Merge branch 'main' into gh-144278

Author: zooba

Android/android.py

@@ -15,6 +15,7 @@
 from asyncio import wait_for
 from contextlib import asynccontextmanager
 from datetime import datetime, timezone
+from enum import IntEnum, auto
 from glob import glob
 from os.path import abspath, basename, relpath
 from pathlib import Path
@@ -61,6 +62,19 @@
 hidden_output = []
 
 
+# Based on android/log.h in the NDK.
+class LogPriority(IntEnum):
+    UNKNOWN = 0
+    DEFAULT = auto()
+    VERBOSE = auto()
+    DEBUG = auto()
+    INFO = auto()
+    WARN = auto()
+    ERROR = auto()
+    FATAL = auto()
+    SILENT = auto()
+
+
 def log_verbose(context, line, stream=sys.stdout):
     if context.verbose:
         stream.write(line)
@@ -505,47 +519,47 @@ async def logcat_task(context, initial_devices):
     pid = await wait_for(find_pid(serial), startup_timeout)
 
     # `--pid` requires API level 24 or higher.
-    args = [adb, "-s", serial, "logcat", "--pid", pid,  "--format", "tag"]
+    #
+    # `--binary` mode is used in order to detect which messages end with a
+    # newline, which most of the other modes don't indicate (except `--format
+    # long`). For example, every time pytest runs a test, it prints a "." and
+    # flushes the stream. Each "." becomes a separate log message, but we should
+    # show them all on the same line.
+    args = [adb, "-s", serial, "logcat", "--pid", pid,  "--binary"]
     logcat_started = False
     async with async_process(
-        *args, stdout=subprocess.PIPE, stderr=subprocess.STDOUT,
+        *args, stdout=subprocess.PIPE, stderr=None
     ) as process:
-        while line := (await process.stdout.readline()).decode(*DECODE_ARGS):
-            if match := re.fullmatch(r"([A-Z])/(.*)", line, re.DOTALL):
+        while True:
+            try:
+                priority, tag, message = await read_logcat(process.stdout)
                 logcat_started = True
-                level, message = match.groups()
-            else:
-                # If the regex doesn't match, this is either a logcat startup
-                # error, or the second or subsequent line of a multi-line
-                # message. Python won't produce multi-line messages, but other
-                # components might.
-                level, message = None, line
+            except asyncio.IncompleteReadError:
+                break
 
             # Exclude high-volume messages which are rarely useful.
             if context.verbose < 2 and "from python test_syslog" in message:
                 continue
 
             # Put high-level messages on stderr so they're highlighted in the
             # buildbot logs. This will include Python's own stderr.
-            stream = (
-                sys.stderr
-                if level in ["W", "E", "F"]  # WARNING, ERROR, FATAL (aka ASSERT)
-                else sys.stdout
-            )
-
-            # To simplify automated processing of the output, e.g. a buildbot
-            # posting a failure notice on a GitHub PR, we strip the level and
-            # tag indicators from Python's stdout and stderr.
-            for prefix in ["python.stdout: ", "python.stderr: "]:
-                if message.startswith(prefix):
-                    global python_started
-                    python_started = True
-                    stream.write(message.removeprefix(prefix))
-                    break
+            stream = sys.stderr if priority >= LogPriority.WARN else sys.stdout
+
+            # The app's stdout and stderr should be passed through transparently
+            # to our own corresponding streams.
+            if tag in ["python.stdout", "python.stderr"]:
+                global python_started
+                python_started = True
+                stream.write(message)
+                stream.flush()
             else:
                 # Non-Python messages add a lot of noise, but they may
-                # sometimes help explain a failure.
-                log_verbose(context, line, stream)
+                # sometimes help explain a failure. Format them in the same way
+                # as `logcat --format tag`.
+                formatted = f"{priority.name[0]}/{tag}: {message}"
+                if not formatted.endswith("\n"):
+                    formatted += "\n"
+                log_verbose(context, formatted, stream)
 
         # If the device disconnects while logcat is running, which always
         # happens in --managed mode, some versions of adb return non-zero.
@@ -556,6 +570,44 @@ async def logcat_task(context, initial_devices):
             raise CalledProcessError(status, args)
 
