Simplify away safe_text_dupfile/EncodedFile

I tried to understand what the `safe_text_dupfile()` function and
`EncodedFile` class do. Outside tests, `EncodedFile` is only used by
`safe_text_dupfile`, and `safe_text_dupfile` is only used by
`FDCaptureBinary.__init__()`. I then started to eliminate always-true
conditions based on the single call site, and in the end nothing was
left.

The old code had some workarounds:

- Ensure `errors` property is set: pytest-dev#555.
  It is set now, although now it will return `replace`, which is the
  actual method used also before, instead of `strict`. AFAIU, the issue
  just wanted it to be set to something, not `strict` specifically.

- Infinite recursion trying to pickle `EncodedFile` (f7282b8).
  Not sure, I don't think this object should be pickle-able at all (and
  it isn't...).

- A bug in `logging.Handler.__repr__` with int `file.name`.
  pytest-dev#2555
  Was fixed in Python>=3.7.4: https://bugs.python.org/issue36015.
  TODO: This workaround is probably still needed?

src/_pytest/capture.py

@@ -9,9 +9,7 @@
 import sys
 from io import UnsupportedOperation
 from tempfile import TemporaryFile
-from typing import BinaryIO
 from typing import Generator
-from typing import Iterable
 from typing import Optional
 
 import pytest
@@ -390,55 +388,6 @@ def disabled(self):
             yield
 
 
-def safe_text_dupfile(f, mode, default_encoding="UTF8"):
-    """ return an open text file object that's a duplicate of f on the
-        FD-level if possible.
-    """
-    encoding = getattr(f, "encoding", None)
-    try:
-        fd = f.fileno()
-    except Exception:
-        if "b" not in getattr(f, "mode", "") and hasattr(f, "encoding"):
-            # we seem to have a text stream, let's just use it
-            return f
-    else:
-        newfd = os.dup(fd)
-        if "b" not in mode:
-            mode += "b"
-        f = os.fdopen(newfd, mode, 0)  # no buffering
-    return EncodedFile(f, encoding or default_encoding)
-
-
-class EncodedFile:
-    errors = "strict"  # possibly needed by py3 code (issue555)
-
-    def __init__(self, buffer: BinaryIO, encoding: str) -> None:
-        self.buffer = buffer
-        self.encoding = encoding
-
-    def write(self, s: str) -> int:
-        if not isinstance(s, str):
-            raise TypeError(
-                "write() argument must be str, not {}".format(type(s).__name__)
-            )
-        return self.buffer.write(s.encode(self.encoding, "replace"))
-
-    def writelines(self, lines: Iterable[str]) -> None:
-        self.buffer.writelines(x.encode(self.encoding, "replace") for x in lines)
-
-    @property
-    def name(self) -> str:
-        """Ensure that file.name is a string."""
-        return repr(self.buffer)
-
-    @property
-    def mode(self) -> str:
-        return self.buffer.mode.replace("b", "")
-
-    def __getattr__(self, name):
-        return getattr(object.__getattribute__(self, "buffer"), name)
-
-
 CaptureResult = collections.namedtuple("CaptureResult", ["out", "err"])
 
 
@@ -548,9 +497,7 @@ def __init__(self, targetfd, tmpfile=None):
                 self.syscapture = SysCapture(targetfd)
             else:
                 if tmpfile is None:
-                    f = TemporaryFile()
-                    with f:
-                        tmpfile = safe_text_dupfile(f, mode="wb+")
+                    tmpfile = TemporaryFile("wt+", encoding="utf-8", errors="replace")
                 if targetfd in patchsysdict:
                     self.syscapture = SysCapture(targetfd, tmpfile)
                 else:
@@ -579,7 +526,7 @@ def _start(self):
 
     def snap(self):
         self.tmpfile.seek(0)
-        res = self.tmpfile.read()
+        res = self.tmpfile.buffer.read()
         self.tmpfile.seek(0)
         self.tmpfile.truncate()
         return res
@@ -621,10 +568,10 @@ class FDCapture(FDCaptureBinary):
     EMPTY_BUFFER = str()  # type: ignore
 
     def snap(self):
-        res = super().snap()
-        enc = getattr(self.tmpfile, "encoding", None)
-        if enc and isinstance(res, bytes):
-            res = str(res, enc, "replace")
+        self.tmpfile.seek(0)
+        res = self.tmpfile.read()
+        self.tmpfile.seek(0)
+        self.tmpfile.truncate()
         return res
 
