pylint: disable=consider-using-with
     def _open_expected_file(self):
         try:
-            return open(self._test_file.expected_output)
+            return open(self._test_file.expected_output, encoding="utf-8")
         except FileNotFoundError:
             return StringIO("")
 
     # pylint: disable=consider-using-with
     def _open_source_file(self):
         if self._test_file.base == "invalid_encoded_data":
-            return open(self._test_file.source)
+            return open(self._test_file.source, encoding="utf-8")
         if "latin1" in self._test_file.base:
             return open(self._test_file.source, encoding="latin1")
         return open(self._test_file.source, encoding="utf8")
diff --git a/script/bump_changelog.py b/script/bump_changelog.py
index 278b627da..41c8c3c8a 100644
--- a/script/bump_changelog.py
+++ b/script/bump_changelog.py
@@ -33,10 +33,10 @@ def main() -> None:
     logging.debug(f"Launching bump_changelog with args: {args}")
     if "dev" in args.version:
         return
-    with open(DEFAULT_CHANGELOG_PATH) as f:
+    with open(DEFAULT_CHANGELOG_PATH, encoding="utf-8") as f:
         content = f.read()
     content = transform_content(content, args.version)
-    with open(DEFAULT_CHANGELOG_PATH, "w") as f:
+    with open(DEFAULT_CHANGELOG_PATH, "w", encoding="utf-8") as f:
         f.write(content)
 
 
diff --git a/script/fix_documentation.py b/script/fix_documentation.py
index 242225381..b03f70ead 100644
--- a/script/fix_documentation.py
+++ b/script/fix_documentation.py
@@ -82,7 +82,7 @@ def main(argv: Union[List[str], None] = None) -> int:
 
     return_value: int = 0
     for file_name in args.filenames:
-        with open(file_name) as fp:
+        with open(file_name, encoding="utf-8") as fp:
             orignal_content = fp.read()
         content = orignal_content
         # Modify files
@@ -91,7 +91,7 @@ def main(argv: Union[List[str], None] = None) -> int:
             content = changelog_insert_empty_lines(content, args.subtitle_prefix)
         # If modified, write changes and eventually return 1
         if orignal_content != content:
-            with open(file_name, "w") as fp:
+            with open(file_name, "w", encoding="utf-8") as fp:
                 fp.write(content)
             return_value |= 1
     return return_value
diff --git a/tests/checkers/unittest_similar.py b/tests/checkers/unittest_similar.py
index 8edc87bca..ca64bdba3 100644
--- a/tests/checkers/unittest_similar.py
+++ b/tests/checkers/unittest_similar.py
@@ -449,7 +449,7 @@ def test_get_map_data():
     # Manually perform a 'map' type function
     for source_fname in source_streams:
         sim = similar.SimilarChecker(linter)
-        with open(source_fname) as stream:
+        with open(source_fname, encoding="utf-8") as stream:
             sim.append_stream(source_fname, stream)
         # The map bit, can you tell? ;)
         data.extend(sim.get_map_data())
diff --git a/tests/functional/b/bad_open_mode_py3.py b/tests/functional/b/bad_open_mode_py3.py
index 27aa2e6a2..b68238013 100644
--- a/tests/functional/b/bad_open_mode_py3.py
+++ b/tests/functional/b/bad_open_mode_py3.py
@@ -3,22 +3,22 @@
 

 NAME = "foo.bar"

 open(NAME, "wb")

-open(NAME, "w")

+open(NAME, "w", encoding="utf-8")

 open(NAME, "rb")

-open(NAME, "x")

+open(NAME, "x", encoding="utf-8")

 open(NAME, "br")

-open(NAME, "+r")

+open(NAME, "+r", encoding="utf-8")

 open(NAME, "xb")

-open(NAME, "rwx")  # [bad-open-mode]

-open(NAME, "rr")  # [bad-open-mode]

-open(NAME, "+")  # [bad-open-mode]

