## Include this file in top-level srcdir to compile FSTs
## Copyright: Sámediggi/Divvun/UiT
## Licence: GPL v3+

####### Automake targets: ########

if WANT_DICTIONARIES
if CAN_HFST
GT_ANALYSERS+=analyser-dict-gt-desc.hfstol \
				   analyser-dict-gt-desc-mobile.hfstol
GT_GENERATORS+=generator-dict-gt-norm.hfstol
endif # CAN_HFST

if CAN_XFST
GT_ANALYSERS+=analyser-dict-gt-desc.xfst \
				   analyser-dict-gt-desc-mobile.xfst
GT_GENERATORS+=generator-dict-gt-norm.xfst
endif # CAN_XFST
endif # WANT_DICTIONARIES

##### END Xerox target list #####

# Building dictionary fsts. The dictionary analyser differs from the
# regular analyser by NOT removing variant and homonym tags, so that we can
# generate proper (mini)paradigms for the given lemmas and variants (the
# analysis string is used as input to the generator).
# Tags to be removed:
#   - Sem/-taggar
#   - OLang/*
#   - MWE
analyser-dict-gt-desc.tmp.%: analyser-raw-gt-desc.%            \
					filters/remove-area-tags.%                 \
					filters/remove-dialect-tags.%              \
					filters/remove-orig_lang-tags.%            \
					filters/remove-usage_except_NGminip-tags.% \
					filters/remove-semantic-tags.%             \
					filters/remove-hyphenation-marks.%         \
					filters/remove-infl_deriv-borders.%        \
					filters/remove-word-boundary.%             \
					filters/remove-orthography-tags.%          \
					filters/remove-Orth_IPA-strings.%          \
					orthography/inituppercase.compose.%        \
					orthography/spellrelax.compose.%           \
					orthography/downcase-derived_proper-strings.compose.%
	$(AM_V_XFST_TOOL)$(PRINTF) "read regex \
				@\"filters/remove-area-tags.$*\"                 \
			.o. @\"filters/remove-dialect-tags.$*\"              \
			.o. @\"filters/remove-orig_lang-tags.$*\"            \
			.o. @\"filters/remove-usage_except_NGminip-tags.$*\" \
			.o. @\"filters/remove-semantic-tags.$*\"             \
			.o. @\"filters/remove-orthography-tags.$*\"          \
			.o. @\"filters/remove-Orth_IPA-strings.$*\"          \
			.o. @\"$<\"                                          \
			.o. @\"orthography/downcase-derived_proper-strings.compose.$*\" \
			.o. @\"filters/remove-hyphenation-marks.$*\"         \
			.o. @\"filters/remove-infl_deriv-borders.$*\"        \
			.o. @\"filters/remove-word-boundary.$*\"             \
			; \n\
			define fst \n\
			set flag-is-epsilon ON\n\
			read regex fst \
			.o. @\"orthography/inituppercase.compose.$*\"        \
			.o. @\"orthography/spellrelax.compose.$*\"           \
			;\n\
		 save stack $@\n\
		 quit\n" | $(XFST_TOOL)

# The mobile analyser is just like the regular dictionary analyser,
# except for one additional set of spellrelaxes:
analyser-dict-gt-desc-mobile.tmp.%: analyser-dict-gt-desc.tmp.% \
					orthography/spellrelax-mobile-keyboard.compose.%
	$(AM_V_XFST_TOOL)$(PRINTF) "read regex \
			    @\"$<\" \
			.o. @\"orthography/spellrelax-mobile-keyboard.compose.$*\" \
			;\n\
		 save stack $@\n\
		 quit\n" | $(XFST_TOOL)