 

testing/test_capture.py

@@ -1,16 +1,12 @@
 import contextlib
 import io
 import os
-import pickle
 import subprocess
 import sys
 import textwrap
-from io import StringIO
 from io import UnsupportedOperation
 from typing import BinaryIO
 from typing import Generator
-from typing import List
-from typing import TextIO
 
 import pytest
 from _pytest import capture
@@ -865,49 +861,6 @@ def tmpfile(testdir) -> Generator[BinaryIO, None, None]:
         f.close()
 
 
-@needsosdup
-def test_dupfile(tmpfile) -> None:
-    flist = []  # type: List[TextIO]
-    for i in range(5):
-        nf = capture.safe_text_dupfile(tmpfile, "wb")
-        assert nf != tmpfile
-        assert nf.fileno() != tmpfile.fileno()
-        assert nf not in flist
-        print(i, end="", file=nf)
-        flist.append(nf)
-
-    fname_open = flist[0].name
-    assert fname_open == repr(flist[0].buffer)
-
-    for i in range(5):
-        f = flist[i]
-        f.close()
-    fname_closed = flist[0].name
-    assert fname_closed == repr(flist[0].buffer)
-    assert fname_closed != fname_open
-    tmpfile.seek(0)
-    s = tmpfile.read()
-    assert "01234" in repr(s)
-    tmpfile.close()
-    assert fname_closed == repr(flist[0].buffer)
-
-
-def test_dupfile_on_bytesio():
-    bio = io.BytesIO()
-    f = capture.safe_text_dupfile(bio, "wb")
-    f.write("hello")
-    assert bio.getvalue() == b"hello"
-    assert "BytesIO object" in f.name
-
-
-def test_dupfile_on_textio():
-    sio = StringIO()
-    f = capture.safe_text_dupfile(sio, "wb")
-    f.write("hello")
-    assert sio.getvalue() == "hello"
-    assert not hasattr(f, "name")
-
-
 @contextlib.contextmanager
 def lsof_check():
     pid = os.getpid()
@@ -1355,8 +1308,8 @@ def test_error_attribute_issue555(testdir):
         """
         import sys
         def test_capattr():
-            assert sys.stdout.errors == "strict"
-            assert sys.stderr.errors == "strict"
+            assert sys.stdout.errors == "replace"
+            assert sys.stderr.errors == "replace"
     """
     )
     reprec = testdir.inline_run()
@@ -1431,15 +1384,6 @@ def test_spam_in_thread():
     result.stdout.no_fnmatch_line("*IOError*")
 
 
-def test_pickling_and_unpickling_encoded_file():
-    # See https://bitbucket.org/pytest-dev/pytest/pull-request/194
-    # pickle.loads() raises infinite recursion if
-    # EncodedFile.__getattr__ is not implemented properly
-    ef = capture.EncodedFile(None, None)
-    ef_as_str = pickle.dumps(ef)
-    pickle.loads(ef_as_str)
-
-
 def test_global_capture_with_live_logging(testdir):
     # Issue 3819
     # capture should work with live cli logging
@@ -1555,18 +1499,6 @@ def test_stderr_write_returns_len(capsys):
     assert sys.stderr.write("Foo") == 3
 
 
-def test_encodedfile_writelines(tmpfile: BinaryIO) -> None:
-    ef = capture.EncodedFile(tmpfile, "utf-8")
-    with pytest.raises(AttributeError):
-        ef.writelines([b"line1", b"line2"])  # type: ignore[list-item]  # noqa: F821
-    assert ef.writelines(["line1", "line2"]) is None  # type: ignore[func-returns-value]  # noqa: F821
-    tmpfile.seek(0)
-    assert tmpfile.read() == b"line1line2"
-    tmpfile.close()
-    with pytest.raises(ValueError):
-        ef.read()
-
-
 def test__get_multicapture() -> None:
     assert isinstance(_get_multicapture("fd"), MultiCapture)
     pytest.raises(ValueError, _get_multicapture, "unknown").match(