Skip to content
Tom Kralidis edited this page Apr 4, 2018 · 1 revision

pycsw 3.0

Key Concepts

  • CSW4
  • JSON
  • OpenAPI

Requirements

From @ingenieroariel (2018-04-04)

  • I want pycsw to help me go from a folder of ISO or Dublin core metadata records on a folder structure to one big geojson seq document.
  • I want pycsw to be installable with zero additional dependencies.
  • I want pycsw to serve that big geojson seq document in a way that is compatible with WFS3 and STAC, even if it is slow.
  • I want pycsw to offer me the possibility of indexing that big fat geojson sequence document on a file based database like rocksdb and provide flexible full text search and geometry predicates for searching the catalog.
  • I want pycsw to view any geojson file as a catalog of particular features and serve them as wfs3 with no expectation of minimal fields except those defined by the geojson spec.
  • I want pycsw to be able to serve subsets of collections of objects (maps) with enough metadata to initialize a slippy map without additional requests to the main servers. This can be TMS links, vector tile links, default bbox, srid, list of enabled / disabled layers, tree based layer list.
  • I want pycsw to provide direct links to cloud optimized geotiffs on it's catalog and mark them correctly
  • I want pycsw to use all the metadata info to dynamically create html representations of a single database without bundling any javascript packages (see unpkg.com and d3-require).
  • I want pycsw to support relative links to static data living in the geojson index folder.
  • I want pycsw to easily dump a 50GB metadata index file from a disk to a client, with resumable abilities.
  • I want xml parsing to be optional
  • I want fancy indexing to be optional