✨ feat(Cargo.lock): Add arc-swap, portable-atomic, pyo3-log packages

🔧 fix(Cargo.lock): Update pyo3, pyo3-build-config, pyo3-ffi, pyo3-macros, pyo3-macros-backend versions to 0.20.3
🔧 fix(pyproject.toml): Update shazamio_core version to 1.1.0-rc.2
🔧 fix(Cargo.toml): Update shazamio-core version to 1.1.0-rc.2, add pyo3-log and log dependencies
🔧 fix(src/lib.rs): Add logging initialization and messages, update segment duration to 10 seconds
🔧 fix(src/fingerprinting/algorithm.rs): Update segment duration to 10 seconds
🔧 fix(shazamio_core/shazamio_core.py): Update segment duration to 10 seconds, add type hints and docstrings
🔧 fix(shazamio_core/shazamio_core.pyi): Update segment duration to 10 seconds, add type hints and docstrings
🔧 fix(.github/workflows/CI.yml): Update comments for macOS targets

Author: dotX12

.github/workflows/CI.yml

@@ -44,7 +44,7 @@ jobs:
       fail-fast: false
       matrix:
         include:
-          # macOS 13 (x86_64) со всеми версиями Python
+          # macOS 13 (x86_64)
           - os: macos-13
             target: x86_64-apple-darwin
             python-version: "3.9"
@@ -58,7 +58,7 @@ jobs:
             target: x86_64-apple-darwin
             python-version: "3.12"
 
-          # macOS-latest (arm64) со всеми версиями Python
+          # macOS-latest (arm64)
           - os: macos-latest
             target: aarch64-apple-darwin
             python-version: "3.9"

Cargo.lock

@@ -1,6 +1,6 @@
 [package]
 name = "shazamio-core"
-version = "1.1.0-rc.1"
+version = "1.1.0-rc.2"
 edition = "2021"
 rust-version = "1.62"
 
@@ -23,8 +23,10 @@ futures = { version = "0.3.30", features = [] }
 serde = { version = "1.0.196", features = ["derive"] }
 bytes = "1.5.0"
 tempdir = "0.3.7"
-pyo3 = "0.20.2"
+pyo3 = "0.20.3"
 pyo3-asyncio = { version = "0.20.0", features = ["async-std-runtime", "async-std", "tokio", "tokio-runtime"] }
+pyo3-log = "=0.8.4"
+log = "0.4.20"
 
 [features]
 default = ["pyo3/extension-module"]

Cargo.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "shazamio_core"
-version = "1.1.0-rc.1"
+version = "1.1.0-rc.2"
 description = ""
 authors = ["dotX12 <dev@shitposting.team>"]
 readme = "README.md"

pyproject.toml

@@ -1,4 +1,6 @@
 from dataclasses import dataclass
+from typing import Union
+from os import PathLike
 
 
 @dataclass
@@ -35,21 +37,47 @@ def __repr__(self) -> str:
 
 
 class Recognizer:
-    def __init__(self, segment_duration_seconds: int = 12) -> None:
+    """
+    Recognizer uses a Rust implementation under the hood.
+
+    This class provides an interface for recognizing audio files, but the actual
+    processing logic is implemented in Rust and accessed via FFI.
+    """
+
+    def __init__(self, segment_duration_seconds: int = 10) -> None:
+        """
+        :param segment_duration_seconds: The duration (in seconds) of the audio segment to analyze.
+            - **Default:** 12 seconds.
+            - **If the audio file is longer than this duration**, a centered segment of the specified duration is selected.
+              - Example: If the audio is **60 seconds** and `segment_duration_seconds = 10`, the extracted segment will be **from 25s to 35s**.
+            - **If the audio file is shorter than this duration**, the entire file is used.
+              - Example: If the audio is **8 seconds** and `segment_duration_seconds = 10`, the entire **8-second file** will be processed.
+            - **Audio is always converted to mono and downsampled to 16 kHz** before analysis.
+            - This parameter determines the number of samples used for frequency analysis and fingerprint generation.
+        """
+        self.segment_duration_seconds = segment_duration_seconds
         raise NotImplemented
 
-    async def recognize_path(self, value: str) -> Signature:
+    async def recognize_path(self, value: Union[str, PathLike]) -> Signature:
         """
-        :param value: path file
-        :return: Signature object
-        :raises SignatureError: if there is any error
+        Recognize audio from a file path.
+
+        This method is a Python wrapper around a Rust implementation.
+
+        :param value: Path to an audio file.
+        :return: Signature object.
+        :raises SignatureError: if an error occurs.
         """
         raise NotImplemented
 
     async def recognize_bytes(self, value: bytes) -> Signature:
         """
-        :param value: bytes file
-        :return: Signature object
-        :raises SignatureError: if there is any error
+        Recognize audio from raw bytes.
+
+        This method is a Python wrapper around a Rust implementation.
+
+        :param value: Raw audio file as bytes.
+        :return: Signature object.
+        :raises SignatureError: if an error occurs.
         """
         raise NotImplemented

shazamio_core/shazamio_core.py

@@ -1,47 +1,83 @@
 from dataclasses import dataclass
+from typing import Union
+from os import PathLike
+
 
