Skip to contents

Classify URLs by what serves them

Source: R/urls.R
url_audit.Rd

Pure string work, vectorised, no network. Each URL is one of:

Usage

url_audit(urls, bases = atlas_bases())

Arguments

urls

character vector of URLs

bases

named list from atlas_bases()

Value

a data frame with url, host and class

Details

  • store: under the store base, any extension.

  • computed: a tile / API host (titiler, legacy titiler, STAC API, api, tile, tilecache, titilecache, h3t, pmtiles tile-proxy hosts). A bulk extension inside the query string (?url=x.tif) does not matter.

  • vm_bulk: any OTHER *.marinesensitivity.org / *.oceanmetrics.io host (file, storage, app, preview, ...) whose path ends in a bulk extension (.tif .tiff .pmtiles .parquet .gpkg .duckdb .zip .gz .nc; query and fragment ignored, case-insensitive), or a directory URL (ends in /) under /pmtiles/ or /derived/ on the file host. This is the egress leak.

  • page: the remaining URLs on those project hosts (HTML, JSON, STAC JSON, no extension) and on marinesensitivity.org itself.

  • external: anything else (other organisations' hosts).

NA input gives class NA.

Examples

url_audit(c("https://file.marinesensitivity.org/pmtiles/v9/rng_iucn/1.pmtiles?v=1",
            "https://file.marinesensitivity.org/stac/v9/collection.json"))
#>                                                                    url
#> 1 https://file.marinesensitivity.org/pmtiles/v9/rng_iucn/1.pmtiles?v=1
#> 2           https://file.marinesensitivity.org/stac/v9/collection.json
#>                         host   class
#> 1 file.marinesensitivity.org vm_bulk
#> 2 file.marinesensitivity.org    page