Remote Data Manager¶
There are functions which require files to be downloaded from the internet before they are executed. If the files change on the remote server or if they are tampered with after the file is downloaded, the function may return erroneous results.
This Remote Data Manager provides developers with a way to download the files easily and ensure that the files are intact when they are used. If the file is changed on the remote server, the data manager will raise an error so that the user is made aware that the intended file is not available. If the file has changed on disk, data manager will redownload the file after warning the user.
The manager has to be imported before it is used:
from sunpy.data import manager
require is a decorator which is used to inform the data manager that the function requires a file for its execution.
get function is used to access the file inside the function.
Suppose a function requires a file with url ‘http://data.sunpy.org/sample-data/predicted-sunspot-radio-flux.txt’ which has a SHA256 hash of ‘4c85b04a5528aa97eb84a087450eda0421c71833820576330bba148564089b11’.
The following example will show how this function can be implemented:
@manager.require('test_file', ['http://data.sunpy.org/sample-data/predicted-sunspot-radio-flux.txt'], '4c85b04a5528aa97eb84a087450eda0421c71833820576330bba148564089b11') def test_function(): file_path = manager.get('test_file') # do something with file return file_path
The following example shows how cache can be used:
from sunpy.data import cache def test_function(): file_path = cache.download('http://data.sunpy.org/sample-data/predicted-sunspot-radio-flux.txt') return file_path test_function() # The file will be downloaded with this call test_function() # subsequent calls won't redownload the file
A pytest fixture is provided for ease of mocking network requests when using cache. The following example demonstates the usage of the fixture.:
@pytest.fixture() def local_cache(sunpy_cache): sunpy_cache = sunpy_cache('sunpy.test_module.cache') sunpy_cache.add('http://example.com/test_file', 'test_data_path')
The above snippet creates a pytest fixture called
local_cache. This fixture can be used in wherever the files have to be mocked.
An example is given below.:
def test_test_function(local_cache): # inside this function the mocked cache is used # test_function uses 'http://example.com/test_file' assert test_function() == True