Python script to get specific metadata from GBIF-registered datasets.

get_gbif_datasets.py

# -*- coding: utf-8 -*-
import requests
import json
import csv
import sys

def get_datasets(offset, limit):
    sys.stderr.write('Requesting batch {}\n'.format(offset)) # Use stderr to not contaminate stdout, which is used for the results
    request = requests.get('http://api.gbif.org/v1/dataset/', params={'limit': limit, 'offset': offset})
    results = request.json()['results']
    return results

def parse_metadata(dataset, fields):
    metadata = []
    for field in fields:
        try:
            metadata.append(dataset[field].encode('utf-8'))
        except KeyError, e:
            metadata.append('')
    return metadata

def main(limit=20):
    still_more_datasets = True
    offset = 0
    csvwriter = csv.writer(sys.stdout, lineterminator='\n')

    # Get fields to keep
    fields_to_keep = sys.argv[1:] # Remove first element = script itself

    # Add headers
    csvwriter.writerow(fields_to_keep)

    # Request metadata
    while still_more_datasets:
        next_batch = get_datasets(offset, limit)
        if len(next_batch) == 0:
            # Nothing returned, stop requesting
            still_more_datasets = False
        else:
            for dataset in next_batch:
                metadata = parse_metadata(dataset, fields_to_keep)
                csvwriter.writerow(metadata)
            offset += limit

if __name__ == '__main__': # If run from command line
    sys.exit(main())

Usage: python get_gbif_datasets.py key title type rights > datasets.csv