austin-python 2.1.2__tar.gz → 2.2.0__tar.gz

{austin_python-2.1.2 → austin_python-2.2.0}/.flake8

@@ -1,12 +1,11 @@
 [flake8]
 select = ANN,B,B9,C,D,E,F,W,I
-ignore = ANN101,ANN102,ANN401,B950,D100,D102,D103,D104,D107,E203,E501,I001,I005,W503,W606
+ignore = ANN101,ANN102,ANN401,B905,B950,D100,D102,D103,D104,D107,E203,E501,I001,I005,W503,W606
 exclude =
   austin/format/pprof/profile_pb2.py
   test/data/column.py
 max-line-length = 88
 docstring-convention = google
-import-order-style = pep8
 application-import-names = austin
 per-file-ignores =
   test/*:ANN,D

{austin_python-2.1.2 → austin_python-2.2.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: austin-python
-Version: 2.1.2
+Version: 2.2.0
 Summary: Python wrapper for Austin, the CPython frame stack sampler.
 Project-URL: homepage, https://github.com/P403n1x87/austin-python
 Project-URL: documentation, https://austin-python.readthedocs.io
@@ -15,12 +15,12 @@ Keywords: development,performance,profiling,testing
 Classifier: Development Status :: 5 - Production/Stable
 Classifier: Intended Audience :: Developers
 Classifier: License :: OSI Approved :: GNU General Public License v3 or later (GPLv3+)
-Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3.13
-Requires-Python: >=3.9
+Classifier: Programming Language :: Python :: 3.14
+Requires-Python: >=3.10
 Requires-Dist: protobuf~=3.12
 Requires-Dist: psutil>=5.7.0
 Requires-Dist: toml~=0.10
@@ -187,7 +187,7 @@ raw samples by aggregation.
 # Compatibility
 
 The latest `austin-python` package is tested on Linux, macOS and Windows with
-Python 3.9-3.13.
+Python 3.9-3.14.
 
 
 # Documentation

{austin_python-2.1.2 → austin_python-2.2.0}/README.md

@@ -159,7 +159,7 @@ raw samples by aggregation.
 # Compatibility
 
 The latest `austin-python` package is tested on Linux, macOS and Windows with
-Python 3.9-3.13.
+Python 3.9-3.14.
 
 
 # Documentation

{austin_python-2.1.2 → austin_python-2.2.0}/austin/base.py

@@ -43,7 +43,6 @@ from austin.errors import AustinError
 from austin.events import AustinMetadata
 from austin.events import AustinSample
 
-
 SemVer = Tuple[int, int, int]
 
 

{austin_python-2.1.2 → austin_python-2.2.0}/austin/events.py

@@ -1,7 +1,6 @@
 import typing as t
 from dataclasses import dataclass
 
-
 ProcessId = int
 ThreadName = str
 InterpreterId = int

{austin_python-2.1.2 → austin_python-2.2.0}/austin/format/collapsed_stack.py

@@ -1,6 +1,6 @@
+import typing as t
 from functools import singledispatchmethod
 from pathlib import Path
-import typing as t
 
 from austin.events import AustinEvent
 from austin.events import AustinEventIterator
@@ -10,7 +10,6 @@ from austin.events import AustinMetrics
 from austin.events import AustinSample
 from austin.format.mojo import MojoStreamReader
 
-
 __version__ = "0.1.0"
 
 
@@ -66,17 +65,17 @@ def parse_collapsed_stack(sample: str) -> AustinSample:
         raise InvalidSample(sample)
 
     if sample[0] != "P":
-        raise InvalidSample(f"No process ID in sample '{sample}'")
+        raise InvalidSample(f"No process ID in sample {sample!r}")
 
     head, _, metrics = sample.rpartition(" ")
     process, _, rest = head.partition(";")
     try:
         pid = int(process[1:])
     except ValueError:
-        raise InvalidSample(f"Invalid process ID in sample '{sample}'") from None
+        raise InvalidSample(f"Invalid process ID in sample {sample!r}") from None
 
     if rest[0] != "T":
-        raise InvalidSample(f"No thread ID in sample '{sample}'")
+        raise InvalidSample(f"No thread ID in sample {sample!r}")
 
     thread, _, frames_data = rest.partition(";")
     thread = thread[1:]