# The generator does NOT NGminip tagged strings, since we do
# not want them in the miniparadigms generated by this transducer. At the same
# time, this transducer DOES CONTAIN the homonym and variant tags, to govern
# proper paradigm generation.
# Obligatory tags:
#   - homonymy tags
#   - variant tags
#   - Use/NGminip ??? - check with Lene!
generator-dict-gt-norm.tmp.%: analyser-raw-gt-desc.%           \
					filters/remove-area-tags.%                 \
					filters/make-optional-hyph-tags.%          \
					filters/make-optional-transitivity-tags.%  \
					filters/make-optional-semantic-tags.%      \
					filters/make-optional-v1-tags.%            \
					filters/make-optional-adv_comp-tags.%      \
					filters/remove-dialect-tags.%              \
					filters/remove-orig_lang-tags.%            \
					filters/remove-usage_except_NGminip-tags.% \
					filters/remove-orthography-tags.%          \
					filters/remove-Orth_IPA-strings.%          \
					filters/remove-error-strings.%             \
					filters/remove-hyphenation-marks.%         \
					filters/remove-infl_deriv-borders.%        \
					filters/remove-word-boundary.%             \
					orthography/downcase-derived_proper-strings.compose.%
	$(AM_V_XFST_TOOL)$(PRINTF) "read regex  \
				@\"filters/make-optional-transitivity-tags.$*\"  \
			.o. @\"filters/make-optional-hyph-tags.$*\"          \
			.o. @\"filters/make-optional-semantic-tags.$*\"      \
			.o. @\"filters/make-optional-v1-tags.$*\"            \
			.o. @\"filters/make-optional-adv_comp-tags.$*\"      \
			.o. @\"filters/remove-area-tags.$*\"                \
			.o. @\"filters/remove-dialect-tags.$*\"              \
			.o. @\"filters/remove-orig_lang-tags.$*\"            \
			.o. @\"filters/remove-usage_except_NGminip-tags.$*\" \
			.o. @\"filters/remove-error-strings.$*\"             \
			.o. @\"filters/remove-orthography-tags.$*\"          \
			.o. @\"filters/remove-Orth_IPA-strings.$*\"          \
			.o. @\"$<\"                                          \
			.o. @\"orthography/downcase-derived_proper-strings.compose.$*\" \
			.o. @\"filters/remove-hyphenation-marks.$*\"         \
			.o. @\"filters/remove-infl_deriv-borders.$*\"        \
			.o. @\"filters/remove-word-boundary.$*\"             \
			;\n\
		 save stack $@\n\
		 quit\n" | $(XFST_TOOL)


###############################################################
# Additional targets to support multiple writing systems:

define giella_dict_alt_ws_analysers
analyser-dict-gt-desc.%.tmp.$(1): analyser-raw-gt-desc.$(1)       \
					filters/remove-area-tags.$(1)                 \
					filters/remove-dialect-tags.$(1)              \
					filters/remove-orig_lang-tags.$(1)            \
					filters/remove-usage_except_NGminip-tags.$(1) \
					filters/remove-semantic-tags.$(1)             \
					filters/remove-hyphenation-marks.$(1)         \
					filters/remove-infl_deriv-borders.$(1)        \
					filters/remove-word-boundary.$(1)             \
					filters/remove-orthography-tags.$(1)          \
					filters/remove-Orth_IPA-strings.$(1)          \
					orthography/spellrelax.%.compose.$(1)         \
					orthography/$$(DEFAULT_WS)-to-%.compose.$(1)
	$$(AM_V_XFST_TOOL)$$(PRINTF) "read regex \
				@\"filters/remove-area-tags.$(1)\"                 \
			.o. @\"filters/remove-dialect-tags.$(1)\"              \
			.o. @\"filters/remove-orig_lang-tags.$(1)\"            \
			.o. @\"filters/remove-usage_except_NGminip-tags.$(1)\" \
			.o. @\"filters/remove-semantic-tags.$(1)\"             \
			.o. @\"filters/remove-orthography-tags.$(1)\"          \
			.o. @\"filters/remove-Orth_IPA-strings.$(1)\"          \
			.o. @\"$$<\"                                           \
			.o. @\"filters/remove-hyphenation-marks.$(1)\"         \
			.o. @\"filters/remove-infl_deriv-borders.$(1)\"        \
			.o. @\"filters/remove-word-boundary.$(1)\"             \
			.o. @\"orthography/$$(DEFAULT_WS)-to-$$*.compose.$(1)\"\
			.o. @\"orthography/spellrelax.$$*.compose.$(1)\"       \
			;\n\
		 save stack $$@\n\
		 quit\n" | $$(XFST_TOOL)
endef
$(foreach fst,hfst xfst foma,$(eval $(call giella_dict_alt_ws_analysers,$(fst))))

