[Snowball-discuss] Another Porter2 question: "able" sample seems wrong

[Martin Holmes]

Hi Martin,

I've hit another issue with exactly this point in the algorithm. The 
word "knife" (to take one example) survives unchanged as far as Step 5. 
Then we have the instruction:

e  :  delete if in R2, or in R1 and not preceded by a short syllable

R1 for knife is position 5 (e), and R2 is position 6 (empty).

So:

  - the final e is not in R2; but

  - The final e is in R1, so the second condition applies:

If e is preceded by a short syllable, then it should be deleted.

What precedes the e is "knif". This seems to me to match the definition 
of a short syllable:

"Define a short syllable in a word as either (a) a vowel followed by a 
non-vowel other than w, x or Y and preceded by a non-vowel, or * (b) a 
vowel at the beginning of the word followed by a non-vowel."

Definition a) seems to apply: a vowel (i) followed by a non-vowel other 
than w, x or Y (f) and preceded by a non-vowel (n).

Therefore the e should be deleted. But in the test data it isn't.

What am I missing?

All help appreciated,
Martin


On 2019-06-04 12:36 p.m., Martin Porter wrote:
> yes, I think your last point is effectively the right way of looking
> at it. 'e' is in R1, but 'e' itself is preceded by (immediately
> preceded by) 'abl', which is not a short syllable.
> 
> The idea of all this is to distinguish (for example), ap and ape, so
> ap and aps stem to ap, and ape, apes, aping, aped stem to ape. It is a
> distinction for the "short syllable" case.
> 
> (Incidentally, I am somewhat quoting from memory here, not having
> access to my Linux machine, but I think I've got this right.)
> 
> _______________________________________________
> Snowball-discuss mailing list
> Snowball-discuss at lists.tartarus.org
> https://lists.tartarus.org/mailman/listinfo/snowball-discuss
>