Split large SharpHound datasets (JSON files) into smaller files that can more easily be imported into BloodHound. Especially useful due to the Electron memory limitations and BloodHound CE (running in browser) 1GB file size limitation.

bh_split2.py

#!/usr/bin/python3
# Based on https://gist.github.com/deltronzero/7c23bacf97b4b61c7a2f2950ef6f35d8
# updated to current BloodHound JSON format (compatible with SharpHound 2.x and BloodHound CE)
# pip install simplejson
import simplejson
import sys

def splitfile(file_name, object_limit):
    print(f"[*] Loading {file_name}")
    with open(file_name) as f:
        data = simplejson.load(f)
        total_objects = data['meta']['count']
        #file_type = data['meta']['type']
        file_type = 'data'
        print(f"Total Objects: {total_objects}")
        object_count = 0
        file_count = 0

        while object_count < total_objects:
            a = {}
            a[file_type] = data[file_type][object_count:][:object_limit]
            object_count += len(a[file_type])

            a['meta'] = data['meta']
            a['meta']['count'] = object_count

            f_split = file_name.split("\\")[-1].split(".")
            file_name_out = f"{f_split[0]}_{file_count}.{f_split[1]}"

            print(f"[*] Writing {file_name_out} - {object_count} of {total_objects}")
            f_out = open(file_name_out, "w")
            simplejson.dump(a, f_out)
            f_out.close()
            file_count += 1

def main():
    if len(sys.argv) < 2:
        print(sys.argv[0] + " filename.json [object_limit]")
        return

    if len(sys.argv) < 3:
        splitfile(sys.argv[1], 20000)
    else:
        splitfile(sys.argv[1], int(sys.argv[2]))

if __name__ == "__main__":
    main()