Create generate_csv.py

generate_csv.py

@@ -0,0 +1,19 @@
+import os
+import csv
+
+def generate_csv(split):
+    csv_path = f"dataset_{split}.csv"
+    folder = f"dataset/{split}"
+    with open(csv_path, "w", newline="", encoding="utf-8") as f:
+        writer = csv.writer(f)
+        writer.writerow(["img_path", "latex"])
+        for file in sorted(os.listdir(folder)):
+            if file.endswith(".txt"):
+                img_file = file.replace(".txt", ".png")
+                img_path = f"{split}/{img_file}"
+                with open(os.path.join(folder, file), "r", encoding="utf-8") as label_file:
+                    latex = label_file.read().strip()
+                    writer.writerow([img_path, latex])
+
+generate_csv("train")
+generate_csv("val")