 
+# Read one binary log message from the given StreamReader. The message format is
+# described at https://android.stackexchange.com/a/74660. All supported versions
+# of Android use format version 2 or later.
+async def read_logcat(stream):
+    async def read_bytes(size):
+        return await stream.readexactly(size)
+
+    async def read_int(size):
+        return int.from_bytes(await read_bytes(size), "little")
+
+    payload_len = await read_int(2)
+    if payload_len < 2:
+        # 1 byte for priority, 1 byte for null terminator of tag.
+        raise ValueError(f"payload length {payload_len} is too short")
+
+    header_len = await read_int(2)
+    if header_len < 4:
+        raise ValueError(f"header length {header_len} is too short")
+    await read_bytes(header_len - 4)  # Ignore other header fields.
+
+    priority_int = await read_int(1)
+    try:
+        priority = LogPriority(priority_int)
+    except ValueError:
+        priority = LogPriority.UNKNOWN
+
+    payload_fields = (await read_bytes(payload_len - 1)).split(b"\0")
+    if len(payload_fields) < 2:
+        raise ValueError(
+            f"payload {payload!r} does not contain at least 2 "
+            f"null-separated fields"
+        )
+    tag, message, *_ = [
+        field.decode(*DECODE_ARGS) for field in payload_fields
+    ]
+    return priority, tag, message
+
+
 def stop_app(serial):
     run([adb, "-s", serial, "shell", "am", "force-stop", APP_ID], log=False)
 

Android/testbed/app/build.gradle.kts

@@ -94,10 +94,13 @@ android {
         }
 
         // This controls the API level of the maxVersion managed emulator, which is used
-        // by CI and cibuildwheel. 34 takes up too much disk space (#142289), 35 has
-        // issues connecting to the internet (#142387), and 36 and later are not
-        // available as aosp_atd images yet.
-        targetSdk = 33
+        // by CI and cibuildwheel.
+        //  * 33 has excessive buffering in the logcat client
+        //    (https://cs.android.com/android/_/android/platform/system/logging/+/d340721894f223327339010df59b0ac514308826).
+        //  * 34 consumes too much disk space on GitHub Actions (#142289).
+        //  * 35 has issues connecting to the internet (#142387).
+        //  * 36 and later are not available as aosp_atd images yet.
+        targetSdk = 32
 
         versionCode = 1
         versionName = "1.0"
@@ -130,9 +133,10 @@ android {
         path("src/main/c/CMakeLists.txt")
     }
 
-    // Set this property to something non-empty, otherwise it'll use the default
-    // list, which ignores asset directories beginning with an underscore.
-    aaptOptions.ignoreAssetsPattern = ".git"
+    // Set this property to something nonexistent but non-empty. Otherwise it'll use the
+    // default list, which ignores asset directories beginning with an underscore, and
+    // maybe also other files required by tests.
+    aaptOptions.ignoreAssetsPattern = "android-testbed-dont-ignore-anything"
 
     compileOptions {
         sourceCompatibility = JavaVersion.VERSION_1_8
@@ -234,6 +238,12 @@ androidComponents.onVariants { variant ->
                     from(cwd)
                 }
             }
+
+            // A filename ending with .gz will be automatically decompressed
+            // while building the APK. Avoid this by adding a dash to the end,
+            // and add an extra dash to any filenames that already end with one.
+            // This will be undone in MainActivity.kt.
+            rename(""".*(\.gz|-)""", "$0-")
         }
     }
 

Android/testbed/app/src/main/java/org/python/testbed/MainActivity.kt

@@ -80,7 +80,9 @@ class PythonTestRunner(val context: Context) {
                 continue
             }
             input.use {
-                File(targetSubdir, name).outputStream().use { output ->
+                // Undo the .gz workaround from build.gradle.kts.
+                val outputName = name.replace(Regex("""(.*)-"""), "$1")
+                File(targetSubdir, outputName).outputStream().use { output ->
                     input.copyTo(output)
                 }
             }

Doc/howto/logging-cookbook.rst

@@ -229,7 +229,7 @@ messages should not. Here's how you can achieve this::
    # tell the handler to use this format
    console.setFormatter(formatter)
    # add the handler to the root logger
-   logging.getLogger('').addHandler(console)
+   logging.getLogger().addHandler(console)
 
    # Now, we can log to the root logger, or any other logger. First the root...
    logging.info('Jackdaws love my big sphinx of quartz.')
@@ -650,7 +650,7 @@ the receiving end. A simple way of doing this is attaching a
 
    import logging, logging.handlers
 
-   rootLogger = logging.getLogger('')
+   rootLogger = logging.getLogger()
    rootLogger.setLevel(logging.DEBUG)
    socketHandler = logging.handlers.SocketHandler('localhost',
                        logging.handlers.DEFAULT_TCP_LOGGING_PORT)

Doc/library/os.path.rst

@@ -97,15 +97,17 @@ the :mod:`glob` module.)
 
 .. function:: commonprefix(list, /)
 
-   Return the longest path prefix (taken character-by-character) that is a
-   prefix of all paths in  *list*.  If *list* is empty, return the empty string
+   Return the longest string prefix (taken character-by-character) that is a
+   prefix of all strings in *list*.  If *list* is empty, return the empty string
    (``''``).
 