define giella_dict_alt_ws_generators
generator-dict-gt-norm.%.tmp.$(1): analyser-raw-gt-desc.$(1)      \
					filters/make-optional-hyph-tags.$(1)          \
					filters/make-optional-transitivity-tags.$(1)  \
					filters/make-optional-semantic-tags.$(1)      \
					filters/make-optional-v1-tags.$(1)            \
					filters/make-optional-adv_comp-tags.$(1)      \
					filters/remove-area-tags.$(1)                 \
					filters/remove-dialect-tags.$(1)              \
					filters/remove-orig_lang-tags.$(1)            \
					filters/remove-usage_except_NGminip-tags.$(1) \
					filters/remove-orthography-tags.$(1)          \
					filters/remove-Orth_IPA-strings.$(1)          \
					filters/remove-error-strings.$(1)             \
					filters/remove-hyphenation-marks.$(1)         \
					filters/remove-infl_deriv-borders.$(1)        \
					filters/remove-word-boundary.$(1)             \
					orthography/$$(DEFAULT_WS)-to-%.compose.$(1)
	$$(AM_V_XFST_TOOL)$$(PRINTF) "read regex \
				@\"filters/make-optional-transitivity-tags.$(1)\"  \
			.o. @\"filters/make-optional-hyph-tags.$(1)\"          \
			.o. @\"filters/make-optional-semantic-tags.$(1)\"      \
			.o. @\"filters/make-optional-v1-tags.$(1)\"            \
			.o. @\"filters/make-optional-adv_comp-tags.$(1)\"      \
			.o. @\"filters/remove-area-tags.$(1)\"                 \
			.o. @\"filters/remove-dialect-tags.$(1)\"              \
			.o. @\"filters/remove-orig_lang-tags.$(1)\"            \
			.o. @\"filters/remove-usage_except_NGminip-tags.$(1)\" \
			.o. @\"filters/remove-error-strings.$(1)\"             \
			.o. @\"filters/remove-orthography-tags.$(1)\"          \
			.o. @\"filters/remove-Orth_IPA-strings.$(1)\"          \
			.o. @\"$$<\"                                           \
			.o. @\"filters/remove-hyphenation-marks.$(1)\"         \
			.o. @\"filters/remove-infl_deriv-borders.$(1)\"        \
			.o. @\"filters/remove-word-boundary.$(1)\"             \
			.o. @\"orthography/$$(DEFAULT_WS)-to-$$*.compose.$(1)\"\
			;\n\
		 save stack $$@\n\
		 quit\n" | $$(XFST_TOOL)
endef
$(foreach fst,hfst xfst foma,$(eval $(call giella_dict_alt_ws_generators,$(fst))))

###############################################################
# Additional targets to support alternative orthographies:

define giella_dict_alt_orth_analysers
analyser-dict-gt-desc.%.tmp.$(1): analyser-raw-gt-desc.$(1)       \
					filters/remove-area-tags.$(1)                 \
					filters/remove-dialect-tags.$(1)              \
					filters/remove-orig_lang-tags.$(1)            \
					filters/remove-usage_except_NGminip-tags.$(1) \
					filters/remove-semantic-tags.$(1)             \
					filters/remove-hyphenation-marks.$(1)         \
					filters/remove-infl_deriv-borders.$(1)        \
					filters/remove-word-boundary.$(1)             \
					filters/remove-orthography-tags.$(1)          \
					filters/remove-Orth_IPA-strings.$(1)          \
					orthography/spellrelax.%.compose.$(1)         \
					orthography/raw-to-%.compose.$(1)
	$$(AM_V_XFST_TOOL)$$(PRINTF) "read regex \
				@\"orthography/raw-to-$$*.compose.$(1)\".i \
			.o. @\"filters/make-optional-transitivity-tags.$(1)\"  \
			.o. @\"filters/remove-area-tags.$(1)\"                 \
			.o. @\"filters/remove-dialect-tags.$(1)\"              \
			.o. @\"filters/remove-orig_lang-tags.$(1)\"            \
			.o. @\"filters/remove-usage_except_NGminip-tags.$(1)\" \
			.o. @\"filters/remove-semantic-tags.$(1)\"             \
			.o. @\"filters/remove-orthography-tags.$(1)\"          \
			.o. @\"filters/remove-Orth_IPA-strings.$(1)\"          \
			.o. @\"$$<\"                                           \
			.o. @\"filters/remove-hyphenation-marks.$(1)\"         \
			.o. @\"filters/remove-infl_deriv-borders.$(1)\"        \
			.o. @\"filters/remove-word-boundary.$(1)\"             \
			.o. @\"orthography/raw-to-$$*.compose.$(1)\"\
			.o. @\"orthography/spellrelax.$$*.compose.$(1)\"        \
			;\n\
		 save stack $$@\n\
		 quit\n" | $$(XFST_TOOL)

