fix: wrap mutable ColumnContent defaults in field(default_factory=...) for Python 3.13 dataclass compatibility

src/columns.py

@@ -1,8 +1,4 @@
-from dataclasses import dataclass, make_dataclass
-
-
-def _fields(raw_class):
-    return [v for k, v in raw_class.__dict__.items() if k[:2] != "__" and k[-2:] != "__"]
+from dataclasses import dataclass, field, make_dataclass
 
 
 # These classes are for user facing column names,
@@ -19,37 +15,43 @@ class ColumnContent:
 
 def get_default_auto_eval_column_dict():
     auto_eval_column_dict = []
-    auto_eval_column_dict.append(["rank", ColumnContent, ColumnContent(COL_NAME_RANK, "number", True)])
+    auto_eval_column_dict.append(
+        ["rank", ColumnContent, field(default_factory=lambda: ColumnContent(COL_NAME_RANK, "number", True))]
+    )
     auto_eval_column_dict.append(
         [
             "retrieval_model",
             ColumnContent,
-            ColumnContent(COL_NAME_RETRIEVAL_MODEL, "markdown", True, never_hidden=True),
+            field(default_factory=lambda: ColumnContent(COL_NAME_RETRIEVAL_MODEL, "markdown", True, never_hidden=True)),
         ]
     )
     auto_eval_column_dict.append(
         [
             "reranking_model",
             ColumnContent,
-            ColumnContent(COL_NAME_RERANKING_MODEL, "markdown", True, never_hidden=True),
+            field(default_factory=lambda: ColumnContent(COL_NAME_RERANKING_MODEL, "markdown", True, never_hidden=True)),
         ]
     )
     auto_eval_column_dict.append(
-        ["revision", ColumnContent, ColumnContent(COL_NAME_REVISION, "markdown", True, never_hidden=True)]
+        ["revision", ColumnContent, field(default_factory=lambda: ColumnContent(COL_NAME_REVISION, "markdown", True, never_hidden=True))]
+    )
+    auto_eval_column_dict.append(
+        ["timestamp", ColumnContent, field(default_factory=lambda: ColumnContent(COL_NAME_TIMESTAMP, "date", True, never_hidden=True))]
     )
     auto_eval_column_dict.append(
-        ["timestamp", ColumnContent, ColumnContent(COL_NAME_TIMESTAMP, "date", True, never_hidden=True)]
+        ["average", ColumnContent, field(default_factory=lambda: ColumnContent(COL_NAME_AVG, "number", True))]
     )
-    auto_eval_column_dict.append(["average", ColumnContent, ColumnContent(COL_NAME_AVG, "number", True)])
     auto_eval_column_dict.append(
         [
             "retrieval_model_link",
             ColumnContent,
-            ColumnContent(
-                COL_NAME_RETRIEVAL_MODEL_LINK,
-                "markdown",
-                False,
-                hidden=True,
+            field(
+                default_factory=lambda: ColumnContent(
+                    COL_NAME_RETRIEVAL_MODEL_LINK,
+                    "markdown",
+                    False,
+                    hidden=True,
+                )
             ),
         ]
     )
@@ -57,16 +59,18 @@ def get_default_auto_eval_column_dict():
         [
             "reranking_model_link",
             ColumnContent,
-            ColumnContent(
-                COL_NAME_RERANKING_MODEL_LINK,
-                "markdown",
-                False,
-                hidden=True,
+            field(
+                default_factory=lambda: ColumnContent(
+                    COL_NAME_RERANKING_MODEL_LINK,
+                    "markdown",
+                    False,
+                    hidden=True,
+                )
             ),
         ]
     )
     auto_eval_column_dict.append(
-        ["is_anonymous", ColumnContent, ColumnContent(COL_NAME_IS_ANONYMOUS, "bool", False, hidden=True)]
+        ["is_anonymous", ColumnContent, field(default_factory=lambda: ColumnContent(COL_NAME_IS_ANONYMOUS, "bool", False, hidden=True))]
     )
     return auto_eval_column_dict
 
@@ -76,7 +80,7 @@ def make_autoevalcolumn(cls_name, benchmarks):
     # Leaderboard columns
     for benchmark in list(benchmarks.value):
         auto_eval_column_dict.append(
-            [benchmark.name, ColumnContent, ColumnContent(benchmark.value.col_name, "number", True)]
+            [benchmark.name, ColumnContent, field(default_factory=lambda b=benchmark: ColumnContent(b.value.col_name, "number", True))]
         )
 
     # We use make dataclass to dynamically fill the scores from Tasks
@@ -85,14 +89,19 @@ def make_autoevalcolumn(cls_name, benchmarks):
 
 def get_default_col_names_and_types(benchmarks):
     AutoEvalColumn = make_autoevalcolumn("AutoEvalColumn", benchmarks)
-    col_names = [c.name for c in _fields(AutoEvalColumn) if not c.hidden]
-    col_types = [c.type for c in _fields(AutoEvalColumn) if not c.hidden]
+    col_names = []
+    col_types = []
+    for f in AutoEvalColumn.__dataclass_fields__.values():
+        col = f.default_factory()
+        if not col.hidden:
+            col_names.append(col.name)
+            col_types.append(col.type)
     return col_names, col_types
 
 
 def get_fixed_col_names_and_types():
     fixed_cols = get_default_auto_eval_column_dict()[:-3]
-    return [c.name for _, _, c in fixed_cols], [c.type for _, _, c in fixed_cols]
+    return [c.default_factory().name for _, _, c in fixed_cols], [c.default_factory().type for _, _, c in fixed_cols]
 
 
 COL_NAME_AVG = "Average ⬆️"