{austin_python-2.1.2 → austin_python-2.2.0}/austin/format/compress.py

@@ -62,9 +62,11 @@ def compress(source: AustinFileReader, dest: TextIO, counts: bool = False) -> No
         formatter.format(
             AustinSample.from_key_and_metrics(
                 key,
-                AustinMetrics(memory=metrics)
-                if is_memory
-                else AustinMetrics(time=metrics),
+                (
+                    AustinMetrics(memory=metrics)
+                    if is_memory
+                    else AustinMetrics(time=metrics)
+                ),
             )
         )
         + "\n"
@@ -108,10 +110,10 @@ def main() -> None:
 
     try:
         with (
-            AustinFileReader(args.input) as fin,
+            open(args.input) as _fin,
             open(args.output or args.input, "w") as fout,
         ):
-            compress(fin, fout, args.counts)
+            compress(AustinFileReader(_fin), fout, args.counts)
     except FileNotFoundError:
         print(f"No such input file: {args.input}")
         exit(1)

{austin_python-2.1.2 → austin_python-2.2.0}/austin/format/mojo.py

@@ -1,6 +1,5 @@
 import abc
 import asyncio
-import io
 import typing as t
 from dataclasses import dataclass
 from dataclasses import field
@@ -67,6 +66,7 @@ class MojoEvents:
     METRIC_MEMORY = 10
     STRING = 11
     STRING_REF = 12
+    STACK_REPEAT = 13
 
 
 class MojoEventHandler:
@@ -84,6 +84,7 @@ class MojoEvent:
     """MOJO event."""
 
     EVENT_ID: t.ClassVar = None
+    raw: t.ClassVar[bytes] = b""
 
     def ref(self) -> int:
         return getattr(self, fields(self)[0].name)
@@ -93,9 +94,7 @@ class MojoEvent:
         for f in fields(self):
             value = getattr(self, f.name)
             field_type = (
-                f.type.__args__[0]
-                if isinstance(f.type, t._UnionGenericAlias)
-                else f.type
+                t.get_args(f.type)[0] if t.get_origin(f.type) is t.Union else f.type
             )
             if field_type is str:
                 buffer += value.encode()
@@ -127,9 +126,11 @@ class MojoMetric(MojoEvent):
     def to_bytes(self) -> bytes:
         buffer = bytearray(
             [
-                MojoEvents.METRIC_TIME
-                if self.metric_type is MojoMetricType.TIME
-                else MojoEvents.METRIC_MEMORY
+                (
+                    MojoEvents.METRIC_TIME
+                    if self.metric_type is MojoMetricType.TIME
+                    else MojoEvents.METRIC_MEMORY
+                )
             ]
         )
         buffer += to_varint(self.value)
@@ -184,6 +185,18 @@ class MojoStack(MojoEvent):
     tid: str
 
 
+@dataclass(frozen=True, eq=True)
+class MojoStackRepeat(MojoEvent):
+    """MOJO stack repeat event.
+
+    Signals that the previous sample for this thread provides the base
+    (outermost) frames for the current sample.  The frames accumulated so
+    far in the current sample are the innermost (top) part of the call stack.
+    """
+
+    EVENT_ID = MojoEvents.STACK_REPEAT
+
+
 @dataclass(frozen=True, eq=True)
 class MojoFrame(MojoEvent):
     """MOJO frame."""
@@ -322,6 +335,11 @@ class BaseMojoStreamReader(AustinEventIterator):
         self._str_reader = str_reader()
         next(self._str_reader)
 
+        # Per-thread previous frame list for STACK_REPEAT expansion.
+        # Key: (pid, thread_name); value: list of MojoFrame accumulated by the
+        # last fully-finalised sample for that thread.
+        self._prev_frames: t.Dict[t.Tuple[int, str], t.List[MojoFrame]] = {}
+
         # Austin events
         self.metadata: t.Dict[str, str] = {}
         self.samples: t.List[AustinSample] = []
@@ -381,28 +399,61 @@ class BaseMojoStreamReader(AustinEventIterator):
                         for metric_type, metric in self._running_sample.metrics.items()
                     }
                 ),
-                frames=tuple(
-                    AustinFrame(
-                        filename=mf.filename.value,
-                        function=mf.scope.value,
-                        line=mf.line,
-                        line_end=mf.line_end,
-                        column=mf.column,
-                        column_end=mf.column_end,
+                frames=(
+                    tuple(
+                        AustinFrame(
+                            filename=mf.filename.value,
+                            function=mf.scope.value,
+                            line=mf.line,
+                            line_end=mf.line_end,
+                            column=mf.column,
+                            column_end=mf.column_end,
+                        )
+                        for mf in self._running_sample.frames
                     )
-                    for mf in self._running_sample.frames
-                )
-                if self._running_sample.frames
-                else None,
+                    if self._running_sample.frames
+                    else None
+                ),
                 gc=self._running_sample.gc,
                 idle=self._running_sample.idle,
             )
         )
 