-open(NAME, "xw")  # [bad-open-mode]

+open(NAME, "rwx", encoding="utf-8")  # [bad-open-mode]

+open(NAME, "rr", encoding="utf-8")  # [bad-open-mode]

+open(NAME, "+", encoding="utf-8")  # [bad-open-mode]

+open(NAME, "xw", encoding="utf-8")  # [bad-open-mode]

 open(NAME, "ab+")

 open(NAME, "a+b")

 open(NAME, "+ab")

 open(NAME, "+rUb")

 open(NAME, "x+b")

-open(NAME, "Ua")  # [bad-open-mode]

-open(NAME, "Ur++")  # [bad-open-mode]

-open(NAME, "Ut")

+open(NAME, "Ua", encoding="utf-8")  # [bad-open-mode]

+open(NAME, "Ur++", encoding="utf-8")  # [bad-open-mode]

+open(NAME, "Ut", encoding="utf-8")

 open(NAME, "Ubr")

diff --git a/tests/functional/c/consider/consider_using_with.py b/tests/functional/c/consider/consider_using_with.py
index fc2804455..d05866aaa 100644
--- a/tests/functional/c/consider/consider_using_with.py
+++ b/tests/functional/c/consider/consider_using_with.py
@@ -151,7 +151,7 @@ def test_suppress_in_exit_stack():
     """Regression test for issue #4654 (false positive)"""
     with contextlib.ExitStack() as stack:
         _ = stack.enter_context(
-            open("/sys/firmware/devicetree/base/hwid,location", "r")
+            open("/sys/firmware/devicetree/base/hwid,location", "r", encoding="utf-8")
         )  # must not trigger
 
 
diff --git a/tests/functional/c/consider/consider_using_with_open.py b/tests/functional/c/consider/consider_using_with_open.py
index e93b619b7..30be92def 100644
--- a/tests/functional/c/consider/consider_using_with_open.py
+++ b/tests/functional/c/consider/consider_using_with_open.py
@@ -1,5 +1,5 @@
 # pylint: disable=missing-function-docstring, missing-module-docstring, invalid-name, import-outside-toplevel
-# pylint: disable=missing-class-docstring, too-few-public-methods, unused-variable, multiple-statements
+# pylint: disable=missing-class-docstring, too-few-public-methods, unused-variable, multiple-statements, line-too-long
 """
 The functional test for the standard ``open()`` function has to be moved in a separate file,
 because PyPy has to be excluded for the tests as the ``open()`` function is uninferable in PyPy.
@@ -8,26 +8,26 @@ PyPy from ALL functional tests.
 """
 from contextlib import contextmanager
 
-
-myfile = open("test.txt")  # [consider-using-with]
+myfile = open("test.txt", encoding="utf-8")  # [consider-using-with]
 
 
 def test_open():
-    fh = open("test.txt")  # [consider-using-with]
+    fh = open("test.txt", encoding="utf-8")  # [consider-using-with]
     fh.close()
 
-    with open("test.txt") as fh:  # must not trigger
+    with open("test.txt", encoding="utf-8") as fh:  # must not trigger
         fh.read()
 
 
 def test_open_in_enter():
     """Message must not trigger if the resource is allocated in a context manager."""
+
     class MyContextManager:
         def __init__(self):
             self.file_handle = None
 
         def __enter__(self):
-            self.file_handle = open("foo.txt", "w")  # must not trigger
+            self.file_handle = open("foo.txt", "w", encoding="utf-8")  # must not trigger
 
         def __exit__(self, exc_type, exc_value, traceback):
             self.file_handle.close()
@@ -36,31 +36,31 @@ def test_open_in_enter():
 @contextmanager
 def test_open_in_with_contextlib():
     """Message must not trigger if the resource is allocated in a context manager."""
-    file_handle = open("foo.txt", "w")  # must not trigger
+    file_handle = open("foo.txt", "w", encoding="utf-8")  # must not trigger
     yield file_handle
     file_handle.close()
 
