Update dataset loading script

big_patent.py

@@ -15,7 +15,6 @@
 
 # Lint as: python3
 """BigPatent Dataset."""
-import glob
 import gzip
 import json
 import os
@@ -53,12 +52,8 @@ There are two features:
 
 _LICENSE = "Creative Commons Attribution 4.0 International"
 
-_REPO = "https://huggingface.co/datasets/big_patent/resolve/main/data"
-_URLS = {
-    "train": f"{_REPO}/train.zip",
-    "validation": f"{_REPO}/val.zip",
-    "test": f"{_REPO}/test.zip",
-}
+_SPLIT_NAMES = {datasets.Split.TRAIN: "train", datasets.Split.VALIDATION: "val", datasets.Split.TEST: "test"}
+_URL = "data/{version}/{split_name}.zip"
 
 _DOCUMENT = "description"
 _SUMMARY = "abstract"
@@ -78,38 +73,44 @@ _CPC_DESCRIPTION = {
 # Available versions:
 # 1.0.0 lower cased tokenized words.
 # 2.0.0 cased raw strings.
-# 2.1.0 cased raw strings (fixed).
-# TODO Add raw string versions
+# 2.1.2 cased raw strings (fixed).
 
-_VERSION = "1.0.0"
+_VERSION = "2.1.2"
 
 
 class BigPatentConfig(datasets.BuilderConfig):
     """BuilderConfig for BigPatent."""
 
-    def __init__(self, *args, cpc_codes=None, **kwargs):
+    def __init__(self, codes="all", version=_VERSION, **kwargs):
         """BuilderConfig for BigPatent.
         Args:
-            cpc_codes: str, cpc_codes
+            codes (str or list, default 'all'): CPC codes. Either 'all' or a combination
+                of {'a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'y'}.
             **kwargs: keyword arguments forwarded to super.
         """
-        super().__init__(*args, version=_VERSION, **kwargs)
-        self.cpc_codes = cpc_codes
+        if isinstance(codes, str):
+            codes = [codes]
+        name = "+".join(codes)
+        if name == "all":
+            codes = list(_CPC_DESCRIPTION)
+        if version != _VERSION:
+            name = f"{name}-{version}"
+        super().__init__(name=name, version=version, **kwargs)
+        self.codes = codes
 
 
 class BigPatent(datasets.GeneratorBasedBuilder):
     """BigPatent datasets."""
 
+    BUILDER_CONFIG_CLASS = BigPatentConfig
     BUILDER_CONFIGS = [
         BigPatentConfig(
-            cpc_codes=list(_CPC_DESCRIPTION),
-            name="all",
+            codes="all",
             description="Patents under all categories.",
         ),
     ] + [
-        BigPatentConfig(  # pylint:disable=g-complex-comprehension
-            cpc_codes=[k],
-            name=k,
+        BigPatentConfig(
+            codes=k,
             description=f"Patents under Cooperative Patent Classification (CPC) {k}: {v}",
         )
         for k, v in sorted(_CPC_DESCRIPTION.items())
@@ -129,22 +130,30 @@ class BigPatent(datasets.GeneratorBasedBuilder):
 
     def _split_generators(self, dl_manager):
         """Returns SplitGenerators."""
-        dl_paths = dl_manager.download_and_extract(_URLS)
-        split_dirs = {datasets.Split.TRAIN: "train", datasets.Split.VALIDATION: "val", datasets.Split.TEST: "test"}
+        urls = {
+            split: _URL.format(version=self.config.version, split_name=split_name)
+            for split, split_name in _SPLIT_NAMES.items()
+        }
+        dl_paths = dl_manager.download_and_extract(urls)
+        paths = {
+            split: [
+                dl_manager.iter_files(os.path.join(dl_paths[split], split_name, code)) for code in self.config.codes
+            ]
+            for split, split_name in _SPLIT_NAMES.items()
+        }
         return [
             datasets.SplitGenerator(
                 name=split,
-                gen_kwargs={"path": dl_paths[split], "split_dir": split_dirs[split]},
+                gen_kwargs={"paths": paths[split]},
             )
-            for split in split_dirs
+            for split in _SPLIT_NAMES
         ]
 
-    def _generate_examples(self, path=None, split_dir=None):
+    def _generate_examples(self, paths=None):
         """Yields examples."""
-        for cpc_code in self.config.cpc_codes:
-            filenames = glob.glob(os.path.join(path, split_dir, cpc_code, "*"))
-            for filename in sorted(filenames):
-                with open(filename, "rb") as fin:
+        for paths_per_code in paths:
+            for path in paths_per_code:
+                with open(path, "rb") as fin:
                     fin = gzip.GzipFile(fileobj=fin)
                     for row in fin:
                         json_obj = json.loads(row)