Explore integration of code into CEDA ESGF STAC architecture

[huard]

CEDA has developped a nice architecture to parse metadata out of a filesystem. Some of the THREDDS parsing code could be integrated over there.

[huard]

• Asked permission to create PR for NcML attribute extractor plugin: https://github.com/cedadev/asset-scanner/issues/39
• Created NcML header_extract backend in asset-scanner
• Installed crim-ca/stac-sandbox (comment nginx in docker-compose - hardcoded ssh key; pip install uvicorn)
• Installed crim-ca/stac-ingest
• pip install -e asset-generator item-generator asset-scanner
• pip install siphon xarray lxml
• adjust asset-scanner-example (change paths to reflect the fact that we use ncml, not dodsC), use access_urls.NCML
• Opened PR: https://github.com/cedadev/asset-scanner/pull/42

[huard]

Opened PR for intake-esm output plugin https://github.com/cedadev/asset-scanner/pull/46 Wrote a Controlled Vocabulary post extraction processor, but it doesn't seem to be run.

[huard]

Opened PR for metadata validation: https://github.com/cedadev/asset-scanner/pull/48