Utility to recursively scrape ArcGIS MapServer data using the ArcGIS REST API.

ArcGIS MapServer REST queries are limited to 1,000 objects in some cases. This tool makes repeated calls until all data for a given layer (and all sub-layers) is extracted. Output can be GeoJSON file format or data may be written directly to Postgres database tables in PostGIS format. GIS clients - e.g. QGIS - can be configured to use vector layer data from PostGIS sources.

Ruby 2.0 or above - see Travis badge for tested Ruby versions.

A Postgres database with the PostGIS extension enabled for database export.

For data import to a database GDAL must be installed and specifically the ogr2ogr executable must be available in your path.

*NIX systems only - Linux/Mac OS X. ArcGIS MapServer data is readable directly by ArcGIS Windows clients 😉

The following esri geometry types are so far supported:

• esriGeometryPoint, esriGeometryMultipoint, esriGeometryLine, esriGeometryPolyline, esriGeometryPolygon

Annotation layers are ignored, as are layers with no esri geometryType.

Currently the JSON data for a whole layer is held in memory before being output. For large layers - e.g. >100,000 objects - this can be multiple GB of memory. Is this causes a problem for you please add a comment to issue #4.

Add this line to your application's Gemfile:

gem 'gis_scraper'

And then execute:

$ bundle

Or install it yourself as:

$ gem install gis_scraper

Configuration options may be set via a hash or specified in a Yaml file. The following options are available:

• :threads Scraping is multi-threaded. The number of threads to use may be set with this option (default: 8)
• :output_path For JSON output, the path used to write files to (default: '~/Desktop')

The following options are used to connect to a database:

• :host (default: 'localhost')
• :port (default: 5432)
• :dbname (default: 'postgres')
• :user (default: 'postgres')
• :password (default: nil)

These additional options are available when using output to a database and are applied to the ogr2ogr command:

• :srs Used to overide the source spacial reference system. Currently only EPSG string format is valid - e.g. 'EPSG:3109' (default: no overide)

To set via a hash

GisScraper.configure(:threads => 16)

Using a Yaml configuration file

GisScraper.configure_with 'path-to-Yaml-file'

GisScraper.config # returns the hash of configuration values

A LayerWriter object must be instantiated with one required arg - a Service/Layer URL (ending in an integer representing the layer number). Example:

writer = LayerWriter.new('http://gps.digimap.gg/arcgis/rest/services/StatesOfJersey/JerseyMappingOL/MapServer/0')

An optional second argument for the output path for JSON files may be specified. If so this overides the configuration option. Example:

writer = LayerWriter.new('http://gps.digimap.gg/arcgis/rest/services/StatesOfJersey/JerseyMappingOL/MapServer/0', '~/Desktop')

JSON output

writer.output_json

If the layer is type 'Feature Layer', a single file of JSON data will be saved (named the same as the layer). If the layer is type 'Group Layer', the sub-group structure is traversed recursively thus: Directories for each sub-group layer are created and JSON data files for each constituent feature layer written to them.

Output to a database

Valid database config options must be set. The following command will convert JSON files, create tables for each layer (& sub-layers, if any) and import the data. Table names are lowercased, prefixed '' and have spaces replaced with undescores. If a table with the same name exists the name is appended with ''.

writer.output_to_db

For the full specification clone this repo and run:

rake spec

Bug reports, pull requests (and feature requests) are welcome on GitHub at https://github.com/MatzFan/gis_scraper.

The gem is available as open source under the terms of the MIT License