update billsum to 2.0.0

PiperOrigin-RevId: 282681798

Author: yaozhaogoogle

tensorflow_datasets/summarization/billsum.py

@@ -20,6 +20,7 @@
 from __future__ import print_function
 
 import json
+import os
 
 import tensorflow as tf
 import tensorflow_datasets.public_api as tfds
@@ -47,14 +48,7 @@
   - sum_len: number of chars in summary.
 """
 
-_URLS = {
-    "us_train":
-        "https://datahub.io/akornilo/billsum/r/us_train_data_final.jsonl",
-    "us_test":
-        "https://datahub.io/akornilo/billsum/r/us_test_data_final.jsonl",
-    "ca_test":
-        "https://datahub.io/akornilo/billsum/r/ca_test_data_final.jsonl",
-}
+_URL = "https://datahub.io/akornilo/billsum/r/billsum.zip"
 
 _DOCUMENT = "text"
 _SUMMARY = "summary"
@@ -63,7 +57,8 @@
 class Billsum(tfds.core.GeneratorBasedBuilder):
   """BillSum Dataset."""
 
-  VERSION = tfds.core.Version("1.0.0")
+  # data source updated to filter near duplicates.
+  VERSION = tfds.core.Version("2.0.0")
 
   def _info(self):
     return tfds.core.DatasetInfo(
@@ -81,26 +76,26 @@ def _info(self):
 
   def _split_generators(self, dl_manager):
     """Returns SplitGenerators."""
-    dl_paths = dl_manager.download(_URLS)
+    dl_path = dl_manager.download_and_extract(_URL)
     return [
         tfds.core.SplitGenerator(
             name=tfds.Split.TRAIN,
             gen_kwargs={
-                "path": dl_paths["us_train"],
+                "path": os.path.join(dl_path, "us_train_data_final_v2.jsonl"),
                 "key": "bill_id"
             },
         ),
         tfds.core.SplitGenerator(
             name=tfds.Split.TEST,
             gen_kwargs={
-                "path": dl_paths["us_test"],
+                "path": os.path.join(dl_path, "us_test_data_final_v2.jsonl"),
                 "key": "bill_id"
             },
         ),
         tfds.core.SplitGenerator(
             name="ca_test",
             gen_kwargs={
-                "path": dl_paths["ca_test"],
+                "path": os.path.join(dl_path, "ca_test_data_final.jsonl"),
                 "key": "external_id"
             },
         ),

tensorflow_datasets/summarization/billsum_test.py

@@ -30,12 +30,7 @@ class BillsumTest(testing.DatasetBuilderTestCase):
       "test": 1,  # Number of fake test example
       "ca_test": 1  # Number of fake test example
   }
-  DL_EXTRACT_RESULT = {
-      "us_train": "us_train.jsonl",
-      "us_test": "us_test.jsonl",
-      "ca_test": "ca_test.jsonl"
-  }
-
+  DL_EXTRACT_RESULT = ""
 
 if __name__ == "__main__":
   testing.test_main()

tensorflow_datasets/testing/test_data/fake_examples/billsum/ca_test.jsonl renamed to tensorflow_datasets/testing/test_data/fake_examples/billsum/ca_test_data_final.jsonl

@@ -1,3 +1 @@
-https://datahub.io/akornilo/billsum/r/ca_test_data_final.jsonl 15086790 4db74f4ffed4ba6692a14f8823e719dbc93b4aa79acd9fedf3825e7f656b73ed
-https://datahub.io/akornilo/billsum/r/us_test_data_final.jsonl 59068884 6285ff95d86d0c2f4e4b42954a7aae0709439bf3f448da66fb822ab681c69489
-https://datahub.io/akornilo/billsum/r/us_train_data_final.jsonl 335338742 baa27a45443541ff77f610f8353c75b121c66ae29517629eeee9154bdce7ab82
+https://datahub.io/akornilo/billsum/r/billsum.zip 68989684 f20c03fd95a0b457a39d31b23c7fa677466ce53b3cb9cf8ae4f256e074f01f13