Closed sidneymbell closed 2 years ago
Thanks to @jacoblerner-czi for consulting on this ticket from a security standpoint! Takeaways:
Requirements and pointers on where to start (cc @oliviabholmes to add / adjust) --
ingestNextstrain(path/to/treefile)
& dispatch({type: "tree file uploaded", data: output_from_ingest_nextstrain})
. find_leaf_by_name(name, traverse_preorder(tree))
--> dispatch({type: "mrca clicked", data: output_from_find_leaf_by_name})
division
(state) and location
(county)const { tidyMetadata, metadataCensus } = ingestCSVMetadata(results.data); dispatch({ type: "metadata uploaded and parsed", data: { metadataCensus: metadataCensus, tidyMetadata: tidyMetadata }, });
@sidneymbell how are we currently ingesting JSON? Trying to understand why this is a blocker
@oliviabholmes -- currently via direct upload
It's a blocker because it's required to take in data from CZ GEN EPI via the planned integration, which uses the same pre-signed URLs for trees that we use to link out to nextstrain.
Related to #145 and #147
Notes from @vincent-czi --
For folks coming from CZ GEN EPI, don't dump them straight into the analysis page; rather, prepopulate the upload screen modal with their json and preliminary guess as to their jurisdiction of interest (based on their DPH home jurisdiction set in cz gen epi). This allows them to also upload metadata and/or change jurisdiction of interest without requiring #162 or #161 .
One way to do this --
division
and location
) in the URL alongside the tree jsondivision
and location
division
+/- location
are valid options, then set those as default for the user
(Or even use fuzzy string matching to choose the closest valid option)Status update: fetch is working for JSON and URL
Propose splitting the ticket to add the query params late next week (technically after launch).
Closed as main chunk of work -- taking in the external URL JSON -- is complete. Follow up work for doing query / search params filed as #186
Similar to nextstrain.org/fetch/URL