osmapiR
An R interface to OpenStreetMap API for fetching and saving data from/to the OpenStreetMap database
osmapiR is the rare API client that tracks its server's wiki revision numbers in the changelog.
◆Recent moves
- 5mo ago
bbox arguments accept sf and terra objects
Bounding boxes can now be passed as character, matrix, vector, an sf `bbox`, or a terra `SpatExtent`, removing the conversion step between the two dominant R spatial stacks and this client. Also adds changeset discussion comment search and the release's customary sync against the OSM wiki revision range.
View source ↗ - 0y ago
Note search defaults to creation order; JSON for GPX metadata
A default worth changing — `osm_search_notes()` sorts by `created_at` rather than `updated_at`, which is what someone looking for new notes actually wants. JSON output arrives for `osm_get_gpx_metadata()`, internal endpoint functions are renamed with a dot prefix, and an httr2 1.2.0 compatibility fix lands ahead of that release.
View source ↗ - 1y ago
Note subscriptions and user block endpoints added
Extends coverage into moderation and participation: subscribe and unsubscribe from notes, create and read user blocks, list active ones, with a vectorised `osm_get_user_blocks()` alongside the single-record reader. Deprecated endpoints are retired in the same pass.
View source ↗ - 1y ago
Changeset queries gain from and to parameters
Adds `from` and `to` to `osm_query_changesets()` and fixes the older `time`/`time_2` pair, which had not been working as documented. Also picks up `httr2::oauth_cache_clear()` for token cache management.
View source ↗ - 1y ago
JOSS citation added; single-tag conversion fixed
Records the JOSS article in CITATION and fixes `tags_list2wide()` for objects carrying exactly one tag. Small, though the single-tag case is the sort of edge condition that silently produces the wrong shape.
View source ↗ - 2y ago
Complete OSM API coverage arrives in one release
⚡ SPARKThe initial release is unusually complete: every API endpoint, OAuth2 where required, pagination handled internally, and vectorised atomic calls. Starting from full coverage is why subsequent releases have been able to track the server's changes rather than chase missing features.
View source ↗