ESIP2018: Difference between revisions
From OPeNDAP Documentation
⧼opendap2-jumptonavigation⧽
(Created page with "These JSON dialects create 1:1 onto mappings to the native "binary" representation / source encoding. NCO-JSON <-> nc4 hdf-json<-> hdf5 DAP2 JSON <-> DAP2 clowder: https:/...") |
No edit summary |
||
| Line 1: | Line 1: | ||
== JSON == | |||
These JSON dialects create 1:1 onto mappings to the native "binary" representation / source encoding. | These JSON dialects create 1:1 onto mappings to the native "binary" representation / source encoding. | ||
* NCO-JSON <-> nc4 | |||
* hdf-json<-> hdf5 | |||
* DAP2 JSON <-> DAP2 | |||
== BALTO == | |||
clowder: https://opensource.ncsa.illinois.edu/confluence/display/CATS/Home | clowder: https://opensource.ncsa.illinois.edu/confluence/display/CATS/Home | ||
==DSR== | |||
How to generate DSR? | How to generate DSR? | ||
* Bake the response into the code (current) | * Bake the response into the code (current) | ||
* Dynamically generate the response by: | * Dynamically generate the response by: | ||
** Get DDX/DMR and examine content in OLFS: Probably the most efficient but requires that we duplicate BES logic in the OLFS. For example: How do we determine if a dataset is a candidate for a covjson response? We have to examine the DDS/DAS/DDX/DMR to determine of the metadata is sufficient | ** Have OLFS ask BES what responses are supported for a particular dataset:The covjson handler in the BES does this as part of generating the covjson response. We could duplicate that work in the OLFS (yuck) or we could add code to covjson that allows us to ask the question. | ||
** Get DDX/DMR and examine content in OLFS: Probably the most efficient but requires that we duplicate BES logic in the OLFS. For example: How do we determine if a dataset is a candidate for a covjson response? We have to examine the DDS/DAS/DDX/DMR to determine of the metadata is sufficient and the BES code already does this | |||
** | ** | ||
Revision as of 19:20, 19 July 2018
JSON
These JSON dialects create 1:1 onto mappings to the native "binary" representation / source encoding.
- NCO-JSON <-> nc4
- hdf-json<-> hdf5
- DAP2 JSON <-> DAP2
BALTO
clowder: https://opensource.ncsa.illinois.edu/confluence/display/CATS/Home
DSR
How to generate DSR?
- Bake the response into the code (current)
- Dynamically generate the response by:
- Have OLFS ask BES what responses are supported for a particular dataset:The covjson handler in the BES does this as part of generating the covjson response. We could duplicate that work in the OLFS (yuck) or we could add code to covjson that allows us to ask the question.
- Get DDX/DMR and examine content in OLFS: Probably the most efficient but requires that we duplicate BES logic in the OLFS. For example: How do we determine if a dataset is a candidate for a covjson response? We have to examine the DDS/DAS/DDX/DMR to determine of the metadata is sufficient and the BES code already does this