+        # Save fully-expanded frame list for future STACK_REPEAT events from
+        # this thread.
+        self._prev_frames[(self._running_sample.pid, self._running_sample.thread)] = (
+            list(self._running_sample.frames)
+        )
+
         self._running_sample = None
 
         return sample
 
+    @staticmethod
+    def _is_native_frame(frame: MojoFrame) -> bool:
+        name = frame.filename.value
+        return not name.endswith(".py") and not (
+            name.startswith("<") and name.endswith(">")
+        )
+
+    def get_stack_repeat(self) -> MojoStackRepeat:
+        """Handle a STACK_REPEAT event.
+
+        The previous sample's frames (minus any trailing native frames that
+        belonged to the old top-of-stack above the eval frame) are prepended
+        to the frames accumulated so far.
+        """
+        assert self._running_sample is not None
+        key = (self._running_sample.pid, self._running_sample.thread)
+        prev = list(self._prev_frames.get(key, []))
+        # Strip native frames from the innermost (top) end of the previous
+        # stack — those were above the eval frame and have now been replaced
+        # by the new frames in the current sample.
+        while prev and self._is_native_frame(prev[-1]):
+            prev.pop()
+        self._running_sample.frames = prev + self._running_sample.frames
+        return MojoStackRepeat()
+
     def get_stack(self, pid: int, iid: t.Optional[int], thread: str) -> MojoStack:
         """Parse a stack."""
         if self._running_sample is not None:
@@ -611,6 +662,11 @@ class MojoStreamReader(BaseMojoStreamReader):
         """Parse string reference."""
         return self.get_string_ref(self.read_int())
 
+    @handles(MojoEvents.STACK_REPEAT)
+    def parse_stack_repeat(self) -> MojoStackRepeat:
+        """Parse a stack repeat event."""
+        return self.get_stack_repeat()
+
     def parse_event(self) -> t.Optional[MojoEvent]:
         """Parse a single event."""
         try:
@@ -670,7 +726,9 @@ class MojoStreamReader(BaseMojoStreamReader):
         if self._running_sample is not None:
             yield self._finalize_sample()
 
-    def hexdump(self, start: int, end: int, highlight: t.Set[int] = set()) -> None:  # noqa: B006
+    def hexdump(
+        self, start: int, end: int, highlight: t.Set[int] = set()  # noqa: B006
+    ) -> None:
         """Print a hexdump of the MOJO file."""
         self.mojo.seek(start)
         data = self.mojo.read(end - start)
@@ -794,6 +852,11 @@ class AsyncMojoStreamReader(BaseMojoStreamReader):
         """Parse string reference."""
         return self.get_string_ref(await self.read_int())
 
+    @handles(MojoEvents.STACK_REPEAT)
+    async def parse_stack_repeat(self) -> MojoStackRepeat:
+        """Parse a stack repeat event."""
+        return self.get_stack_repeat()
+
     async def parse_event(self) -> t.Optional[MojoEvent]:
         """Parse a single event."""
         try:
@@ -855,6 +918,8 @@ class AsyncMojoStreamReader(BaseMojoStreamReader):
 
 
 class BaseMojoStreamWriter(abc.ABC):
+    """Base class for MOJO stream writers."""
+
     HEADER = b"MOJ\x03"
 
     def __init__(self, mojo: t.Any) -> None:
@@ -904,11 +969,13 @@ class BaseMojoStreamWriter(abc.ABC):
             return mojo_frame
 
     @abc.abstractmethod
-    def write(self, event: AustinEvent) -> int: ...
+    def write(self, event: AustinEvent) -> int: ...  # noqa: E704
 
 
 class MojoStreamWriter(BaseMojoStreamWriter):
-    def __init__(self, mojo: io.BytesIO):
+    """MOJO stream writer."""
+
+    def __init__(self, mojo: t.BinaryIO) -> None:
         super().__init__(mojo)
 
         mojo.write(self.HEADER)
@@ -939,7 +1006,7 @@ class MojoStreamWriter(BaseMojoStreamWriter):
             for frame in frames:
                 size += self.mojo.write(MojoFrameReference(frame).to_bytes())
 