 
 def test_open_outside_assignment():
-    open("foo").read()  # [consider-using-with]
-    content = open("foo").read()  # [consider-using-with]
+    open("foo", encoding="utf-8").read()  # [consider-using-with]
+    content = open("foo", encoding="utf-8").read()  # [consider-using-with]
 
 
 def test_open_inside_with_block():
-    with open("foo") as fh:
-        open("bar")  # [consider-using-with]
+    with open("foo", encoding="utf-8") as fh:
+        open("bar", encoding="utf-8")  # [consider-using-with]
 
 
 def test_ternary_if_in_with_block(file1, file2, which):
     """Regression test for issue #4676 (false positive)"""
-    with (open(file1) if which else open(file2)) as input_file:  # must not trigger
+    with (open(file1, encoding="utf-8") if which else open(file2, encoding="utf-8")) as input_file:  # must not trigger
         return input_file.read()
 
 
 def test_single_line_with(file1):
-    with open(file1): return file1.read()  # must not trigger
+    with open(file1, encoding="utf-8"): return file1.read()  # must not trigger
 
 
 def test_multiline_with_items(file1, file2, which):
-    with (open(file1) if which
-          else open(file2)) as input_file:  return input_file.read()
+    with (open(file1, encoding="utf-8") if which
+        else open(file2, encoding="utf-8")) as input_file: return input_file.read()
diff --git a/tests/functional/c/consider/consider_using_with_open.txt b/tests/functional/c/consider/consider_using_with_open.txt
index 63cd08b7c..ea3503123 100644
--- a/tests/functional/c/consider/consider_using_with_open.txt
+++ b/tests/functional/c/consider/consider_using_with_open.txt
@@ -1,5 +1,5 @@
-consider-using-with:12:9::Consider using 'with' for resource-allocating operations
-consider-using-with:16:9:test_open:Consider using 'with' for resource-allocating operations
+consider-using-with:11:9::Consider using 'with' for resource-allocating operations
+consider-using-with:15:9:test_open:Consider using 'with' for resource-allocating operations
 consider-using-with:45:4:test_open_outside_assignment:Consider using 'with' for resource-allocating operations
 consider-using-with:46:14:test_open_outside_assignment:Consider using 'with' for resource-allocating operations
 consider-using-with:51:8:test_open_inside_with_block:Consider using 'with' for resource-allocating operations
diff --git a/tests/functional/d/defined_and_used_on_same_line.py b/tests/functional/d/defined_and_used_on_same_line.py
index a9b9eb7d5..d5bdb4db0 100644
--- a/tests/functional/d/defined_and_used_on_same_line.py
+++ b/tests/functional/d/defined_and_used_on_same_line.py
@@ -1,5 +1,5 @@
 """Check for definitions and usage happening on the same line."""
-#pylint: disable=missing-docstring,multiple-statements,no-absolute-import,parameter-unpacking,wrong-import-position,unnecessary-comprehension
+#pylint: disable=missing-docstring,multiple-statements,no-absolute-import,parameter-unpacking,wrong-import-position,unnecessary-comprehension,unspecified-encoding
 from __future__ import print_function
 print([index
        for index in range(10)])
diff --git a/tests/functional/d/disabled_msgid_in_pylintrc.py b/tests/functional/d/disabled_msgid_in_pylintrc.py
index 09ddf1147..91cb155e1 100644
--- a/tests/functional/d/disabled_msgid_in_pylintrc.py
+++ b/tests/functional/d/disabled_msgid_in_pylintrc.py
@@ -1,6 +1,6 @@
 """https://github.com/PyCQA/pylint/issues/4265"""
 
