Automated Code Change

PiperOrigin-RevId: 944183338

Author: h-joo

tensorflow_datasets/datasets/patch_camelyon/patch_camelyon_dataset_builder.py

@@ -54,13 +54,13 @@ def _split_generators(self, dl_manager):
     paths = dl_manager.download_and_extract(resources)
     return [
         tfds.core.SplitGenerator(
-            name=tfds.Split.TEST, gen_kwargs=dict(split='test', paths=paths)
+            name=tfds.Split.TEST, gen_kwargs=dict(split='test', paths=paths)  # pyrefly: ignore[missing-attribute]
         ),
         tfds.core.SplitGenerator(
-            name=tfds.Split.TRAIN, gen_kwargs=dict(split='train', paths=paths)
+            name=tfds.Split.TRAIN, gen_kwargs=dict(split='train', paths=paths)  # pyrefly: ignore[missing-attribute]
         ),
         tfds.core.SplitGenerator(
-            name=tfds.Split.VALIDATION,
+            name=tfds.Split.VALIDATION,  # pyrefly: ignore[missing-attribute]
             gen_kwargs=dict(split='valid', paths=paths),
         ),
     ]

tensorflow_datasets/datasets/paws_wiki/paws_wiki_dataset_builder.py

@@ -44,9 +44,9 @@
     _UNLABELED_FINAL: ["train", "dev"],
 }
 _SPLIT_MAPPINGS = {
-    "train": tfds.Split.TRAIN,
-    "dev": tfds.Split.VALIDATION,
-    "test": tfds.Split.TEST,
+    "train": tfds.Split.TRAIN,  # pyrefly: ignore[missing-attribute]
+    "dev": tfds.Split.VALIDATION,  # pyrefly: ignore[missing-attribute]
+    "test": tfds.Split.TEST,  # pyrefly: ignore[missing-attribute]
 }
 
 _CLASS_LABELS = ["different_meaning", "paraphrase"]
@@ -101,10 +101,10 @@ def _info(self):
   def _split_generators(self, dl_manager):
     """Returns SplitGenerators."""
     dl_paths = dl_manager.download_and_extract(_DOWNLOAD_URLS)
-    subset = self.builder_config.subset
+    subset = self.builder_config.subset  # pyrefly: ignore[missing-attribute]
     labels_path = os.path.join(dl_paths[subset], _EXTRACTED_FOLDERS[subset])
 
-    if self.builder_config.tokenized:
+    if self.builder_config.tokenized:  # pyrefly: ignore[missing-attribute]
       mappings_path = ""
       tags2texts = {}
     else:
@@ -193,7 +193,7 @@ def _generate_examples(
           key = row["id"]
           label_str = (
               "noisy_label"
-              if self.builder_config.subset == _UNLABELED_FINAL
+              if self.builder_config.subset == _UNLABELED_FINAL  # pyrefly: ignore[missing-attribute]
               else "label"
           )
           example = {

tensorflow_datasets/datasets/paws_x_wiki/paws_x_wiki_dataset_builder.py

@@ -75,32 +75,32 @@ def _split_generators(self, dl_manager):
     base_path = os.path.join(dl_path["x-final"], "x-final")
 
     # Name of file for training for 'en' is different from other languages
-    if self.builder_config.language == "en":
+    if self.builder_config.language == "en":  # pyrefly: ignore[missing-attribute]
       training_path = os.path.join(
           base_path, self.builder_config.language, "train.tsv"
       )
     else:
-      training_path = os.path.join(
+      training_path = os.path.join(  # pyrefly: ignore[no-matching-overload]
           base_path, self.builder_config.language, "translated_train.tsv"
       )
 
