Matching of hybrid taxon names

[japonicus]

A hybrid taxon such as: Carex acuta x elata = C. x prolixa

might reasonably be searched for using any of:

"Carex elata x acuta" "Carex x prolixa" "Carex prolixa"

these search terms do not currently result in any matches

[sacrevert]

Note that the latter two cases appear to work now, and it is only the following that does not (i.e. when the user reverses the order of the parent taxa): "Carex elata x acuta"

[kitenetter]

relates to #16

[kazlauskis]

@sacrevert @japonicus are there any particular rules I should follow when inverting the hybrid name order for searches? i.e. GENUS {NAME1} x {NAME2} SOMETHING ELSE - is this the matter of just swapping NAME1 and NAME2 on the sides of the 'x'? Or are there other variations of this?

[sacrevert]

The only thing I can think of that might affect this is for the roses when the order of the names matters (because technically {NAME1} is the female parent. In these cases the suffix (f x m) or (female x male) is appended, but I don't know whether this is technically part of the {NAME2} string or not. Maybe this doesn't matter for this type of pattern matching however... @japonicus will have a more educated view!

[japonicus]

Thanks for picking this up.

In general the hybrid name parts can be swapped with impunity. A few edge cases to consider would be:

Triple hybrids, e.g. Salix aurita x lapponum x herbacea

Hybrids with a hybrid parent, e.g. Populus x jackii x nigra - in this case the two swappable parts are 'x jackii' and 'nigra'

Hybrids with parents specified to an infra-specific level, e.g. Salix cinerea subsp. cinerea × subsp. oleifolia - where the two parts are 'subsp. cinerea' and 'subsp. oleifolia' Salix alba var vitellina × babylonica var. pekinensis - where the parts are 'alba var vitellina' and 'babylonica var. pekinensis'

My inclination would be to ignore the more complex cases and deal only with the simple form of 'species1' x 'species2' [x 'species3']. If the name doesn't match that pattern then don't swap the parts, as a more complex hybrids could easily be mangled in odd ways. e.g. Populus × ⁠canadensis 'Serotina' × x jackii is quite nasty

As @sacrevert noted, there are some roses where the hybrid parent order is significant and should not be reversed. These will have an '(f x m)' qualifier (short for female x male), fortunately I don't think these are included in the UKSI at present so can be ignored.

[sacrevert]

Again, not sure what implications are, but the (f x m) Rosa stuff is already in the UKSI and therefore iRecord. Note that I think that there are also some instances where the suffix is given as (female x male) rather than (f x m).

[japonicus]

I think the f x m entries are all currently marked as synonyms (as in your screenshot) so wouldn't feature in the app.

Depending on how Chris interprets BSBI's recent UKSI revisions then the f x m names could come back as accepted names. I half wish I'd not resurrected them...

Essentially if a name includes the 'f x m' or 'female x male' qualifier then don't swap any hybrid parts. I don't think anyone using the app will actually want to enter (f x m) names in the field, so if/when the names return then possibly they might be excluded from the app dictionary altogether - but I'll raise a separate ticket for that, as there are probably other concepts that ought to be excluded as they would never be likely to be used validly for app based field recording.