analyser-dict-gt-desc.%.tmp.$(1): analyser-raw-gt-desc.$(1)       \
					filters/remove-area-tags.$(1)                 \
					filters/remove-dialect-tags.$(1)              \
					filters/remove-orig_lang-tags.$(1)            \
					filters/remove-usage_except_NGminip-tags.$(1) \
					filters/remove-semantic-tags.$(1)             \
					filters/remove-hyphenation-marks.$(1)         \
					filters/remove-infl_deriv-borders.$(1)        \
					filters/remove-word-boundary.$(1)             \
					filters/remove-orthography-tags.$(1)          \
					filters/remove-Orth_IPA-strings.$(1)          \
					orthography/spellrelax.%.compose.$(1)         \
					orthography/$(DEFAULT_ORTH)-to-%.compose.$(1)
	$$(AM_V_XFST_TOOL)$$(PRINTF) "read regex \
				@\"orthography/$$(DEFAULT_ORTH)-to-$$*.compose.$(1)\".i \
			.o. @\"filters/make-optional-transitivity-tags.$(1)\"  \
			.o. @\"filters/remove-area-tags.$(1)\"                 \
			.o. @\"filters/remove-dialect-tags.$(1)\"              \
			.o. @\"filters/remove-orig_lang-tags.$(1)\"            \
			.o. @\"filters/remove-usage_except_NGminip-tags.$(1)\" \
			.o. @\"filters/remove-semantic-tags.$(1)\"             \
			.o. @\"filters/remove-orthography-tags.$(1)\"          \
			.o. @\"filters/remove-Orth_IPA-strings.$(1)\"          \
			.o. @\"$$<\"                                           \
			.o. @\"filters/remove-hyphenation-marks.$(1)\"         \
			.o. @\"filters/remove-infl_deriv-borders.$(1)\"        \
			.o. @\"filters/remove-word-boundary.$(1)\"             \
			.o. @\"orthography/$$(DEFAULT_ORTH)-to-$$*.compose.$(1)\"\
			.o. @\"orthography/spellrelax.$$*.compose.$(1)\"        \
			;\n\
		 save stack $$@\n\
		 quit\n" | $$(XFST_TOOL)
endef
$(foreach fst,hfst xfst foma,$(eval $(call giella_dict_alt_orth_analysers,$(fst))))

define giella_dict_alt_orth_generators
generator-dict-gt-norm.%.tmp.$(1): analyser-raw-gt-desc.$(1)      \
					filters/make-optional-hyph-tags.$(1)          \
					filters/make-optional-transitivity-tags.$(1)  \
					filters/make-optional-semantic-tags.$(1)      \
					filters/make-optional-v1-tags.$(1)            \
					filters/make-optional-adv_comp-tags.$(1)      \
					filters/remove-area-tags.$(1)                 \
					filters/remove-dialect-tags.$(1)              \
					filters/remove-orig_lang-tags.$(1)            \
					filters/remove-usage_except_NGminip-tags.$(1) \
					filters/remove-orthography-tags.$(1)          \
					filters/remove-Orth_IPA-strings.$(1)          \
					filters/remove-error-strings.$(1)             \
					filters/remove-hyphenation-marks.$(1)         \
					filters/remove-infl_deriv-borders.$(1)        \
					filters/remove-word-boundary.$(1)             \
					orthography/$(DEFAULT_ORTH)-to-%.compose.$(1)
	$$(AM_V_XFST_TOOL)$$(PRINTF) "read regex \
				@\"orthography/$$(DEFAULT_ORTH)-to-$$*.compose.$(1)\".i \
			.o. @\"filters/make-optional-transitivity-tags.$(1)\"  \
			.o. @\"filters/make-optional-hyph-tags.$(1)\"          \
			.o. @\"filters/make-optional-semantic-tags.$(1)\"      \
			.o. @\"filters/make-optional-v1-tags.$(1)\"            \
			.o. @\"filters/make-optional-adv_comp-tags.$(1)\"      \
			.o. @\"filters/remove-area-tags.$(1)\"                 \
			.o. @\"filters/remove-dialect-tags.$(1)\"              \
			.o. @\"filters/remove-orig_lang-tags.$(1)\"            \
			.o. @\"filters/remove-usage_except_NGminip-tags.$(1)\" \
			.o. @\"filters/remove-error-strings.$(1)\"             \
			.o. @\"filters/remove-orthography-tags.$(1)\"          \
			.o. @\"filters/remove-Orth_IPA-strings.$(1)\"          \
			.o. @\"$$<\"                                           \
			.o. @\"filters/remove-hyphenation-marks.$(1)\"         \
			.o. @\"filters/remove-infl_deriv-borders.$(1)\"        \
			.o. @\"filters/remove-word-boundary.$(1)\"             \
			.o. @\"orthography/$$(DEFAULT_ORTH)-to-$$*.compose.$(1)\"\
			;\n\
		 save stack $$@\n\
		 quit\n" | $$(XFST_TOOL)