-            if self._gc:
+            if event.gc:
                 size += self.mojo.write(bytes([MojoEvents.GC]))
 
             if self._mode == "full":

austin_python-2.2.0/austin/format/mojo_compress.py

@@ -0,0 +1,122 @@
+# This file is part of "austin-python" which is released under GPL.
+#
+# See file LICENCE or go to http://www.gnu.org/licenses/ for full license
+# details.
+#
+# austin-python is a Python wrapper around Austin, the CPython frame stack
+# sampler.
+#
+# Copyright (c) 2018-2020 Gabriele N. Tornetta <phoenix1987@gmail.com>.
+# All rights reserved.
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+import typing as t
+from io import BytesIO
+
+from austin.events import AustinMetadata
+from austin.events import AustinMetrics
+from austin.events import AustinSample
+from austin.format.mojo import MojoStreamReader
+from austin.format.mojo import MojoStreamWriter
+
+
+def compress(source: t.BinaryIO, dest: t.BinaryIO) -> None:
+    """Compress a MOJO source stream.
+
+    Aggregates all metrics on equal stacks, where equality is determined by
+    pid, interpreter id, thread, frames, gc flag, and idle flag.
+    """
+    time_stats: t.Dict[AustinSample.Key, int] = {}
+    memory_stats: t.Dict[AustinSample.Key, int] = {}
+    metadata: t.List[AustinMetadata] = []
+    samples: t.Dict[AustinSample.Key, AustinSample] = {}
+
+    reader = MojoStreamReader(source)
+
+    for event in reader:
+        if isinstance(event, AustinMetadata):
+            metadata.append(event)
+            continue
+
+        assert isinstance(event, AustinSample)
+
+        key = event.key()
+        if key not in samples:
+            samples[key] = event
+        if event.metrics.time is not None:
+            time_stats[key] = time_stats.get(key, 0) + event.metrics.time
+        if event.metrics.memory is not None:
+            memory_stats[key] = memory_stats.get(key, 0) + event.metrics.memory
+
+    writer = MojoStreamWriter(dest)
+
+    for m in metadata:
+        writer.write(m)
+
+    for key in samples:
+        writer.write(
+            AustinSample.from_key_and_metrics(
+                key,
+                AustinMetrics(
+                    time=time_stats.get(key),
+                    memory=memory_stats.get(key),
+                ),
+            )
+        )
+
+
+def main() -> None:
+    """mojo-compress entry point."""
+    from argparse import ArgumentParser
+
+    arg_parser = ArgumentParser(
+        prog="mojo-compress",
+        description=(
+            "Compress a MOJO sample file by aggregating the collected samples."
+        ),
+    )
+
+    arg_parser.add_argument(
+        "input",
+        type=str,
+        help="The input MOJO file.",
+    )
+    arg_parser.add_argument(
+        "output",
+        type=str,
+        help="The output MOJO file; defaults to the input file for in-place compression.",
+        nargs="?",
+        default=None,
+    )
+
+    arg_parser.add_argument("-V", "--version", action="version", version="0.1.0")
+
+    args = arg_parser.parse_args()
+
+    try:
+        with open(args.input, "rb") as fin:
+            buffer = BytesIO()
+            compress(fin, buffer)
+
+        output_path = args.output or args.input
+        with open(output_path, "wb") as fout:
+            fout.write(buffer.getvalue())
+
+    except FileNotFoundError:
+        print(f"No such input file: {args.input}")
+        exit(1)
+
+
+if __name__ == "__main__":
+    main()

{austin_python-2.1.2 → austin_python-2.2.0}/austin/format/pprof/__init__.py

@@ -33,7 +33,6 @@ from austin.format import Mode
 from austin.format.pprof.profile_pb2 import Profile
 from austin.format.pprof.profile_pb2 import Sample
 
-
 PROCESS_ID_LABEL = "Process ID"
 THREAD_ID_LABEL = "Thread ID"
 INTERPRETER_ID_LABEL = "Interpreter ID"

{austin_python-2.1.2 → austin_python-2.2.0}/austin/format/speedscope.py

@@ -38,8 +38,6 @@ from austin.events import AustinMetadata
 from austin.events import AustinSample
 from austin.format import Mode
 from austin.format.collapsed_stack import AustinFileReader
-from austin.format.collapsed_stack import InvalidSample
-
 
 __version__ = "0.3.1"
 

{austin_python-2.1.2 → austin_python-2.2.0}/austin/stats.py

@@ -42,7 +42,6 @@ from austin.events import AustinSample
 from austin.events import ProcessId
 from austin.format.collapsed_stack import AustinEventCollapsedStackFormatter
 
