|
| 1 | +#!/usr/bin/env python3 |
| 2 | +"""Script to retrieve and process NEMAR datasets.""" |
| 3 | + |
| 4 | +import logging |
| 5 | +import os |
| 6 | +import sys |
| 7 | +from typing import Dict, List, Optional |
| 8 | + |
| 9 | +import requests |
| 10 | +import urllib3 |
| 11 | + |
| 12 | +# Add the project root to the Python path |
| 13 | +sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__)))) |
| 14 | + |
| 15 | +import eegdash.dataset |
| 16 | + |
| 17 | +# Disable SSL warnings since we're using verify=False |
| 18 | +urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning) |
| 19 | + |
| 20 | +# Configure logging |
| 21 | +logging.basicConfig(level=logging.INFO, format="%(message)s") |
| 22 | +logger = logging.getLogger(__name__) |
| 23 | + |
| 24 | + |
| 25 | +class NemarAPI: |
| 26 | + """Client for interacting with the NEMAR API.""" |
| 27 | + |
| 28 | + def __init__(self, token: Optional[str] = None): |
| 29 | + """Initialize NEMAR API client. |
| 30 | +
|
| 31 | + Args: |
| 32 | + token: NEMAR access token. If not provided, will look for NEMAR_TOKEN env variable. |
| 33 | +
|
| 34 | + Raises: |
| 35 | + ValueError: If no token is provided or found in environment. |
| 36 | +
|
| 37 | + """ |
| 38 | + self.base_url = "https://nemar.org/api/dataexplorer/datapipeline" |
| 39 | + self.token = token or os.environ.get("NEMAR_TOKEN") |
| 40 | + if not self.token: |
| 41 | + raise ValueError( |
| 42 | + "NEMAR token must be provided either as argument or NEMAR_TOKEN environment variable" |
| 43 | + ) |
| 44 | + |
| 45 | + def get_datasets(self, start: int = 0, limit: int = 500) -> Optional[Dict]: |
| 46 | + """Get list of datasets from NEMAR. |
| 47 | +
|
| 48 | + Args: |
| 49 | + start: Starting index for pagination. |
| 50 | + limit: Maximum number of datasets to return. |
| 51 | +
|
| 52 | + Returns: |
| 53 | + JSON response containing dataset information or None if request fails. |
| 54 | +
|
| 55 | + """ |
| 56 | + payload = { |
| 57 | + "nemar_access_token": self.token, |
| 58 | + "table_name": "dataexplorer_dataset", |
| 59 | + "start": start, |
| 60 | + "limit": limit, |
| 61 | + } |
| 62 | + |
| 63 | + try: |
| 64 | + response = requests.post( |
| 65 | + f"{self.base_url}/list", |
| 66 | + headers={"Content-Type": "application/json"}, |
| 67 | + json=payload, |
| 68 | + verify=False, |
| 69 | + ) |
| 70 | + response.raise_for_status() |
| 71 | + return response.json() |
| 72 | + except requests.exceptions.RequestException as e: |
| 73 | + logger.error("Error fetching datasets: %s", e) |
| 74 | + return None |
| 75 | + |
| 76 | + @staticmethod |
| 77 | + def extract_dataset_info(datasets_response: Dict) -> List[Dict]: |
| 78 | + """Extract relevant information from datasets response. |
| 79 | +
|
| 80 | + Args: |
| 81 | + datasets_response: Response from get_datasets(). |
| 82 | +
|
| 83 | + Returns: |
| 84 | + List of dictionaries containing dataset information. |
| 85 | +
|
| 86 | + """ |
| 87 | + if not datasets_response or "entries" not in datasets_response: |
| 88 | + return [] |
| 89 | + |
| 90 | + return [ |
| 91 | + { |
| 92 | + "id": data["id"], |
| 93 | + "name": data["name"], |
| 94 | + "modalities": data["modalities"], |
