def main():
  '''
  Program wrapper.

  '''
  print('%s Creating Dataverse instance.' % item('bullet'))
  d = Dataverse('dataverse.harvard.edu', 'IFPRI')

  print('%s Collecting all content from Dataverse.' % item('bullet'))
  contents = d.contents()

  #
  #  Collects data and organizes
  #  in lists and dictionaries.
  #
  datasets = []
  resources = []
  for dataset in contents:
    print('%s Collecting data for ID %s.' % (item('bullet'), dataset['id']))

    o = Dataset(dataset['id']).info()
    if o.get('status') is 'ERROR' or None:
      continue

    try:
      parsed_data = parse_dataset(o)
    except ValueError:
      print('%s Missing metadata. Not parsing.' % item('warn'))
      continue

    else:
      datasets.append(parsed_data['metadata'])
      resources += parsed_data['resources']

  #
  #  Exports JSON files to disk..
  #
  export_json(datasets, 'data/datasets.json')
  export_json(resources, 'data/resources.json')
  print('%s Total datasets downloaded %s' % (item('success'), str(len(datasets))))
Пример #2
0
def LoadConfig(config_path="config/dev.json", verbose=True):
    """Load configuration parameters."""

    try:
        with open(config_path) as json_file:
            config = json.load(json_file)

    except Exception as e:
        print("%s Couldn't load configuration." % item("error"))
        if verbose:
            print(e)
        return False

    return config
def main(verbose=True):
  '''
  Wrapper to run all the scheduled tasks.

  '''

  if verbose:
    print('%s Running scheduler.' % item('bullet'))

  try:
    while True:
      schedule.run_pending()
      time.sleep(1)

  except Exception as e:
    print(e)
    return False
 def test_item_returns_correct_type(self):
   '''
   item() utility function returns the right type.
   '''
   result = item('bullet')
   self.assertIs(type(result), str)
      continue

    try:
      parsed_data = parse_dataset(o)
    except ValueError:
      print('%s Missing metadata. Not parsing.' % item('warn'))
      continue

    else:
      datasets.append(parsed_data['metadata'])
      resources += parsed_data['resources']

  #
  #  Exports JSON files to disk..
  #
  export_json(datasets, 'data/datasets.json')
  export_json(resources, 'data/resources.json')
  print('%s Total datasets downloaded %s' % (item('success'), str(len(datasets))))


if __name__ == '__main__':
  try:
    main()
    print('%s Successfully collected IFPRI data.' % item('success'))
    scraperwiki.status('ok')

  except Exception as e:
    print('%s Failed to collected IFPRI data.' % item('error'))
    print(e)
    scraperwiki.status('error', 'Failed to collect data.')