 try:
-    f = open('test')
+    f = open('test', encoding="utf-8")
 except Exception:
     pass
diff --git a/tests/functional/n/non/non_iterator_returned.py b/tests/functional/n/non/non_iterator_returned.py
index 94c3601f5..4d0dd2a39 100644
--- a/tests/functional/n/non/non_iterator_returned.py
+++ b/tests/functional/n/non/non_iterator_returned.py
@@ -67,7 +67,7 @@ class FileBasedIterator(object):
     def __iter__(self):
         if self.file is not None:
             self.file.close()
-        self.file = open(self.path)
+        self.file = open(self.path, encoding="utf-8")
         # self file has two infered values: None and <instance of 'file'>
         # we don't want to emit error in this case
         return self.file
diff --git a/tests/functional/r/redefined_argument_from_local.py b/tests/functional/r/redefined_argument_from_local.py
index 820252781..7ccff1d18 100644
--- a/tests/functional/r/redefined_argument_from_local.py
+++ b/tests/functional/r/redefined_argument_from_local.py
@@ -1,31 +1,33 @@
 # pylint: disable=missing-docstring, unused-variable, unused-argument
 # pylint: disable=redefined-outer-name, invalid-name
 
+
 def test_redefined_in_with(name):
-    with open('something') as name: # [redefined-argument-from-local]
+    with open("something", encoding="utf-8") as name:  # [redefined-argument-from-local]
         pass
-    with open('something') as (second, name): # [redefined-argument-from-local]
+    with open("something", encoding="utf-8") as (second, name):  # [redefined-argument-from-local]
         pass
-    with open('something') as (second, (name, third)): # [redefined-argument-from-local]
+    with open("something", encoding="utf-8") as (
+        second,
+        (name, third),  # [redefined-argument-from-local]
+    ):
         pass
     other = None
-    with open('something') as other:
+    with open("something", encoding="utf-8") as other:
         pass
 
 
-
 def test_not_redefined_in_with(name):
-    with open('something') as test_redefined_in_with:
+    with open("something", encoding="utf-8") as test_redefined_in_with:
         pass
 
 
-
 def test_redefined_in_for(name):
-    for name in []: # [redefined-argument-from-local]
+    for name in []:  # [redefined-argument-from-local]
         pass
-    for (name, is_) in []: # [redefined-argument-from-local]
+    for (name, is_) in []:  # [redefined-argument-from-local]
         pass
-    for (is_, (name, _)) in []: # [redefined-argument-from-local]
+    for (is_, (name, _)) in []:  # [redefined-argument-from-local]
         pass
     for _ in []:
         pass
@@ -45,7 +47,7 @@ def test_not_redefined_in_for(name):
 def test_redefined_in_except_handler(name):
     try:
         1 / 0
-    except ZeroDivisionError as name: # [redefined-argument-from-local]
+    except ZeroDivisionError as name:  # [redefined-argument-from-local]
         pass
 
 
@@ -58,7 +60,7 @@ def test_not_redefined_in_except_handler(name):
 
 def test_not_redefined(name):
     if not name:
-        name = ''
+        name = ""
 
