python - the usage or API of tf.app.flags -

when reading cifar10 example, can see following code segment, said follow google commandline standard. in specific, code segment do? did not find api document cover tf.app.flags.define_string

flags = tf.app.flags.flags  tf.app.flags.define_string('train_dir', '/tmp/cifar10_train',                        """directory write event logs """                        """and checkpoint.""") tf.app.flags.define_integer('max_steps', 1000000,                         """number of batches run.""") tf.app.flags.define_boolean('log_device_placement', false,                         """whether log device placement.""") 

my experience tensorflow looking @ source code more useful ctrl+f in api doc. keep pycharm open tensorflow project, , can search either example of how (e.g., custom reader).

in particular case, want @ what's going on in tensorflow/python/platform/flags.py. it's thin wrapper around argparse.argumentparser(). in particular, of define_* end adding arguments _global_parser, example, through helper function:

def _define_helper(flag_name, default_value, docstring, flagtype):     """registers 'flag_name' 'default_value' , 'docstring'."""     _global_parser.add_argument("--" + flag_name,                                 default=default_value,                                 help=docstring,                                 type=flagtype) 

so flags api same find argumentparser.