Closed 7yl4r closed 6 years ago
ACK @7yl4r I also noticed this when I was working with the examples. We should probably investigate how far this goes into the code. The examples I was working with were pretty timid to be honest, they did not really provide any advanced or testing functionality from CMR.
I've forked and been testing/improving some more, and may be able to submit some more pull requests when I have more concrete results, but what I really want is to get the download URL.
I just opened CMR#401 to address the lack of download url specifically since I don't really want to rewrite pyCMR to use the json endpoint.
Ah, here is the trouble. It is not a different version that is needed, pyCMR specifically targets the ECHO10-formatted xml endpoint.
CMR returns a simplified xml reference
if the endpoint is not specified, so the fix is to ensure you ask for echo10 in the built url. This is easy to do in `cmr.cfg
search_granule_url = https://%(cmr_host)s/search/granules.echo10
I'm seeing a number of issues with key names leading me to believe that there have been significant changes to the CMR API since this was implemented. I believe #27 is a symptom of this larger issue.
However, the CMR changelog doesn't indicate any major changes that I can correlate to API changes.
What version of the CMR API does this work with?
Below is an illustration of the expected vs actual API results as related to #27. The expected xml should contain a url, but the actual result from the API contains no url at all!
expected
aka something like
actual
example query
Note that the ftp download url can be found in the equivalent json API response. (one of the
feed.entry[n].links
)