1 files changed, 28 insertions, 0 deletions
diff --git a/dataset_utils.py b/dataset_utils.py
new file mode 100644
index 0000000..865d5be
--- /dev/null
+++ b/dataset_utils.py
@@ -0,0 +1,28 @@
+import numpy
+from PIL import Image
+from pathlib import Path
+
+"""Class to interface the training and testing data."""
+class DatasetUtils:
+    def __init__(self) -> None:
+        self.data_path = Path('./data')
+        # self.current_directory = 'I'
+
+    """Convert the dataset to a 2 dimension array."""
+    def data_to_arrays(self):
+        for dir in self.data_path.iterdir():
+            if not dir.is_dir(): continue
+
+            for file in dir.glob('/*.png'):
+                image = Image.open(str(file)).convert(mode='L')
+                # Return the image's pixel values as an array alongside
+                # the character that it represents.
+                yield (dir.name, numpy.asarray(image))
+
+    """Search for a file in the dataset."""
+    def search(self, filename: str):
+        dataset = [image for image in self.data_path.rglob('*.png')]
+        
+
+    def get_random_sample(self):
+        pass