9.14. Shapefile Converter

The shapefile converter processes shapefiles. Since a shapefile is a collection of files, and not a single input, there are a few constraints that must be observed:

Shapefile converters are not usable in distributed converter jobs, as the map/reduce paradigm does not work well with the collection of related files that comprise a shapefile. In addition, when using a shapefile converter it is important to set the input file path in the evaluation context. This is handled automatically by the GeoMesa CLI tools, but if used programmatically "inputFilePath" must be set in the evaluation context global parameters.

9.14.1. Configuration

The shapefile converter supports the following configuration keys:

Key	Required	Type	Description
type	yes	String	Must be the string shp.

9.14.2. Transform Functions

The transform element supports referencing each attribute in the input shapefile by its column number using $. $0 refers to the feature ID, then the first attribute is $1, etc. Each attribute will be typed according to the schema of the shapefile.

In addition to the standard Transformation Functions, the shapefile converter provides the following functions:

9.14.2.1. shp

For ease of use, this will access an attribute in the shapefile by name, e.g. shp('my_attribute'). Note that attributes can also be referenced by number, as described above.

9.14.2.2. shpFeatureId

For ease of use, this will access the feature ID of the current shapefile feature, e.g. shpFeatureId(). Note that the feature ID can also be referenced by $0, as described above.

9.14.3. Example Usage

This example is for the Tiger US States boundary file available from the US Census Bureau.

The shapefile has the following columns:

*the_geom:MultiPolygon,STATEFP:String,STATENS:String,AFFGEOID:String,GEOID:String,STUSPS:String,NAME:String,LSAD:String,ALAND:Long,AWATER:Long

From these, we will only consider a subset, using the following SimpleFeatureType:

geomesa.sfts.cb_2017_us_state_20m = {
  attributes = [
    { name = "name",    type = "String"       }
    { name = "usps",    type = "String"       }
    { name = "area",    type = "Long"         }
    { name = "geom",    type = "MultiPolygon" }
  ]
}

You could ingest with the following converter:

geomesa.converters.cb_2017_us_state_20m = {
  type = "shp"
  id-field = "$0"
  fields = [
    { name = "name", transform = "$7" }, // example of lookup by field number
    { name = "usps", transform = "shp('STUSPS')" }, // example of lookup by name
    { name = "area", transform = "add(shp('ALAND'),shp('AWATER'))" },
    { name = "geom", transform = "shp('the_geom')" },
  ]
}