question about cf. in input

[abubelinha]

Some results of my Python calls to https://verifier.globalnames.org/api/v1/verifications/

• Input 1: Phacopsis oxyspora var. defecta Triebel & Rambold
  Bestresult: [Index Fungorum (5) 413758]: Phacopsis oxyspora var. defecta Triebel & Rambold 1995

• Input 2: Phacopsis oxyspora cf. var. defecta Triebel & Rambold
  Bestresult: [Catalogue of Life (1) 4FKN3]: Phacopsis oxyspora (Tul.) Triebel & Rambold
  Preferred: [Index Fungorum (5)134632]: Phacopsis oxyspora (Tul.) Triebel & Rambold 1988

I would expect that with or without "cf.", the input should match the infraspecific rank.
But when "cf." is present gnames matched the next higher taxonomic rank (not sure if this is on purpose or not)

I have only seen this with infraspecific ranks input. With species input, that is not happening:

• Input 3: Phacopsis cf. oxyspora
  Bestresult: [Catalogue of Life (1) 4FKN3]: Phacopsis oxyspora (Tul.) Triebel & Rambold
  Preferred: [Index Fungorum (5)134632]: Phacopsis oxyspora (Tul.) Triebel & Rambold 1988

May this be a bug, or should I take care of removing "cf." myself from input, to get my expected results in output?
I linked years above to a different question. Thanks!

[dimus]

The reason behind it is the semantic meaning of 'cf.' in the name https://en.wikipedia.org/wiki/Cf.

GNparser considers names with cf. to be be surrogates, assuming someone could not identify species correctly, and compared them to something that they are 'not'. So in this case it would be a mistake to match such name with details after cf. annotation.

[abubelinha]

Yes, I understand what you mean.
The part of the name after cf. is just a guess/suggestion (might be or might be not correct). So gnverifier takes the more conservative approach and decides not to take it into account:

So in this case it would be a mistake to match such name with details after cf. annotation.

No problems with that (if I do prefer to be risky, I could just remove " cf. " from the names in advance)

But the same gnverifier logic should be applied to both specific and infraspecific ranks. Shouldn't it?

You say Phacopsis cf. oxyspora should be matched to Phacopsis (and not Phacopsis oxyspora),
But that's not happening (see api link or 3rd example above).
Am I missing something?

[dimus]

You say Phacopsis cf. oxyspora should be matched to Phacopsis (and not Phacopsis oxyspora),

looks like a bug in gnparser, adding an issue there

[dimus]

gnames/gnparser#216

[dimus]

I was wrong actually @abubelinha, gnparser suppose to stop parsing if 'aff.' used, while for 'cf.' we decided to parse name anyway,
but issue a warning and bring quality to 4. So you are correct that Phacopsis oxyspora cf. var. defecta Triebel & Rambold should give the same canonical as Phacopsis oxyspora var. defecta Triebel & Rambold according to our rules. I just did not implement 'cf.' for infraspecies yet. So I will change gnames/gnparser#216 accordingly

[abubelinha]

Thanks @dimus
I agree: aff. xxx means "this resembles but is different to xxx" (a negative meaning).
But cf. xxx means "this could be xxx" (a positive meaning, saying nothing against xxx).