Menu

#223 étkezési is not stemmed correctly

v1.0 (example)
closed-fixed
nobody
None
5
2014-05-28
2012-07-31
Anonymous
No

The word étkezési is not stemmed correctly. In fact, it cannot be stemmed at all:

spell(étkezési, ...): 1, info: 0, root: étkezik compound: 0, forbidden: 0, allcap: 0, nocap:0, title: 0

stem(étkezési): 0

analyze(étkezési): 2
st:étkezik po:vrb ts:PRES_INDIC_INDEF_SG_3 ds:Ás_PROCESS/RESULT_noun is:i_PLACE/TIME_adj ts:NOM
st:étek po:noun ts:PLUR ts:NOM ds:z_ACTION_vrb ds:Ás_PROCESS/RESULT_noun is:i_PLACE/TIME_adj ts:NOM

I don't know the official standpoint, but from the user's point of view, stem() should NOT return 0, if the word is in the dictionary, and can be analyzed.

This bug occurs in both 1.2 and 1.3.

Discussion

  • Németh László

    A dictionary bug (missing terminating suffix after the derivative suffix).

    It can be solveable by this sed conversion (affected 110 lines of the 20 thousand affix rules):

    sed 's/\( ds:[^:]* \)is:/\1ts:NOM is:/' < hu_HU.aff.orig  >hu_HU.aff
    

    Thanks for the bug report.

     
  • Németh László

    • status: open --> closed-fixed
    • Group: --> v1.0 (example)
     
  • Németh László

    This dictionary problem will be fixed in the next release of the Hungarian Hunspell dictionary, too.