-   .. note::
+   .. warning::
 
       This function may return invalid paths because it works a
-      character at a time.  To obtain a valid path, see
-      :func:`commonpath`.
+      character at a time.
+      If you need a **common path prefix**, then the algorithm
+      implemented in this function is not secure. Use
+      :func:`commonpath` for finding a common path prefix.
 
       ::
 

Doc/library/stdtypes.rst

@@ -2163,6 +2163,8 @@ expression support in the :mod:`re` module).
 
    .. doctest::
 
+      >>> 'spam, spam, spam'.index('spam')
+      0
       >>> 'spam, spam, spam'.index('eggs')
       Traceback (most recent call last):
         File "<python-input-0>", line 1, in <module>
@@ -2546,6 +2548,20 @@ expression support in the :mod:`re` module).
 
    Like :meth:`rfind` but raises :exc:`ValueError` when the substring *sub* is not
    found.
+   For example:
+
+   .. doctest::
+
+      >>> 'spam, spam, spam'.rindex('spam')
+      12
+      >>> 'spam, spam, spam'.rindex('eggs')
+      Traceback (most recent call last):
+        File "<stdin-0>", line 1, in <module>
+          'spam, spam, spam'.rindex('eggs')
+          ~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^
+      ValueError: substring not found
+
+   See also :meth:`index` and :meth:`find`.
 
 
 .. method:: str.rjust(width, fillchar=' ', /)

Doc/library/tomllib.rst

@@ -4,19 +4,25 @@
 .. module:: tomllib
    :synopsis: Parse TOML files.
 
-.. versionadded:: 3.11
-
 .. moduleauthor:: Taneli Hukkinen
 .. sectionauthor:: Taneli Hukkinen
 
 **Source code:** :source:`Lib/tomllib`
 
 --------------
 
-This module provides an interface for parsing TOML 1.0.0 (Tom's Obvious Minimal
+This module provides an interface for parsing TOML 1.1.0 (Tom's Obvious Minimal
 Language, `https://toml.io <https://toml.io/en/>`_). This module does not
 support writing TOML.
 
+.. versionadded:: 3.11
+   The module was added with support for TOML 1.0.0.
+
+.. versionchanged:: next
+   Added TOML 1.1.0 support.
+   See the :ref:`What's New <whatsnew315-tomllib-1-1-0>` for details.
+
+
 .. seealso::
 
     The :pypi:`Tomli-W package <tomli-w>`

Doc/whatsnew/3.15.rst

@@ -867,6 +867,56 @@ tkinter
   with outdated names.
   (Contributed by Serhiy Storchaka in :gh:`143754`)
 
+
+.. _whatsnew315-tomllib-1-1-0:
+
+tomllib
+-------
+
+* The :mod:`tomllib` module now supports TOML 1.1.0.
+  This is a backwards compatible update, meaning that all valid TOML 1.0.0
+  documents are parsed the same way.
+
+  The changes, according to the `official TOML changelog`_, are:
+
+  - Allow newlines and trailing commas in inline tables.
+
+    Previously an inline table had to be on a single line and couldn't end
+    with a trailing comma. This is now relaxed so that the following is valid:
+
+    .. syntax highlighting needs TOML 1.1.0 support in Pygments,
+       see https://github.com/pygments/pygments/issues/3026
+
+    .. code-block:: text
+
+       tbl = {
+          key      = "a string",
+          moar-tbl =  {
+             key = 1,
+          },
+       }
+
+  - Add ``\xHH`` notation to basic strings for codepoints under 255,
+    and the ``\e`` escape for the escape character:
+
+    .. code-block:: text
+
+       null = "null byte: \x00; letter a: \x61"
+       csi = "\e["
+
+  - Seconds in datetime and time values are now optional.
+    The following are now valid:
+
+    .. code-block:: text
+
+       dt = 2010-02-03 14:15
+       t  = 14:15
+
+  (Contributed by Taneli Hukkinen in :gh:`142956`.)
+
+.. _official TOML changelog: https://github.com/toml-lang/toml/blob/main/CHANGELOG.md
+
+
 types
 ------
 

Lib/_android_support.py

@@ -168,6 +168,13 @@ def write(self, prio, tag, message):
         # message.
         message = message.replace(b"\x00", b"\xc0\x80")
 
+        # On API level 30 and higher, Logcat will strip any number of leading
+        # newlines. This is visible in all `logcat` modes, even --binary. Work
+        # around this by adding a leading space, which shouldn't make any
+        # difference to the log's usability.
+        if message.startswith(b"\n"):
+            message = b" " + message
+
         with self._lock:
             now = time()
             self._bucket_level += (