Datasets:

ArXiv:
License:
dp-bench / src /utils.py
shinseung428's picture
bugfix and add remaining dataset
6c2a241
import json
from pathlib import Path
from typing import List
def read_file(path: str, supported_formats: str = ".json") -> dict:
"""Read a file and return its content as a string
Args:
path (str): the path to the file to read
supported_formats (str, optional): the supported file formats. Defaults to ".json".
Returns:
dict: the json content of the file
Raises:
FileNotFoundError: if the file does not exist
ValueError: if the file format is not supported
"""
path = Path(path)
# check if the file exists and is a file
if not path.exists() or not path.is_file():
raise FileNotFoundError(f"File {path} not found")
# check if the file format is supported
if path.suffix not in supported_formats:
raise ValueError(f"File format {path.suffix} not supported")
with path.open("r") as file:
file_content = json.load(file)
return file_content
def create_directory(path: str) -> None:
"""Create a directory if it does not exist
Args:
path (str): the path to the directory to create
"""
path = Path(path)
if not path.exists():
path.mkdir(parents=True, exist_ok=True)
def read_file_paths(path: str, supported_formats: List[str] = [".jpg"]) -> List[str]:
"""Read files in a directory and return their content as a list of strings
Args:
path (str): the path to the directory containing the file paths to read
supported_formats (List[str], optional): the supported file formats. Defaults to [".jpg"].
Returns:
list: list of valid file paths
Raises:
FileNotFoundError: if the directory does not exist
"""
path = Path(path)
# check if the directory exists and is a directory
if not path.exists() or not path.is_dir():
raise FileNotFoundError(f"Directory {path} not found")
# get the list of files in the directory
file_paths = [file for file in path.iterdir() if file.is_file()]
# filter file paths based on the supported formats
if supported_formats:
file_paths = [file for file in file_paths if file.suffix in supported_formats]
else:
file_paths = []
return file_paths
def check_dataset_format(data: dict, image_key: str) -> None:
"""Check the format of the dataset
Args:
data (dict): the gt/prediction dataset to check
image_key (str): the image name acting as the key in the dataset
Raises:
ValueError: if a key is missing in the dataset
"""
if data[image_key].get("elements") is None:
raise ValueError(
f"{image_key} does not have 'elements' key in the json file. "
"Check if you are passing the correct data."
)
elements = data[image_key]["elements"]
for elem in elements:
if elem.get("category") is None:
raise ValueError(
f"{image_key} does not have 'category' key in the ground truth file. "
"Check if you are passing the correct data."
)
if elem.get("content") is None:
raise ValueError(
f"{image_key} does not have 'content' key in the ground truth file. "
"Check if you are passing the correct data."
)
else:
content = elem["content"]
if content.get("text") is None:
raise ValueError(
f"{image_key} does not have 'text' key in the ground truth file. "
"Check if you are passing the correct data."
)
def check_data_validity(gt_data: dict, pred_data: dict) -> None:
"""Check the validity of the ground truth and prediction data
Args:
gt_data (dict): the ground truth data
pred_data (dict): the prediction data
Raises:
ValueError: if the ground truth or prediction data is invalid
"""
if not gt_data:
raise ValueError("Ground truth data is empty")
if not pred_data:
raise ValueError("Prediction data is empty")
for image_key in gt_data.keys():
pred_elem = pred_data.get(image_key)
if pred_data is None:
raise ValueError(
f"{image_key} not found in prediction. "
"Check if you are passing the correct data."
)
for image_key in gt_data.keys():
check_dataset_format(gt_data, image_key)
for image_key in pred_data.keys():
check_dataset_format(pred_data, image_key)