 
 def apply_filter(objects, filt=lambda obj: True):
diff --git a/tests/functional/r/redefined_argument_from_local.txt b/tests/functional/r/redefined_argument_from_local.txt
index 28ec50da1..b811971c6 100644
--- a/tests/functional/r/redefined_argument_from_local.txt
+++ b/tests/functional/r/redefined_argument_from_local.txt
@@ -1,7 +1,7 @@
-redefined-argument-from-local:5:30:test_redefined_in_with:Redefining argument with the local name 'name'
-redefined-argument-from-local:7:39:test_redefined_in_with:Redefining argument with the local name 'name'
-redefined-argument-from-local:9:40:test_redefined_in_with:Redefining argument with the local name 'name'
-redefined-argument-from-local:24:8:test_redefined_in_for:Redefining argument with the local name 'name'
-redefined-argument-from-local:26:9:test_redefined_in_for:Redefining argument with the local name 'name'
-redefined-argument-from-local:28:15:test_redefined_in_for:Redefining argument with the local name 'name'
-redefined-argument-from-local:48:4:test_redefined_in_except_handler:Redefining argument with the local name 'name'
+redefined-argument-from-local:6:48:test_redefined_in_with:Redefining argument with the local name 'name'
+redefined-argument-from-local:8:57:test_redefined_in_with:Redefining argument with the local name 'name'
+redefined-argument-from-local:12:9:test_redefined_in_with:Redefining argument with the local name 'name'
+redefined-argument-from-local:26:8:test_redefined_in_for:Redefining argument with the local name 'name'
+redefined-argument-from-local:28:9:test_redefined_in_for:Redefining argument with the local name 'name'
+redefined-argument-from-local:30:15:test_redefined_in_for:Redefining argument with the local name 'name'
+redefined-argument-from-local:50:4:test_redefined_in_except_handler:Redefining argument with the local name 'name'
diff --git a/tests/functional/r/regression/regression_4612_crash_pytest_fixture.py b/tests/functional/r/regression/regression_4612_crash_pytest_fixture.py
index f0aa0f5d1..f7680f1fb 100644
--- a/tests/functional/r/regression/regression_4612_crash_pytest_fixture.py
+++ b/tests/functional/r/regression/regression_4612_crash_pytest_fixture.py
@@ -5,5 +5,5 @@ import pytest
 