| 95 | + "participants": data["participants"], |
| 96 | + "file_size": data["file_size"], |
| 97 | + "file_size_gb": float(data["file_size"]) / (1024 * 1024 * 1024), |
| 98 | + "tasks": data.get("tasks", ""), |
| 99 | + "authors": data.get("Authors", ""), |
| 100 | + "doi": data.get("DatasetDOI", ""), |
| 101 | + } |
| 102 | + for _, data in datasets_response["entries"].items() |
| 103 | + ] |
| 104 | + |
| 105 | + |
| 106 | +def fetch_all_datasets() -> List[Dict]: |
| 107 | + """Fetch all available datasets from NEMAR. |
| 108 | +
|
| 109 | + Returns: |
| 110 | + List of dataset information dictionaries. |
| 111 | +
|
| 112 | + """ |
| 113 | + try: |
| 114 | + nemar = NemarAPI() |
| 115 | + except ValueError as e: |
| 116 | + logger.error("Error: %s", e) |
| 117 | + logger.error( |
| 118 | + "Please set your NEMAR token using: export NEMAR_TOKEN='your_token_here'" |
| 119 | + ) |
| 120 | + return [] |
| 121 | + |
| 122 | + all_datasets = [] |
| 123 | + start = 0 |
| 124 | + batch_size = 500 |
| 125 | + |
| 126 | + logger.info("Fetching datasets...") |
| 127 | + while True: |
| 128 | + datasets = nemar.get_datasets(start=start, limit=batch_size) |
| 129 | + if not datasets or not datasets.get("entries"): |
| 130 | + break |
| 131 | + |
| 132 | + batch_info = nemar.extract_dataset_info(datasets) |
| 133 | + if not batch_info: |
| 134 | + break |
| 135 | + |
| 136 | + all_datasets.extend(batch_info) |
| 137 | + logger.info("Retrieved %d datasets so far...", len(all_datasets)) |
| 138 | + |
| 139 | + if len(batch_info) < batch_size: |
| 140 | + break |
| 141 | + |
| 142 | + start += batch_size |
| 143 | + |
| 144 | + return all_datasets |
| 145 | + |
| 146 | + |
| 147 | +def find_undigested_datasets() -> List[Dict]: |
| 148 | + """Find datasets that haven't been digested into eegdash yet. |
| 149 | +
|
| 150 | + Returns: |
| 151 | + List of dataset information dictionaries for undigested datasets. |
| 152 | +
|
| 153 | + """ |
| 154 | + # Get all available datasets from NEMAR |
| 155 | + all_datasets = fetch_all_datasets() |
| 156 | + |
| 157 | + # Get all classes from eegdash.dataset |
| 158 | + eegdash_classes = dir(eegdash.dataset) |
| 159 | + |
| 160 | + # Filter for undigested datasets |
| 161 | + undigested = [] |
| 162 | + for dataset in all_datasets: |
| 163 | + # Convert dataset ID to expected class name format (e.g., ds001785 -> DS001785) |
| 164 | + class_name = dataset["id"].upper() |
| 165 | + |
| 166 | + # Check if this dataset exists as a class in eegdash.dataset |
| 167 | + if class_name not in eegdash_classes: |
| 168 | + undigested.append(dataset) |
| 169 | + |
| 170 | + return undigested |
| 171 | + |
| 172 | + |
| 173 | +def main(): |
| 174 | + """Main function to find and output undigested datasets.""" |
| 175 | + undigested = find_undigested_datasets() |
| 176 | + |
| 177 | + # Print just the dataset IDs and names |
| 178 | + print("\nUndigested Datasets:") |
| 179 | + print("-" * 80) |
| 180 | + for dataset in undigested: |
| 181 | + print(f"{dataset['id']}: {dataset['name']}") |
| 182 | + print(f"\nTotal undigested datasets: {len(undigested)}") |
| 183 | + |
| 184 | + |
| 185 | +if __name__ == "__main__": |
| 186 | + main() |
0 commit comments