Suspect behavior while using name_backbone() with arg strict = TRUE

[damianooldoni]

Hi 👋

I was doing some "strict" name matching, but I got some higherrank matches, which should be not allowed by definition, or at least by documentation, see https://docs.ropensci.org/rgbif/reference/name_backbone.html#arg-strict

strict (logical) If TRUE it (fuzzy) matches only the given name, but never a taxon in the upper classification (optional)

Example:

library(rgbif)
#> Warning: package 'rgbif' was built under R version 4.5.2
strict_true <- rgbif::name_backbone(name = "Achelia sawayai s.l. Marcus, 1940", strict = TRUE)
stritc_false <- rgbif::name_backbone(name = "Achelia sawayai s.l. Marcus, 1940", strict = FALSE)
strict_true$matchType
#> [1] "HIGHERRANK"
stritc_false$matchType
#> [1] "HIGHERRANK"

^{Created on 2025-11-04 with reprex v2.1.1}

I remember I was using strict = TRUE explicitly to avoid higher rank matches. Did something change meanwhile? Thanks in advance.

[jhnwllr]

@damianooldoni
Potentially related to : #564

According to the tech docs:

https://techdocs.gbif.org/en/openapi/v1/species#/Searching%20names/matchNames

If set to true, fuzzy matches only the given name, but never a taxon in the upper classification.

So (I think) if a name in the upper classification is an exact match, it will still match. Just won't fuzzy match a name in the upper classification.
Any comments @mdoering?

BTW, we will be switch to API v2 soon.
#800

[mdoering]

Indeed the matching mode only determines how to match against the scientific name, not to avoid higher matches.
The documentation is wrong in that regard. From the java docs:

    /**
     * Fuzzy matching mode that allows for some minor differences in the scientific name.
     */
    FUZZY,

    /**
     * Strict matching mode that requires an exact match of the scientific name.
     */
    STRICT