Refactor: update README with dataset configuration and remove unused files

README.md

@@ -38,6 +38,21 @@ source_datasets:
 task_categories:
   - other
 pretty_name: EU Law Dataset - Category 15.10
+
+configs:
+- config_name: EU_Law
+  data_files:
+    - split: ground_truth
+      path: "subsets/ground_truth/*"
+    - split: json
+      path: "subsets/json/*"
+    - split: pymupdf
+      path: "subsets/pymupdf/*"
+    - split: unstructured
+      path: "subsets/unstructured/*"
+    - split: nougat
+      path: "subsets/nougat/*"
+
 ---
 
 # EU Law Dataset – Category 15.10: Environment

__init__.py

@@ -1 +0,0 @@
-from .eu_law import EULawDataset

eu_Law.py

@@ -1,56 +0,0 @@
-import os
-import csv
-import datasets
-import sys
-_DESCRIPTION = "EU Law dataset with multiple subsets, each corresponding to a different document conversion method."
-
-
-# Configuration for each subset (conversion format)
-class EuLawConfig(datasets.BuilderConfig):
-    def __init__(self, **kwargs):
-        super().__init__(**kwargs)
-
-# Dataset class
-class EuLaw(datasets.GeneratorBasedBuilder):
-    BUILDER_CONFIGS = [
-        EuLawConfig(name="json", version=datasets.Version("1.0.0"), description="Subset in JSON format"),
-        EuLawConfig(name="pymupdf", version=datasets.Version("1.0.0"), description="Subset from PyMuPDF"),
-        EuLawConfig(name="unstructured", version=datasets.Version("1.0.0"), description="Unstructured output"),
-        EuLawConfig(name="nougat", version=datasets.Version("1.0.0"), description="Subset using Nougat .mmd"),
-        EuLawConfig(name="ground_truth", version=datasets.Version("1.0.0"), description="Ground truth data"),
-    ]
-
-    def _info(self):
-        return datasets.DatasetInfo(
-            description=_DESCRIPTION,
-            features=datasets.Features({
-                "pdf_path": datasets.Value("string"),
-                "conversion": datasets.Value("string"),
-            }),
-            supervised_keys=None,
-            homepage="https://huggingface.co/datasets/nargesbh/eu_Law",
-        )
-
-    def _split_generators(self, dl_manager):
-        subset = self.config.name
-        csv_path = f"subsets/{subset}/conversion_{subset}.csv"
-        downloaded_path = dl_manager.download(csv_path)
-
-        return [
-            datasets.SplitGenerator(
-                name=datasets.Split.TRAIN,
-                gen_kwargs={"filepath": downloaded_path}
-            )
-        ]
-
-    def _generate_examples(self, filepath):
-        csv.field_size_limit(sys.maxsize)
-        with open(filepath, encoding="utf-8") as f:
-            reader = csv.DictReader(f)
-            for idx, row in enumerate(reader):
-                if "pdf_path" not in row or "conversion" not in row:
-                    continue
-                yield idx, {
-                    "pdf_path": row["pdf_path"].strip(),
-                    "conversion": row["conversion"].strip()
-                }