## Process this file with automake to produce Makefile.in ## Copyright: Sámediggi/Divvun/UiT ## Licence: GPL v3+ # The current dir must be built after weighting, but before the rest. SUBDIRS = weighting . hfst vfst # Name of speller corpus sans suffix: CORPUSNAME=spellercorpus # Specify whether to use corpus weights for frequency ranking of suggestions. # Make sure you test the suggestion quality both before and after changing this # setting. ENABLE_CORPUS_WEIGHTS=no # Number of lines (= word forms) in the sort-uniqued corpus. If empty the whole # corpus will be used. Too large a corpus can make the acceptor fst unneccessary # big without improving the suggestion quality in a noticable way. A starting # point could be to only include word forms occurring at least twice. Try # different values to find the best compromise between fst size and suggestion # quality. CORPUS_SIZE= ### BEGIN: Local build rules: *.hfst: *.tmp.hfst ### #### END: Local build rules: *.hfst: *.tmp.hfst #### include $(top_srcdir)/am-shared/tools-spellcheckers-fstbased-mobile-dir-include.am