-
 ThreadInfo = namedtuple("ThreadInfo", ["thread", "iid"])
 
 
@@ -123,9 +122,11 @@ class FrameStats(HierarchicalStats):
         if self.own:
             yield (
                 [frame],
-                AustinMetrics(time=self.own)
-                if stats_type in {AustinStatsType.CPU, AustinStatsType.WALL}
-                else AustinMetrics(memory=self.own),
+                (
+                    AustinMetrics(time=self.own)
+                    if stats_type in {AustinStatsType.CPU, AustinStatsType.WALL}
+                    else AustinMetrics(memory=self.own)
+                ),
             )
 
         if self.children:
@@ -146,14 +147,15 @@ class ThreadStats(HierarchicalStats):
         self, stats_type: AustinStatsType
     ) -> Generator[Tuple[ThreadInfo, List[AustinFrame], AustinMetrics], None, None]:
         """Collapse the hierarchical statistics."""
-
         if self.own:
             yield (
                 self.label,
                 [],
-                AustinMetrics(time=self.own)
-                if stats_type in {AustinStatsType.CPU, AustinStatsType.WALL}
-                else AustinMetrics(memory=self.own),
+                (
+                    AustinMetrics(time=self.own)
+                    if stats_type in {AustinStatsType.CPU, AustinStatsType.WALL}
+                    else AustinMetrics(memory=self.own)
+                ),
             )
 
         if self.children:

{austin_python-2.1.2 → austin_python-2.2.0}/austin/tools/diff.py

@@ -30,9 +30,11 @@ from typing import TextIO
 from typing import Tuple
 
 from austin.events import AustinFrame
+from austin.events import AustinMetrics
+from austin.format.collapsed_stack import InvalidSample
+from austin.format.collapsed_stack import parse_collapsed_stack
 from austin.format.compress import compress
 
-
 FoldedStack = List[AustinFrame]
 
 
@@ -41,7 +43,7 @@ def _similarities(
 ) -> List[Tuple[Tuple[int, int], float]]:
     """O(n * log(n)), n = len(x) * len(y)."""
 
-    def score(a: List[Frame], b: List[Frame]) -> float:
+    def score(a: List[AustinFrame], b: List[AustinFrame]) -> float:
         """Score two folded stacks."""
         if not len(a) and not len(b):
             return 1
@@ -90,12 +92,12 @@ def diff(a: TextIO, b: TextIO) -> str:
         compress(source, _)
         return _.getvalue()
 
-    def get_frames(text: str) -> List[Tuple[FoldedStack, Metric]]:
+    def get_frames(text: str) -> List[Tuple[FoldedStack, AustinMetrics]]:
         """Get the folded stacks and metrics from a string of samples."""
         x = []
         for _ in text.splitlines(keepends=False):
             try:
-                sample = Sample.parse(_)
+                sample = parse_collapsed_stack(_)
                 x.append((sample.frames, sample.metrics))
             except InvalidSample:
                 continue

{austin_python-2.1.2 → austin_python-2.2.0}/austin/tools/mojodbg.py

@@ -29,7 +29,6 @@ from traceback import print_exc
 from austin.format.mojo import MojoEvent
 from austin.format.mojo import MojoStreamReader
 
-
 __version__ = "0.1.0"
 
 

{austin_python-2.1.2 → austin_python-2.2.0}/austin/tools/resolve.py

@@ -30,7 +30,6 @@ from austin.format.mojo import MojoStreamReader
 from austin.format.mojo import MojoString
 from austin.format.mojo import to_varint
 
-
 __version__ = "0.2.0"
 
 
@@ -176,8 +175,10 @@ def resolve_mojo(input: str, output: str) -> None:
                     filename, scope, line = resolved
                     new_value = filename or scope
 
-                    event.raw = event.raw.replace(
-                        event.value.encode(), new_value.encode()
+                    object.__setattr__(
+                        event,
+                        "raw",
+                        event.raw.replace(event.value.encode(), new_value.encode()),
                     )
 
                     if filename is not None:
@@ -185,9 +186,11 @@ def resolve_mojo(input: str, output: str) -> None:
 
             elif isinstance(event, MojoFrame):
                 if event.filename.key in maps.lines:
-                    event.raw = (
+                    object.__setattr__(
+                        event,
+                        "raw",
                         event.raw[: -len(to_varint(event.line))]
-                        + maps.lines[event.filename.key]
+                        + maps.lines[event.filename.key],
                     )
 
             fout.write(event.raw)