Skip to content

Latest commit

 

History

History
143 lines (107 loc) · 6.55 KB

README.md

File metadata and controls

143 lines (107 loc) · 6.55 KB
Raw

Mapboard topology manager

This project a set of PostgreSQL/PostGIS schema definitions and procedures to iteratively solve the linework for a geologic map. These procedures are wrapped in a high-level Python module that allows the topology to be managed programmatically, using a command-line interface, or a "watcher" process.

This project, originally called postgis-geologic-map, was renamed to reflect its close association with the Mapboard GIS iPad app. This application drives the topology management in the app's tethered mapping mode and is a core part of the in-development Mapboard Platform server application. Version 4 was rewritten in Python to support easier integration with other GIS applications, such as Macrostrat's map platform.

The most important elements of this tool are its database models and procedural SQL. The Python module and CLI largely wrap these elements.

This tool is similar to Luca Penasa's Mappy QGIS plugin. However, it relies heavily on the PostGIS spatial database system and focuses on iterative use and speed with large datasets.

Interfaces

Command-line interface

The topo command-line interface (CLI) is the primary way to interact with the topology manager tool.

Topology watcher

The watcher process, invoked using topo update --watch waits for changes to the underlying map data and runs a sequence of procedures to fill interstitial polygons. It supports multiple topologies (e.g. overlapping bedrock and surficial units) and line types.

Python module

In version 4, we inagurated a new Python-based design, with a mapboard.topology_manager module for library use. This module is in an early phase, but it will eventually be released on PyPI. For now, it underpins the command-line interface and the watcher process.

Workflow

Set up the database

This project relies on two PostgreSQL schemas, named map_data and map_topology by default. The map_data schema holds the source data for the map: linework and polygons (used to assign map units to the eventual space-filling polygons), along with map units and line types. The map_topology schema contains solved topological relationships, including polygonal space-filling units.

Currently, environment variables are used to configure the database connection, schema names, and SRID. See .env.example for an example of the required variables. The schemas for mapping and topology data can be configured, but a fairly specific structure is assumed. The minimal schema can be created using the topo create-tables command.

Editing the map

Add geometries to the map_data.linework and map_data.polygon topologies using the GIS platform of your choice. Units and line types are managed by foreign keys to the map_data.linework_type and map_data.polygon_type tables.

After linework and polygons are added to the database, the topology can be updated using the command topo update [--watch]. The output of topology building can be found in the map_topology.map_face layer.

Watch mode

The optional --watch flag enables the topology watcher daemon, to rebuild the topology concurrently with modifications (using --watch mode).

In watch mode, modifications to the map_data.linework and map_data.polygon tables are picked up automatically. In practice, this means that any GIS platform that can connect to PostGIS can be used to propagate changes. QGIS has been tested extensively, and ArcGIS support should be available depending on the version and its support for native PostGIS feature layers.

Removing the topology

The topology can be rebuilt from scratch by calling DROP SCHEMA map_topology CASCADE, without destroying mapping data.

Installation

The project can be installed as a Python package on recent versions of Python (3.10+). It also requires a PostgreSQL database with PostGIS installed. Notionally, all versions greater than PostgreSQL 10 and PostGIS 2.3 should work, but the project is currently tested on PostgreSQL 14 and greater.

Local installation

  1. Install Poetry with pip install poetry.
  2. Install Python dependencies with poetry install.
  3. Create a new PostgreSQL database to hold the mapping data (or you can specify an existing one!).
  4. Create an .env file to configure the application using the .env.example file as a template. Make sure to change the database connection info to the right values for your PostgreSQL connection.
  5. Run the application with poetry run topo. This will show a help page listing available commands.
  6. Create tables: topo create-tables.

Docker installation

Note: Docker installation is broken in Version 4. It will be fixed soon.

  1. Make sure Docker and docker-compose are installed using the instructions for your platform.
  2. Modify the docker-assets/docker-map-config.json configuration file to suit your needs (typically, this involves changing the srid and tolerance fields). A better way to configure the application in Docker is forthcoming.
  3. Run docker-compose up --build. No need for a local PostgreSQL installation!
  4. Connect to the geologic_map database on local port 54321.

"Hybrid" installation

Development with Docker tends to be slow unless heavily optimized, since the app and its relatively heavy Python dependencies must be recompiled on each build. One nice alternative is to run the database server in Docker while running the rest of the app locally. This is the approach taken by the CI GitHub workflow.

Contributing

Contributions in the form of raised issues or proposed changes are welcome. The core database code is a strong foundation, and the quality of the rest of the software around it needs much improvement.

TODO

  • Improve documentation and onboarding process.
  • Improve configurability and stability of Docker version
  • Move map_topology.subtopology table to map_digitizer schema (it currently breaks rule of no dependencies between the schemas). This is now outmoded by the mapboard.map_layer construct.
  • Stabilize and document vector-tile generation functionality. Vector tile creation has been moved out of this library.
  • TESTS!