     return [
         tfds.core.SplitGenerator(
-            name=tfds.Split.TRAIN,
+            name=tfds.Split.TRAIN,  # pyrefly: ignore[missing-attribute]
             gen_kwargs={"path": training_path},
         ),
         tfds.core.SplitGenerator(
-            name=tfds.Split.TEST,
+            name=tfds.Split.TEST,  # pyrefly: ignore[missing-attribute]
             gen_kwargs={
-                "path": os.path.join(
+                "path": os.path.join(  # pyrefly: ignore[no-matching-overload]
                     base_path, self.builder_config.language, "test_2k.tsv"
                 )
             },
         ),
         tfds.core.SplitGenerator(
-            name=tfds.Split.VALIDATION,
+            name=tfds.Split.VALIDATION,  # pyrefly: ignore[missing-attribute]
             gen_kwargs={
-                "path": os.path.join(
+                "path": os.path.join(  # pyrefly: ignore[no-matching-overload]
                     base_path, self.builder_config.language, "dev_2k.tsv"
                 )
             },

tensorflow_datasets/datasets/penguins/penguins_dataset_builder.py

@@ -159,11 +159,11 @@ class Builder(tfds.core.GeneratorBasedBuilder):
   def _info(self) -> tfds.core.DatasetInfo:
     """Returns the dataset metadata."""
     supervised_keys = None
-    features = self.builder_config.features
+    features = self.builder_config.features  # pyrefly: ignore[missing-attribute]
     supervised_features = features.copy()
-    label_name = self.builder_config.label
+    label_name = self.builder_config.label  # pyrefly: ignore[missing-attribute]
 
-    if self.builder_config.name == 'processed':
+    if self.builder_config.name == 'processed':  # pyrefly: ignore[missing-attribute]
       label_feature = supervised_features.pop(label_name, None)
       supervised_keys = ('features', label_name)
       features = {
@@ -187,7 +187,7 @@ def _info(self) -> tfds.core.DatasetInfo:
 
   def _split_generators(self, dl_manager: tfds.download.DownloadManager):
     """Returns SplitGenerators."""
-    path = dl_manager.download(_PENGUINS_PATH + self.builder_config.file_name)
+    path = dl_manager.download(_PENGUINS_PATH + self.builder_config.file_name)  # pyrefly: ignore[missing-attribute]
     return {'train': self._generate_examples(path)}
 
   def _generate_examples(self, path):
@@ -197,8 +197,8 @@ def _generate_examples(self, path):
         row = {f: self._clean_up(f, v) for f, v in row.items()}
 
         # Pack features if requested.
-        if self.builder_config.name == 'processed':
-          label_name = self.builder_config.label
+        if self.builder_config.name == 'processed':  # pyrefly: ignore[missing-attribute]
+          label_name = self.builder_config.label  # pyrefly: ignore[missing-attribute]
           label = row.pop(label_name, None)
           row = list(row.values())
           yield i, {'features': row, label_name: label}
@@ -207,21 +207,21 @@ def _generate_examples(self, path):
 
   def _clean_up(self, field, value):
     """Applies field-level pre-processing, if needed."""
-    if not self.builder_config.cleanup:
+    if not self.builder_config.cleanup:  # pyrefly: ignore[missing-attribute]
       return value
     if field not in self.builder_config.cleanup:
       return value
 
-    feature_type = self.builder_config.features[field]
+    feature_type = self.builder_config.features[field]  # pyrefly: ignore[missing-attribute]
     if feature_type == np.float32:
       # Field is a float. If it won't parse, clean it up.
       try:
         return float(value)
       except ValueError:
-        return self.builder_config.cleanup[field]
+        return self.builder_config.cleanup[field]  # pyrefly: ignore[bad-index]
     elif isinstance(feature_type, tfds.features.ClassLabel):
       # Field is a class. If it's OOV, clean it up.
       if value not in feature_type.names:
-        return self.builder_config.cleanup[field]
+        return self.builder_config.cleanup[field]  # pyrefly: ignore[bad-index]
 
     return value

tensorflow_datasets/datasets/pet_finder/pet_finder_dataset_builder.py

@@ -100,15 +100,15 @@ def _split_generators(self, dl_manager):
 
