Skip to content


Subversion checkout URL

You can clone with
Download ZIP
A collector/downloader for IOOS like metadata
Branch: master

Fetching latest commit…

Cannot retrieve the latest commit at this time

Failed to load latest commit information.


Metadown is a programatic collector/downloader for IOOS like ISO 19115-2 metadata.

Services supported:


metadown is available on pypi and is easiest installed using pip.

pip install metadown

lxml, requests, and thredds_crawler will be installed automatically



The ThreddsCollector can take three optional arguments. Both are strongly suggested so you don't crawl an entire THREDDS server (unless that is what you want to do).

  • debug (bool) - Print what is happening to the console. Useful for debugging what to put into selects and skips.
  • selects (list) - Select datasets based on their THREDDS ID. Python regex is supported.
  • skips (list) - Skip datasets based on their name and catalogRefs based on their xlink:title. By default, the crawler uses some common regular expressions to skip lists of thousands upon thousands of individual files that are part of aggregations or FMRCs (they are below.) Setting the skip parameter to anything other than a superset of the defaults below runs the risk of having some angry system admins after you.

    • .*files.*
    • .*Individual Files.*
    • .*File_Access.*
    • .*Forecast Model Run.*
    • .*Constant Forecast Offset.*
    • .*Constant Forecast Date.*

You can access the default skip list through the ThreddsCollector.SKIPS class variable

> from metadown.collectors.thredds import ThreddsCollector
> print ThreddsCollector.SKIPS
  '.*Individual Files.*',
  '.*Forecast Model Run.*',
  '.*Constant Forecast Offset.*',
  '.*Constant Forecast Date.*'

If you need to remove or add a new skip, it is strongly encouraged you use the SKIPS class variable as a starting point!

> from metadown.collectors.thredds import ThreddsCollector
> skips = ThreddsCollector.SKIPS + [".*-Day-Aggregation"]
> metadata_urls = ThreddsCollector("", selects=[".*-Agg"], skips=skips).run()
> print metadata_urls


from metadown.collectors.geonetwork import GeoNetworkCollector

gnc = GeoNetworkCollector("")
metadata_urls =
print metadata_urls

The WaterQualityDataUs constructor accepts any parameters that the webservice supports. Please see parameter table here for a list of options.

from metadown.collectors.waterqualitydataus import WaterQualityDataUs

bbox = (-78, 28, -70, 32)
isos = WaterQualityDataUs(bBox=bbox).run()
for iso in isos:
    with open("your_save_path.xml", "w") as f:

You can also load the ISO into an XML object. For more information on the ISO object, please see the OWSLib project.

from metadown.collectors.waterqualitydataus import WaterQualityDataUs
from owslib.iso import MD_Metadata

bbox = (-78, 28, -70, 32)
isos = WaterQualityDataUs(bBox=bbox).run()
for iso in isos:
    xml_obj  = MD_Metadata(etree.fromstring(str(iso)))

Note: The WaterQualityDataUs object does not return URLs to ISO files, it creates the ISO files manually from metadata avaialble through the webservice. You will need to save the XML strings to a file yourself.

Downloading resulting ISO files

Once you have the metadata urls for the data you want, do whatever you want! If you would like to rename or modify the metadata files, there is a helper class called XmlDownloader

XmlDownloader takes in three parameters:

  • url_list (required) - a list of URLs
  • download_path (required) - folder to download files to on your local machine
  • namer (optional) - a python function for renaming the metadata files before saving them to your local machine. It should take in a single url and return a string filename for the url to be saved as.

    Example namer function that renames GeoNetwork URLs

    from urlparse import urlsplit
    def geonetwork_renamer(url, **kwargs):
        uid = urlsplit(url).query
        uid = uid[uid.index("=")+1:]
        return "GeoNetwork-" + uid + ".xml"
  • modifier (optional) - a python function for full control over the metadata content. It should take in a single url and return a str representation of ISO19115-2.

    Example modifier function that translates GeoNetwork's ISO19139 to ISO19115-2

    from metadown.utils.etree import etree
    def geonetwork_modifier(url, **kwargs):
        gmi_ns = ""
        new_root = etree.Element("{%s}MI_Metadata" % gmi_ns)
        old_root = etree.parse(url).getroot()
        # carry over an attributes we need
        [new_root.set(k,v) for k,v in old_root.attrib.items()]
        # carry over children
        [new_root.append(e) for e in old_root]
        return etree.tostring(new_root, encoding="UTF-8", pretty_print=True, xml_declaration=True)
from metadown.downloader import XmlDownloader, download_directory, namer=geonetwork_renamer, modifier=geonetwork_modifier)
Something went wrong with that request. Please try again.