underpass

underpass is a data mining tool inspired by overpass turbo, aiming to implement extra processing/filtering

the implementation differs highly from overpass-turbo's, as underpass performs the fetching and filtering steps on a backend, instead of directly on the user's browser

building

underpass required rust, cargo, node, and npm to be installed, and optionally the just command runner

manually

frontend is built by running either npm run build (for development), npm run watch (with hot reloading), or npm run prod (for production). typescript can be typechecked by running npm run typecheck

rust can be built and run by cargo run, and cargo run --release for production

remember to build the frontend first, as it gets included into the rust binary during compilation

with just

just run: build and run for development
just release: build a binary for release
just test: run js and rust tests, typecheck typescript
just get-data: fetch required datasets

datasets

underpass needs two datasets: taginfo and elevation. these are not included in this repo, and have to be generated before running underpass. they can be fetched with just get-data.

these datasets are optional, and underpass will run without them. in the case of missing elevation data, all points will be considered as having elevation 0. in the case of missing taginfo, no information about tags will be used.

improvements over overpass-turbo

first and foremost, node popups include a link to google maps and a link to copy coordinates for the node. this helps smooth the workflow of reviewing results

code editor

the code editor implements an overpass ql parser, which enables better syntax highlighting and smarter autocomplete

it also adds some features that are common in more fully featured, such as Ctrl+F to find/replace, and Ctrl+D to select next find match

overpass ql extensions

read docs/overpass-ql-extensions.md

node editor

underpass implements a node editor that allows for powerful filtering that cannot be simply with the overpass API. it can be used to set up advanced filters, such as filtering roads by their bearing. not many filters have been implemented yet, but more are comming soon

map

the map is implemented using the maplibre gl library, which has a better performance when dealing with larger amounts of nodes

annieversary/underpass