GAE MapReduce Datastore to Cloud Storage

Datastore to Cloud Storage MapReduce

import webapp2
from google.appengine.ext import ndb

from mapreduce import mapreduce_pipeline
from mapreduce import base_handler

PROJECT_ID = 'your-project-name' 
GS_BUCKET = 'your-project-bucket-name'
ENTITY_KIND = 'your.Entity'
PIPELINE_NAME = 'some-name'
NUM_SHARDS = 16

def datastore_map(entity):
  yield("%s\n" % 'format-your-entity-here')

class DataBackupPipeline(base_handler.PipelineBase):
    
  def run(self):
    output = yield mapreduce_pipeline.MapperPipeline(
      PIPELINE_NAME, 
      "backup.datastore_map", 
      "mapreduce.input_readers.DatastoreInputReader", 
      output_writer_spec="mapreduce.output_writers.GoogleCloudStorageConsistentOutputWriter",
      params={
        "input_reader":{
          "entity_kind": ENTITY_KIND
        },
        "output_writer":{
          "filesystem": "gs",
          "bucket_name": GS_BUCKET,
          "output_sharding":"input",
        }
      },
      shards=NUM_SHARDS)
      yield CloudStorageWriter(output)

class ResultFile(ndb.Model):
  file_name = ndb.StringProperty()
  date = ndb.DateTimeProperty(auto_now_add=True)
  
class CloudStorageWriter(base_handler.PipelineBase):
    
    def run(self, csv_output):
      # Store all the file names
      files = [str(f.replace('/gs/', 'gs://')) for f in csv_output]
      for f in files:
        entry = ResultFile(file_name=f)
        entry.put()

class BackupHandler(webapp2.RequestHandler):
    
    def get(self):
      # Start the MR job and redirect to the status page
      pipeline = DataBackupPipeline()
      pipeline.start()
      path = pipeline.base_path + "/status?root=" + pipeline.pipeline_id
      self.redirect(path)

I am trying to use my ndb datastore, but getting model is not JSON serializable error. How to resolve it

The yield statement on line 35 appears to have broken indentation -- no?