Make Semitic mode compatible with inflection operators R, C and D

[martinec]

Submitted by @eric-laporte

Feature Request

In the Unitex/GramLab code for inflection of simple words, the Semitic mode (user manual, section 3.5.4) is presently incompatible with R, C and D, which are 3 of the inflection operators described in the manual (section 3.5.1).

Current work with Prihantoro (Diponegoro University) suggests that it would be useful for the description of Indonesian to use both the Semitic mode and the R operator in the same inflection graphs: the Semitic mode for reduplication, and R for morphological variations of the stem. Some Indonesian words undergo both reduplication and morphological variations of the stem.

Example

• DELAs entry: balik
• DELAF entry: balak-balik

The C and D operators are likely to be useful in the same conditions.

I suggest the Semitic mode should be made compatible with the R, C and D operators.

It is already compatible with the L, <R=?>, <I=?> and <X=n> operators.

In the attached logs, the dictionary invokes 2 inflection graphs:

• V232LLak-ref.grf without the Semitic mode,
• V232LLak.grf in the Semitic mode.

The first graph produces the morphological variants (e.g. balak) but not the reduplication.
The second graph produces the reduplication, but not all the morphological variants work.

Thanks,

Eric