Removed old data module

batdetect2/compat/data.py

@@ -1,6 +1,5 @@
 """Compatibility functions between old and new data structures."""
 
-import json
 import os
 import uuid
 from pathlib import Path
@@ -18,7 +17,8 @@ PathLike = Union[Path, str, os.PathLike]
 
 __all__ = [
     "convert_to_annotation_group",
-    "load_annotation_project_from_dir",
+    "load_file_annotation",
+    "annotation_to_sound_event",
 ]
 
 SPECIES_TAG_KEY = "species"
@@ -304,85 +304,3 @@ def list_file_annotations(path: PathLike) -> List[Path]:
     """List all annotations in a directory."""
     path = Path(path)
     return [file for file in path.glob("*.json")]
-
-
-def load_annotation_project_from_file(
-    path: PathLike,
-    name: Optional[str] = None,
-    audio_dir: Optional[PathLike] = None,
-) -> data.AnnotationProject:
-    old_annotations = json.loads(Path(path).read_text())
-
-    annotations = []
-    tasks = []
-
-    for ann in old_annotations:
-        try:
-            ann = FileAnnotation.model_validate(ann)
-        except ValueError:
-            continue
-
-        try:
-            clip = file_annotation_to_clip(ann, audio_dir=audio_dir)
-        except FileNotFoundError:
-            continue
-
-        annotations.append(file_annotation_to_clip_annotation(ann, clip))
-        tasks.append(file_annotation_to_annotation_task(ann, clip))
-
-    return data.AnnotationProject(
-        name=name or str(path),
-        clip_annotations=annotations,
-        tasks=tasks,
-    )
-
-
-def load_annotation_project_from_dir(
-    path: PathLike,
-    name: Optional[str] = None,
-    audio_dir: Optional[PathLike] = None,
-) -> data.AnnotationProject:
-    """Convert annotations to annotation project."""
-    audio_dir = audio_dir or Path.cwd()
-
-    paths = list_file_annotations(path)
-
-    if name is None:
-        name = str(path)
-
-    annotations = []
-    tasks = []
-
-    for p in paths:
-        try:
-            file_annotation = load_file_annotation(p)
-        except FileNotFoundError:
-            continue
-
-        try:
-            clip = file_annotation_to_clip(
-                file_annotation,
-                audio_dir=audio_dir,
-            )
-        except FileNotFoundError:
-            continue
-
-        annotations.append(
-            file_annotation_to_clip_annotation(
-                file_annotation,
-                clip,
-            )
-        )
-
-        tasks.append(
-            file_annotation_to_annotation_task(
-                file_annotation,
-                clip,
-            )
-        )
-
-    return data.AnnotationProject(
-        name=name,
-        clip_annotations=annotations,
-        tasks=tasks,
-    )

batdetect2/compat/params.py

@@ -1,7 +1,6 @@
 from batdetect2.preprocess import (
     AmplitudeScaleConfig,
     AudioConfig,
-    STFTConfig,
     FrequencyConfig,
     LogScaleConfig,
     PcenScaleConfig,
@@ -10,6 +9,7 @@ from batdetect2.preprocess import (
     Scales,
     SpecSizeConfig,
     SpectrogramConfig,
+    STFTConfig,
 )
 from batdetect2.preprocess.spectrogram import get_spectrogram_resolution
 from batdetect2.terms import TagInfo

batdetect2/data.py

@@ -1,119 +0,0 @@
-from pathlib import Path
-from typing import List, Literal, Optional, Tuple, Union
-
-from pydantic import Field
-from soundevent import data, io
-
-from batdetect2.compat.data import (
-    load_annotation_project_from_dir,
-    load_annotation_project_from_file,
-)
-from batdetect2.configs import BaseConfig, load_config
-
-__all__ = [
-    "load_datasets_from_config",
-]
-
-
-class BatDetect2AnnotationFiles(BaseConfig):
-    format: Literal["batdetect2"] = "batdetect2"
-    path: Path
-
-
-class BatDetect2AnnotationFile(BaseConfig):
-    format: Literal["batdetect2_file"] = "batdetect2_file"
-    path: Path
-
-
-class AOEFAnnotationFile(BaseConfig):
-    format: Literal["aoef"] = "aoef"
-    path: Path
-
-
-AnnotationFormats = Union[
-    BatDetect2AnnotationFiles,
-    BatDetect2AnnotationFile,
-    AOEFAnnotationFile,
-]
-
-
-class DatasetInfo(BaseConfig):
-    name: str
-    audio_dir: Path
-    annotations: AnnotationFormats = Field(discriminator="format")
-
-
-class DatasetsConfig(BaseConfig):
-    train: List[DatasetInfo] = Field(default_factory=list)
-    test: List[DatasetInfo] = Field(default_factory=list)
-
-
-def load_dataset(
-    info: DatasetInfo,
-    audio_dir: Optional[Path] = None,
-    base_dir: Optional[Path] = None,
-) -> data.AnnotationProject:
-    audio_dir = (
-        info.audio_dir if base_dir is None else base_dir / info.audio_dir
-    )
-
-    path = (
-        info.annotations.path
-        if base_dir is None
-        else base_dir / info.annotations.path
-    )
-
-    if info.annotations.format == "batdetect2":
-        return load_annotation_project_from_dir(
-            path,
-            name=info.name,
-            audio_dir=audio_dir,
-        )
-
-    if info.annotations.format == "batdetect2_file":
-        return load_annotation_project_from_file(
-            path,
-            name=info.name,
-            audio_dir=audio_dir,
-        )
-
-    if info.annotations.format == "aoef":
-        return io.load(  # type: ignore
-            info.annotations.path,
-            audio_dir=audio_dir,
-        )
-
-    raise NotImplementedError(
-        f"Unknown annotation format: {info.annotations.name}"
-    )
-
-
-def load_datasets(
-    config: DatasetsConfig,
-    base_dir: Optional[Path] = None,
-) -> Tuple[List[data.ClipAnnotation], List[data.ClipAnnotation]]:
-    test_annotations = []
-    train_annotations = []
-
-    for dataset in config.train:
-        project = load_dataset(dataset, base_dir=base_dir)
-        train_annotations.extend(project.clip_annotations)
-
-    for dataset in config.test:
-        project = load_dataset(dataset, base_dir=base_dir)
-        test_annotations.extend(project.clip_annotations)
-
-    return train_annotations, test_annotations
-
-
-def load_datasets_from_config(
-    path: data.PathLike,
-    field: Optional[str] = None,
-    base_dir: Optional[Path] = None,
-):
-    config = load_config(
-        path=path,
-        schema=DatasetsConfig,
-        field=field,
-    )
-    return load_datasets(config, base_dir=base_dir)