improve schema generation:

- use oneOf instead of anyOf
- remove duplicated required strings automatically (not by hand)

schema/content_list.json

@@ -1,7 +1,7 @@
 {
     "type": "array",
     "items": {
-        "anyOf": [
+        "oneOf": [
             {
                 "$ref": "#/$defs/EpisodeContent"
             },

src/content/base_class.py

@@ -28,14 +28,11 @@ class ContentDict(TypedDict):
     scanned_file: ScannedFile
 
 
-
-
 @dataclass(slots=True, repr=True)
 class Content:
     __type: ContentType = field(metadata=alias("type"))
     __scanned_file: ScannedFile = field(metadata=alias("scanned_file"))
 
-
     def summary(self: Self, _detailed: bool = False) -> Summary:
         raise MissingOverrideError
 
@@ -66,7 +63,6 @@ def scan(
         raise MissingOverrideError
 
 
-
 def process_folder(
     directory: Path,
     callback: Callback[Content, ContentCharacteristic, Manager],

src/content/collection_content.py

@@ -9,7 +9,7 @@
     cast,
 )
 
-from apischema import alias
+from apischema import alias, schema
 from classifier import Classifier
 from enlighten import Manager
 from typing_extensions import override
@@ -27,6 +27,7 @@
     NameParser,
     ScannedFile,
     Summary,
+    deduplicate_required,
 )
 from content.series_content import SeriesContent
 
@@ -36,6 +37,7 @@ class CollectionContentDict(ContentDict):
     series: list[SeriesContent]
 
 
+@schema(extra=deduplicate_required)
 @dataclass(slots=True, repr=True)
 class CollectionContent(Content):
     __type: Literal[ContentType.collection] = field(metadata=alias("type"))

src/content/episode_content.py

@@ -9,7 +9,7 @@
     Self,
 )
 
-from apischema import alias
+from apischema import alias, schema
 from classifier import Classifier, FileMetadataError, Language, WAVFile
 from enlighten import Manager
 from typing_extensions import override
@@ -22,6 +22,7 @@
     NameParser,
     ScannedFile,
     Summary,
+    deduplicate_required,
 )
 
 
@@ -50,6 +51,7 @@ def itr_print_percent() -> None:
     print(f"{percent:.02f} %")
 
 
+@schema(extra=deduplicate_required)
 @dataclass(slots=True, repr=True)
 class EpisodeContent(Content):
     __type: Literal[ContentType.episode] = field(metadata=alias("type"))

src/content/general.py

@@ -488,3 +488,12 @@ def safe_index(ls: list[SF], item: SF) -> Optional[int]:
         return ls.index(item)
     except ValueError:
         return None
+
+
+def deduplicate_required(schema: dict[str, Any]) -> None:
+    if schema.get("required") is not None and isinstance(schema["required"], list):
+        result: list[Any] = []
+        for element in schema["required"]:
+            if element not in result:
+                result.append(element)
+        schema["required"] = result

src/content/season_content.py

@@ -10,7 +10,7 @@
     cast,
 )
 
-from apischema import alias
+from apischema import alias, schema
 from classifier import Classifier
 from enlighten import Manager
 from typing_extensions import override
@@ -29,6 +29,7 @@
     ScannedFile,
     SeasonDescription,
     Summary,
+    deduplicate_required,
 )
 
 
@@ -37,6 +38,7 @@ class SeasonContentDict(ContentDict):
     episodes: list[EpisodeContent]
 
 
+@schema(extra=deduplicate_required)
 @dataclass(slots=True, repr=True)
 class SeasonContent(Content):
     __type: Literal[ContentType.season] = field(metadata=alias("type")) # TODO: submit upstream path, to allow this: (to not add "type" in the required field twice)

src/content/series_content.py

@@ -10,7 +10,7 @@
     cast,
 )
 
-from apischema import alias
+from apischema import alias, schema
 from classifier import Classifier
 from enlighten import Manager
 from typing_extensions import override
@@ -28,6 +28,7 @@
     ScannedFile,
     SeriesDescription,
     Summary,
+    deduplicate_required,
 )
 from content.season_content import SeasonContent
 
@@ -37,6 +38,7 @@ class SeriesContentDict(ContentDict):
     seasons: list[SeasonContent]
 
 
+@schema(extra=deduplicate_required)
 @dataclass(slots=True, repr=True)
 class SeriesContent(Content):
     __type: Literal[ContentType.series] = field(metadata=alias("type"))

src/main.py

@@ -4,9 +4,9 @@
 import json
 from os import makedirs
 from pathlib import Path
-from typing import TYPE_CHECKING, Any, Optional, Self, TypedDict
+from typing import TYPE_CHECKING, Annotated, Any, Optional, Self, TypedDict
 
-from apischema import deserialize, serialize
+from apischema import deserialize, schema, serialize
 from apischema.json_schema import (
     deserialization_schema,
     serialization_schema,
@@ -202,7 +202,20 @@ def __del__(self: Self) -> None:
         self.__manager.stop()
 
 
-AllContent = EpisodeContent | SeasonContent | SeriesContent | CollectionContent
+# from: https://wyfo.github.io/apischema/0.18/json_schema/
+# schema extra can be callable to modify the schema in place
+def to_one_of(schema: dict[str, Any]) -> None:
+    if "anyOf" in schema:
+        schema["oneOf"] = schema.pop("anyOf")
+
+
+OneOf = schema(extra=to_one_of)
+
+
+AllContent = Annotated[
+    EpisodeContent | SeasonContent | SeriesContent | CollectionContent,
+    OneOf,
+]
 
 
 def load_from_file(file_path: Path) -> list[Content]: