Closed fb71 closed 8 years ago
Reminds a bit of Common Google Maps Engine Workflows, now in CartoDB.
Other aspects:
Data sources could also be specialized importers/adapters to different external system. For example i would implement a TomTom Webfleet Importer. Therefore i need the login data to the external system from the user and a schedule interval. With this schedule interval, the TomTom Webfleet Importer connects to the external system, reads all car informations including the current GPS Position and updates the internal feature collection.
I think, these part, specialized importers for enterprise APIs could be a real benefit.
Additionally, a ArcGIS Importer makes sense in this scenario. Internal its a specilization of a WMS/WFS importer, but for the user it requires only the ArcGIS login/pw and shows the user a list of all possible importable Datasets from his ArcGIS account. So the user has no need to find the WMS URL and copy/paste them into mapzone.
If i import a PNG file, i don't see any error message. And also in the logs i got
[WARN] UploadService : Upload handler already registered for id: 1378271589
Feature import basically works.
To do:
Open enhancement and bugs are worked on in separate issues.
Sub tasks:
Specification
The goal of an "import" ist to transform a given input into something that can be transfered into a standard data store. Currently we are about to import feature data und image data (WMS) but this might be expanded to work with raster data in the future. It is out of scope of the import to shape or beautify data. This ist done on the imported data.
There is a huge number of input sources and formats in many different flavours. A lot of them are not well specified and even if so the particular input might diverge. So the import is about assumtions, guesses and questions. For example:
The basic job of the import is to:
Verifies assumptions and answered questions are decisions. Decisions may depend on other decisions. Decisions are:
Data source / input