 @dataclass
 class Geolocation:
     altitude: int
     latitude: int
     longitude: int
 
+
 @dataclass
 class SignatureSong:
     samples: int
     timestamp: int
     uri: str
 
+
 @dataclass
 class Signature:
     geolocation: Geolocation
     signature: SignatureSong
     timestamp: int
     timezone: str
 
+
 class SignatureError(Exception):
     def __init__(self, message: str):
         self.message = message
+
     def __str__(self) -> str:
         return self.message
+
     def __repr__(self) -> str:
         return f"SignatureError({self.message})"
 
+
 class Recognizer:
-    def __init__(self, segment_duration_seconds: int = 12) -> None:
+    """
+    Recognizer uses a Rust implementation under the hood.
+
+    This class provides an interface for recognizing audio files, but the actual
+    processing logic is implemented in Rust and accessed via FFI.
+    """
+
+    def __init__(self, segment_duration_seconds: int = 10) -> None:
+        """
+        :param segment_duration_seconds: The duration (in seconds) of the audio segment to analyze.
+            - **Default:** 12 seconds.
+            - **If the audio file is longer than this duration**, a centered segment of the specified duration is selected.
+              - Example: If the audio is **60 seconds** and `segment_duration_seconds = 10`, the extracted segment will be **from 25s to 35s**.
+            - **If the audio file is shorter than this duration**, the entire file is used.
+              - Example: If the audio is **8 seconds** and `segment_duration_seconds = 10`, the entire **8-second file** will be processed.
+            - **Audio is always converted to mono and downsampled to 16 kHz** before analysis.
+            - This parameter determines the number of samples used for frequency analysis and fingerprint generation.
+        """
+        self.segment_duration_seconds = segment_duration_seconds
         raise NotImplemented
 
-    async def recognize_path(self, value: str) -> Signature:
+    async def recognize_path(self, value: Union[str, PathLike]) -> Signature:
         """
-        :param value: path file
-        :return: Signature object
-        :raises SignatureError: if there is any error
+        Recognize audio from a file path.
+
+        This method is a Python wrapper around a Rust implementation.
+
+        :param value: Path to an audio file.
+        :return: Signature object.
+        :raises SignatureError: if an error occurs.
         """
         raise NotImplemented
+
     async def recognize_bytes(self, value: bytes) -> Signature:
         """
-        :param value: bytes file
-        :return: Signature object
-        :raises SignatureError: if there is any error
+        Recognize audio from raw bytes.
+
+        This method is a Python wrapper around a Rust implementation.
+
+        :param value: Raw audio file as bytes.
+        :return: Signature object.
+        :raises SignatureError: if an error occurs.
         """
         raise NotImplemented

shazamio_core/shazamio_core.pyi

@@ -35,7 +35,7 @@ impl SignatureGenerator {
         let raw_pcm_samples: Vec<i16> = converted_file.collect();
 
         // Process the PCM samples as in make_signature_from_buffer
-        let duration_seconds = segment_duration_seconds.unwrap_or(12);
+        let duration_seconds = segment_duration_seconds.unwrap_or(10);
         let sample_rate = 16000;
         let segment_samples = (duration_seconds * sample_rate) as usize;
 
@@ -74,8 +74,8 @@ impl SignatureGenerator {
         }
 
         // Downsample the raw PCM samples to 16 KHz, and skip to the middle of the file
-        // in order to increase recognition odds. Take N (12 default) seconds of sample.
-        let duration_seconds = segment_duration_seconds.unwrap_or(12);
+        // in order to increase recognition odds. Take N (10 default) seconds of sample.
+        let duration_seconds = segment_duration_seconds.unwrap_or(10);
         let sample_rate = 16000;
         let segment_samples = (duration_seconds * sample_rate) as usize;
 
@@ -187,10 +187,8 @@ impl SignatureGenerator {
                 .max(spread_fft_results[position + 2]);
         }
 
-        // Сначала скопируем данные, чтобы избежать одновременной мутации.
         let spread_fft_results_copy = spread_fft_results.clone();
 
-        // Теперь, используя копию, мы можем обновить исходные данные без конфликта мутации.
         for position in 0..=1024 {
             for former_fft_number in &[1, 3, 6] {
                 let former_fft_output = &mut self.spread_fft_outputs