Update chess_ground-targz.py

chess_ground-targz.py

@@ -1,10 +1,11 @@
+import os
 import json
+import tarfile
 import datasets
-import os
 
 # Description of the dataset
 _DESCRIPTION = """\
-Dataset for extracting notations from chess scoresheets.
+Dataset for extracting notations from chess scoresheets, integrating both image and text data.
 """
 
 # BibTeX citation for the dataset
@@ -21,10 +22,10 @@ _LICENSE = "Creative Commons Attribution 3.0"
 
 
 class ChessImageTextDataset(datasets.GeneratorBasedBuilder):
-    """Dataset for linking chess scoresheets images and their ground truth text."""
+    """Dataset for linking chess scoresheet images with ground truth text."""
 
     def _info(self):
-        # Define the features of your dataset (image and text)
+        # Define the features of your dataset (images + text)
         features = datasets.Features(
             {
                 "image": datasets.Image(),  # Image feature for chess scoresheets
@@ -44,8 +45,8 @@ class ChessImageTextDataset(datasets.GeneratorBasedBuilder):
 
     def _split_generators(self, dl_manager):
         """Define the splits of the dataset."""
-        
-        # Load the image dataset
+
+        # Load the image dataset (tar.gz file)
         image_dataset_url = "https://huggingface.co/datasets/Chesscorner/chess-images/resolve/main/flat_images.tar.gz"
         extracted_image_path = dl_manager.download_and_extract(image_dataset_url)
 
@@ -57,36 +58,60 @@ class ChessImageTextDataset(datasets.GeneratorBasedBuilder):
             datasets.SplitGenerator(
                 name=datasets.Split.TRAIN,
                 gen_kwargs={
-                    "image_path": extracted_image_path,
-                    "text_filepath": text_filepath,
+                    "image_path": extracted_image_path,  # Path to extracted tar directory
+                    "text_filepath": text_filepath,  # Path to the text dataset
                 },
             ),
         ]
 
     def _generate_examples(self, image_path, text_filepath):
-        """Generate examples linking images to text."""
-
+        """Generate examples by linking images and text."""
         idx = 0
-        # Load the text dataset
+        
+        # Extract and map text IDs to their corresponding images
+        image_mapping = self._extract_images_from_tar(image_path)
+
+        # Load the text dataset (ground truths) from the JSONL file
         with open(text_filepath, encoding="utf-8") as fp:
             for line in fp:
                 obj = json.loads(line)
                 text = obj["text"]
                 
-                # Extract image ID from the text (assuming the first 5 characters of text match image filename)
-                text_id = text[:5]  # Adjust this logic based on the exact pattern you have
-
-                # Find the corresponding image
-                image_file = os.path.join(f"{text_id}.png", image_path)   
+                # Extract the text ID (assuming text ID matches image filename)
+                text_id = text[:5]  # Adjust this based on the actual pattern of text IDs
                 
-                # Check if the image file exists (you can add extra checks here)
-                if os.path.exists(image_file):
+                # Find the corresponding image file
+                image_file = image_mapping.get(f"{text_id}.png")  # Adjust file extension if necessary
+
+                # Ensure the image exists and yield the example
+                if image_file:
                     yield idx, {
                         "image": image_file,
                         "text": text,
                     }
                 else:
-                    print(f"Text not found for ID: {image_path}")
+                    print(f"Image not found for ID: {text_id}")
                 
                 idx += 1
-            
+
+    def _extract_images_from_tar(self, tar_path):
+        """Extracts the images from the tar.gz archive and returns a mapping of image filenames to file paths."""
+        
+        image_mapping = {}
+        
+        # Ensure extraction of individual files, not directories
+        with tarfile.open(tar_path, "r:gz") as tar:
+            tar.extractall(path="extracted_images")  # Extract images to a specific directory
+            
+            # Iterate over the extracted files and map them
+            for member in tar.getmembers():
+                if member.isfile():  # Only include actual files, not directories
+                    image_filename = os.path.basename(member.name)
+                    image_path = os.path.join("extracted_images", member.name)
+                    
+                    # Check if the path points to a valid image file
+                    if os.path.isfile(image_path):
+                        image_mapping[image_filename] = image_path
+        
+        return image_mapping
+