| # pandas: For data manipulation and transformation. | |
| # pyarrow: For reading and writing Parquet files. | |
| import pyarrow.parquet as pq | |
| import os | |
| import json | |
| table = pq.read_table('data/test-00000-of-00001.parquet') | |
| df = table.to_pandas() # DataFrame | |
| print("INICIO", df) | |
| with open("images/images.json") as json_file: | |
| data = json.load(json_file) | |
| image_rows = data["imagesToUpload"] | |
| for img in image_rows: | |
| image_path = f"images/{img["name"]}" | |
| image_text = img["text"] | |
| with open(image_path, "rb") as image: | |
| image_bytes = image.read() | |
| new_row = {'image': {'bytes': image_bytes, 'path': image_path}, 'text': image_text} | |
| df = df._append(new_row, ignore_index=True) | |
| os.remove(image_path) | |
| print("FIN", df) | |
| df.to_parquet('data/test-00000-of-00001.parquet') |