 @pytest.fixture
 def qm_file():
-    qm_file = open("src/test/resources/example_qm_file.csv").read()
+    qm_file = open("src/test/resources/example_qm_file.csv", encoding="utf-8").read()
     return qm_file
diff --git a/tests/functional/u/unspecified_encoding_py38.py b/tests/functional/u/unspecified_encoding_py38.py
new file mode 100644
index 000000000..20d2d7be1
--- /dev/null
+++ b/tests/functional/u/unspecified_encoding_py38.py
@@ -0,0 +1,55 @@
+"""Warnings for using open() without specifying an encoding"""
+# pylint: disable=consider-using-with
+import io
+import locale
+
+FILENAME = "foo.bar"
+open(FILENAME, "w", encoding="utf-8")
+open(FILENAME, "wb")
+open(FILENAME, "w+b")
+open(FILENAME)  # [unspecified-encoding]
+open(FILENAME, "wt")  # [unspecified-encoding]
+open(FILENAME, "w+")  # [unspecified-encoding]
+open(FILENAME, "w", encoding=None)  # [unspecified-encoding]
+open(FILENAME, "r")  # [unspecified-encoding]
+
+with open(FILENAME, encoding="utf8", errors="surrogateescape") as f:
+    pass
+
+LOCALE_ENCODING = locale.getlocale()[1]
+with open(FILENAME, encoding=LOCALE_ENCODING) as f:
+    pass
+
+with open(FILENAME) as f:  # [unspecified-encoding]
+    pass
+
+with open(FILENAME, encoding=None) as f:  # [unspecified-encoding]
+    pass
+
+LOCALE_ENCODING = None
+with open(FILENAME, encoding=LOCALE_ENCODING) as f:  # [unspecified-encoding]
+    pass
+
+io.open(FILENAME, "w+b")
+io.open_code(FILENAME)
+io.open(FILENAME)  # [unspecified-encoding]
+io.open(FILENAME, "wt")  # [unspecified-encoding]
+io.open(FILENAME, "w+")  # [unspecified-encoding]
+io.open(FILENAME, "w", encoding=None)  # [unspecified-encoding]
+
+with io.open(FILENAME, encoding="utf8", errors="surrogateescape") as f:
+    pass
+
+LOCALE_ENCODING = locale.getlocale()[1]
+with io.open(FILENAME, encoding=LOCALE_ENCODING) as f:
+    pass
+
+with io.open(FILENAME) as f:  # [unspecified-encoding]
+    pass
+
+with io.open(FILENAME, encoding=None) as f:  # [unspecified-encoding]
+    pass
+
+LOCALE_ENCODING = None
+with io.open(FILENAME, encoding=LOCALE_ENCODING) as f:  # [unspecified-encoding]
+    pass
diff --git a/tests/functional/u/unspecified_encoding_py38.rc b/tests/functional/u/unspecified_encoding_py38.rc
new file mode 100644
index 000000000..85fc502b3
--- /dev/null
+++ b/tests/functional/u/unspecified_encoding_py38.rc
@@ -0,0 +1,2 @@
+[testoptions]
+min_pyver=3.8
diff --git a/tests/functional/u/unspecified_encoding_py38.txt b/tests/functional/u/unspecified_encoding_py38.txt
new file mode 100644
index 000000000..cde7f2238
--- /dev/null
+++ b/tests/functional/u/unspecified_encoding_py38.txt
@@ -0,0 +1,15 @@
+unspecified-encoding:10:0::"Using open without explicitly specifying an encoding"
+unspecified-encoding:11:0::"Using open without explicitly specifying an encoding"
+unspecified-encoding:12:0::"Using open without explicitly specifying an encoding"
+unspecified-encoding:13:0::"Using open without explicitly specifying an encoding"
+unspecified-encoding:14:0::"Using open without explicitly specifying an encoding"
+unspecified-encoding:23:5::"Using open without explicitly specifying an encoding"
+unspecified-encoding:26:5::"Using open without explicitly specifying an encoding"
+unspecified-encoding:30:5::"Using open without explicitly specifying an encoding"
+unspecified-encoding:35:0::"Using open without explicitly specifying an encoding"
+unspecified-encoding:36:0::"Using open without explicitly specifying an encoding"
+unspecified-encoding:37:0::"Using open without explicitly specifying an encoding"
+unspecified-encoding:38:0::"Using open without explicitly specifying an encoding"
+unspecified-encoding:47:5::"Using open without explicitly specifying an encoding"
+unspecified-encoding:50:5::"Using open without explicitly specifying an encoding"
+unspecified-encoding:54:5::"Using open without explicitly specifying an encoding"
diff --git a/tests/functional/w/with_used_before_assign.py b/tests/functional/w/with_used_before_assign.py
index 64a475af1..ebe7d3093 100644
--- a/tests/functional/w/with_used_before_assign.py
+++ b/tests/functional/w/with_used_before_assign.py
@@ -1,11 +1,12 @@
-'''
+"""
 Regression test for
 https://bitbucket.org/logilab/pylint/issue/128/attributeerror-when-parsing
-'''
+"""
 from __future__ import with_statement
 
+
 def do_nothing():
     """ empty """
-    with open("") as ctx.obj:  # [undefined-variable]
+    with open("", encoding="utf-8") as ctx.obj:  # [undefined-variable]
         context.do()  # [used-before-assignment]
         context = None
diff --git a/tests/functional/w/with_used_before_assign.txt b/tests/functional/w/with_used_before_assign.txt
index 4a98cec60..8e1fe1be8 100644
--- a/tests/functional/w/with_used_before_assign.txt
+++ b/tests/functional/w/with_used_before_assign.txt
@@ -1,2 +1,2 @@
-undefined-variable:9:21:do_nothing:Undefined variable 'ctx'
-used-before-assignment:10:8:do_nothing:Using variable 'context' before assignment
+undefined-variable:10:39:do_nothing:Undefined variable 'ctx'
+used-before-assignment:11:8:do_nothing:Using variable 'context' before assignment
diff --git a/tests/lint/unittest_lint.py b/tests/lint/unittest_lint.py
index 984e9759d..004c29281 100644
--- a/tests/lint/unittest_lint.py
+++ b/tests/lint/unittest_lint.py
@@ -156,7 +156,7 @@ def create_files(paths, chroot="."):
         if not isdir(dirpath):
             os.makedirs(dirpath)
     for filepath in files:
-        with open(filepath, "w"):
+        with open(filepath, "w", encoding="utf-8"):
             pass
 
