Arcgis-proxy is a microservice for the GFW API. It allows making calls to ArcGIS Server using the GFW geostore.
The proxy fetches the geostore geometry, converts it ESRI JSON and projects it into Web Mercator projections and forwards the call to ArcGIS server. It returns the ArcGIS server response.
Dependencies on other Microservices:
In its current version the proxy only supports calls to ImageServer service for the computeHistorgrams function
/v1/arcgis-proxy/ImageServer/computeHistograms
| Parameters | Explanation | Data Type |
|---|---|---|
| server | a known ArcGIS server instance (either gfw or forest-atlas) | string |
| serverUrl | URL to the ArcGIS Server instance (server web adaptor). Use either server or serverUrl | URL |
| service | the service name | String |
| renderingRule | the rendering rule | JSON |
| pixelSize | pixel size for calculation | integer |
| geostore | geostore ID | hash |
Example
http://production-api.globalforestwatch.org//v1/arcgis-proxy/ImageServer/computeHistograms?server=forest-atlas&service=eth/EthiopiaRestoration&geostore=d1193c16181805c30701dea9a173e30b&renderingRule={"rasterFunction":"Arithmetic","rasterFunctionArguments":{"Raster":"$1","Raster2":"$6","Operation":3}}&pixelSize=100
As this microservice relies on Google Earth Engine, tests require a valid storage.json or equivalent file.
At the time of this writing, actual tests use mock calls, so the real credential are only needed because Google's
library actually validates the credentials on startup.
Before you run the tests, be sure to install the necessary development libraries, using pip install -r requirements_dev.txt.
Actual test execution is done by running the pytest executable on the root of the project.