[pet] interactions of external morphology and generic lexical entries

[Stephan Oepen]

hi bernd,

in helping montse add generic entries to the SRG, i noticed that there
is an assumption in the current cheap that would make it impossible, i
think, to combine an external morphology and generic lexical entries.

for the SRG, input tokens come with a list of orthographemic rules to
be applied, i.e. morphological analysis is completed.  with stems for
which there are no native lexical entries, it is desirable to activate
generic entries.  the current code in lexparser::add_generics() (with
comments in the body of that function), however, assumes that generic
entries always correspond to `word' input tokens, i.e. it will invoke
the internal morphology component un-conditionally.

i relaxed that condition in the chart mapping branch, and for the SRG
at least things seem to work fine.  i expect those comments are yours,
so can you recall why it would be problematic to allow generic entries
for `stem' input items too?

for my (trivial) change of today, please see:

  https://pet.opendfki.de/changeset?new=504

i would like to eventually propagate this change into the main branch.
what do you say?

                                                    all best  -  oe

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
+++ Universitetet i Oslo (IFI); Boks 1080 Blindern; 0316 Oslo; (+47) 2284 0125
+++     CSLI Stanford; Ventura Hall; Stanford, CA 94305; (+1 650) 723 0515
+++       --- oe at ifi.uio.no; oe at csli.stanford.edu; stephan at oepen.net ---
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++