generator-dict-gt-norm.%.tmp.$(1): analyser-raw-gt-desc.$(1)      \
					filters/make-optional-hyph-tags.$(1)          \
					filters/make-optional-transitivity-tags.$(1)  \
					filters/make-optional-semantic-tags.$(1)      \
					filters/make-optional-v1-tags.$(1)            \
					filters/make-optional-adv_comp-tags.$(1)      \
					filters/remove-area-tags.$(1)                 \
					filters/remove-dialect-tags.$(1)              \
					filters/remove-orig_lang-tags.$(1)            \
					filters/remove-usage_except_NGminip-tags.$(1) \
					filters/remove-orthography-tags.$(1)          \
					filters/remove-Orth_IPA-strings.$(1)          \
					filters/remove-error-strings.$(1)             \
					filters/remove-hyphenation-marks.$(1)         \
					filters/remove-infl_deriv-borders.$(1)        \
					filters/remove-word-boundary.$(1)             \
					orthography/raw-to-%.compose.$(1)
	$$(AM_V_XFST_TOOL)$$(PRINTF) "read regex \
				@\"orthography/raw-to-$$*.compose.$(1)\".i \
			.o. @\"filters/make-optional-transitivity-tags.$(1)\"  \
			.o. @\"filters/make-optional-hyph-tags.$(1)\"          \
			.o. @\"filters/make-optional-semantic-tags.$(1)\"      \
			.o. @\"filters/make-optional-v1-tags.$(1)\"            \
			.o. @\"filters/make-optional-adv_comp-tags.$(1)\"      \
			.o. @\"filters/remove-area-tags.$(1)\"                 \
			.o. @\"filters/remove-dialect-tags.$(1)\"              \
			.o. @\"filters/remove-orig_lang-tags.$(1)\"            \
			.o. @\"filters/remove-usage_except_NGminip-tags.$(1)\" \
			.o. @\"filters/remove-error-strings.$(1)\"             \
			.o. @\"filters/remove-orthography-tags.$(1)\"          \
			.o. @\"filters/remove-Orth_IPA-strings.$(1)\"          \
			.o. @\"$$<\"                                           \
			.o. @\"filters/remove-hyphenation-marks.$(1)\"         \
			.o. @\"filters/remove-infl_deriv-borders.$(1)\"        \
			.o. @\"filters/remove-word-boundary.$(1)\"             \
			.o. @\"orthography/raw-to-$$*.compose.$(1)\"\
			;\n\
		 save stack $$@\n\
		 quit\n" | $$(XFST_TOOL)
endef
$(foreach fst,hfst xfst foma,$(eval $(call giella_dict_alt_orth_generators,$(fst))))

###############################################################
# Fallback target for mobile variants of the alternating orth, ws analysers:
define giella_dict_alt_ws_mobile_analysers
analyser-dict-gt-desc-mobile.%.tmp.$(1): analyser-dict-gt-desc.%.$(1)
	$$(AM_V_CP)cp -f $$< $$@
endef
$(foreach fst,hfst xfst foma,$(eval $(call giella_dict_alt_ws_mobile_analysers,$(fst))))