 
diff --git a/tests/test_epylint.py b/tests/test_epylint.py
index a53c2b045..525627166 100644
--- a/tests/test_epylint.py
+++ b/tests/test_epylint.py
@@ -13,7 +13,7 @@ def example_path(tmp_path):
         self.hassan()
     """
     path = tmp_path / "my_app.py"
-    with open(path, "w") as f:
+    with open(path, "w", encoding="utf-8") as f:
         f.write(content)
     return path
 
diff --git a/tests/test_func.py b/tests/test_func.py
index e15d3b0f8..8742b42f9 100644
--- a/tests/test_func.py
+++ b/tests/test_func.py
@@ -88,7 +88,7 @@ class LintTestUsingModule:
 
     def _get_expected(self):
         if self._has_output() and self.output:
-            with open(self.output) as fobj:
+            with open(self.output, encoding="utf-8") as fobj:
                 return fobj.read().strip() + "\n"
         else:
             return ""
@@ -103,7 +103,7 @@ class LintTestUpdate(LintTestUsingModule):
         except OSError:
             expected = ""
         if got != expected:
-            with open(self.output, "w") as f:
+            with open(self.output, "w", encoding="utf-8") as f:
                 f.write(got)
 
 
diff --git a/tests/test_functional.py b/tests/test_functional.py
index f3959b8e3..4ab761738 100644
--- a/tests/test_functional.py
+++ b/tests/test_functional.py
@@ -59,7 +59,7 @@ class LintModuleOutputUpdate(testutils.LintModuleTest):
             if os.path.exists(self._test_file.expected_output):
                 os.remove(self._test_file.expected_output)
             return
-        with open(self._test_file.expected_output, "w") as f:
+        with open(self._test_file.expected_output, "w", encoding="utf-8") as f:
             writer = csv.writer(f, dialect="test")
             for line in actual_output:
                 writer.writerow(line.to_csv())
diff --git a/tests/test_import_graph.py b/tests/test_import_graph.py
index 92ca4a510..07a093759 100644
--- a/tests/test_import_graph.py
+++ b/tests/test_import_graph.py
@@ -46,7 +46,7 @@ POSSIBLE_DOT_FILENAMES = ["foo.dot", "foo.gv", "tests/regrtest_data/foo.dot"]
 def test_dependencies_graph(dest):
     """DOC files are correctly generated, and the graphname is the basename"""
     imports._dependencies_graph(dest, {"labas": ["hoho", "yep"], "hoho": ["yep"]})
-    with open(dest) as stream:
+    with open(dest, encoding="utf-8") as stream:
         assert (
             stream.read().strip()
             == """
diff --git a/tests/test_self.py b/tests/test_self.py
index 937ee0394..de63ebd6f 100644
--- a/tests/test_self.py
+++ b/tests/test_self.py
@@ -553,7 +553,7 @@ class TestRunTC:
             # create module under directories which have the same name as reporter.path_strip_prefix
             # e.g. /src/some/path/src/test_target.py when reporter.path_strip_prefix = /src/
             os.makedirs(fake_path)
-            with open(module, "w") as test_target:
+            with open(module, "w", encoding="utf-8") as test_target:
                 test_target.write("a,b = object()")
 
             self._test_output(
diff --git a/tests/unittest_reporting.py b/tests/unittest_reporting.py
index 6b238b484..3946ec61d 100644
--- a/tests/unittest_reporting.py
+++ b/tests/unittest_reporting.py
@@ -128,7 +128,7 @@ def test_multi_format_output(tmp_path):
         linter.reporter.close_output_files()
         del linter.reporter
 
-    with open(json) as f:
+    with open(json, encoding="utf-8") as f:
         assert (
             f.read() == "[\n"
             "    {\n"