     return [
         tfds.core.SplitGenerator(
-            name=tfds.Split.TRAIN,
+            name=tfds.Split.TRAIN,  # pyrefly: ignore[missing-attribute]
             gen_kwargs={
                 "csv_name": "train.csv",
                 "csv_paths": dl_paths["train"],
                 "img_paths": dl_paths["train_images"],
             },
         ),
         tfds.core.SplitGenerator(
-            name=tfds.Split.TEST,
+            name=tfds.Split.TEST,  # pyrefly: ignore[missing-attribute]
             gen_kwargs={
                 "csv_name": "test.csv",
                 "csv_paths": dl_paths["test"],

tensorflow_datasets/datasets/pg19/pg19_dataset_builder.py

@@ -59,21 +59,21 @@ def _split_generators(self, dl_manager):
 
     return [
         tfds.core.SplitGenerator(
-            name=tfds.Split.TRAIN,
+            name=tfds.Split.TRAIN,  # pyrefly: ignore[missing-attribute]
             gen_kwargs={
                 'metadata': metadata_dict,
                 'filepath': os.path.join(_DATA_DIR, 'train'),
             },
         ),
         tfds.core.SplitGenerator(
-            name=tfds.Split.VALIDATION,
+            name=tfds.Split.VALIDATION,  # pyrefly: ignore[missing-attribute]
             gen_kwargs={
                 'metadata': metadata_dict,
                 'filepath': os.path.join(_DATA_DIR, 'validation'),
             },
         ),
         tfds.core.SplitGenerator(
-            name=tfds.Split.TEST,
+            name=tfds.Split.TEST,  # pyrefly: ignore[missing-attribute]
             gen_kwargs={
                 'metadata': metadata_dict,
                 'filepath': os.path.join(_DATA_DIR, 'test'),

tensorflow_datasets/datasets/piqa/piqa_dataset_builder.py

@@ -58,7 +58,7 @@ def _split_generators(self, dl_manager):
     # Specify the splits
     return [
         tfds.core.SplitGenerator(
-            name=tfds.Split.TRAIN,
+            name=tfds.Split.TRAIN,  # pyrefly: ignore[missing-attribute]
             gen_kwargs={
                 "data_path": os.path.join(
                     extracted_path, "physicaliqa-train-dev/train.jsonl"
@@ -69,7 +69,7 @@ def _split_generators(self, dl_manager):
             },
         ),
         tfds.core.SplitGenerator(
-            name=tfds.Split.VALIDATION,
+            name=tfds.Split.VALIDATION,  # pyrefly: ignore[missing-attribute]
             gen_kwargs={
                 "data_path": os.path.join(
                     extracted_path, "physicaliqa-train-dev/dev.jsonl"

tensorflow_datasets/datasets/places365_small/places365_small_dataset_builder.py

@@ -18,7 +18,7 @@
 import os
 
 from etils import epath
-import six.moves.urllib as urllib
+import six.moves.urllib as urllib  # pyrefly: ignore[missing-source-for-stubs]
 import tensorflow_datasets.public_api as tfds
 
 _BASE_URL = "http://data.csail.mit.edu/places/places365/"

tensorflow_datasets/datasets/plant_leaves/plant_leaves_dataset_builder.py

@@ -80,7 +80,7 @@ def _split_generators(self, dl_manager):
     archive_path = dl_manager.download(_DOWNLOAD_URL)
     return [
         tfds.core.SplitGenerator(
-            name=tfds.Split.TRAIN,
+            name=tfds.Split.TRAIN,  # pyrefly: ignore[missing-attribute]
             gen_kwargs={"archive": dl_manager.iter_archive(archive_path)},
         )
     ]

tensorflow_datasets/datasets/plant_village/plant_village_dataset_builder.py

@@ -84,7 +84,7 @@ def _split_generators(self, dl_manager):
     path = dl_manager.download_and_extract(_URL)
     return [
         tfds.core.SplitGenerator(
-            name=tfds.Split.TRAIN, gen_kwargs={"datapath": path}
+            name=tfds.Split.TRAIN, gen_kwargs={"datapath": path}  # pyrefly: ignore[missing-attribute]
         )
     ]