A tool for pretty printing bson string (e.g., output of mongoexport)

bson_pprint.py

from __future__ import print_function
from __future__ import unicode_literals
from __future__ import division

from bson import json_util, ObjectId
from bson.py3compat import string_types
from bson.dbref import DBRef

DEFAULT_INDENT = 4

def format(bson, indent=DEFAULT_INDENT):
    d = _dbref_fix(json_util.loads(l))
    return json_util.dumps(d, indent=indent)

def parse(infile, outfile, indent=DEFAULT_INDENT):
    for l in infile:
        print(format(l, indent=indent), file=outfile)

def _dbref_fix(obj):
    """
    Recursive helper method that converts BSON ObjectId types so they can be
    converted by bson.json_util.dumps into json.

    Created to deal with a situation like this:
    "collection" : { "$ref" : "collection", "$id" : { "$oid" : "505cdd127e89297d6f000421" } }
    """
    if hasattr(obj, 'iteritems') or hasattr(obj, 'items'):  # PY3 support
        return {k: _dbref_fix(v) for k, v in obj.iteritems()}
    elif hasattr(obj, '__iter__') and not isinstance(obj, string_types):
        return [_dbref_fix(v) for v in obj]
    elif isinstance(obj, DBRef):
        id = obj.id
        coll = obj.collection
        db = obj.database
        d = {}
        if id: d['$id'] = {'$oid': str(id)}
        if coll: d['$ref'] = coll
        if db: d['$db'] = db
        return d
    else:
        return obj


if __name__ == '__main__':
    import sys
    import argparse
    parser = argparse.ArgumentParser(description='Pretty print bson dumps.')
    parser.add_argument('-i', '--indent', type=int, default=DEFAULT_INDENT, help='Number of spaces to indent')
    parser.add_argument('infile', nargs='?', type=argparse.FileType('r'),
                        default=sys.stdin)
    parser.add_argument('outfile', nargs='?', type=argparse.FileType('w'),
                        default=sys.stdout)
    args = parser.parse_args()
    parse(args.infile, args.outfile, args.indent)