# -*- makefile -*-
#
# Makefile module for GIZA++ alignment (with mgiza)
# (c) 2011-2012 Ulrich Germann 

# --- Training Parameters ------------------------------------------------------
#
# mkcls: -n: iterations -c: classes. No space allowed after -n/-c!
mkcls_args = -n10 -c50 

# giza training
%/mgiza.cfg: m1=5
%/mgiza.cfg: m2=0
%/mgiza.cfg: mh=5
%/mgiza.cfg: m3=3
%/mgiza.cfg: m4=3
%/mgiza.cfg: nodumps=0
%/mgiza.cfg: onlyaldumps=0
%/mgiza.cfg: model4smoothfactor=0.4
%/mgiza.cfg: nsmooth=4
%/mgiza.cfg: NCPUS=8

# symal 
symal_grow_diag_final_and = -a=g -d=yes -f=yes -b=yes 
symal_args = ${symal_grow_diag_final_and}

# ------------------------------------------------------------------------------
# You should not have to edit anything below this line
# ------------------------------------------------------------------------------
gizaln = ${WDIR}/crp/trn/aln/giza
giztmp = $(gizaln)/tmp
gizout = $(gizaln)
gizaln.in = ${WDIR}/crp/trn/pll/clean/

.PHONY: giza giza-prep

giza: | $(gizout)/${L1}.txt.gz 
giza: | $(gizout)/${L2}.txt.gz 
giza: | $(gizout)/${L1}-${L2}.symal.gz
	@echo "GIZA WORD ALIGNMENT COMPLETE!"

other = $(if $(findstring $(1),${L1}),${L2},${L1})
fwd   = $(1)-$(call other,$(1))
bwd   = $(call other,$(1))-$(1)

$(gizout)/${L1}.txt.gz: a3file = $(gizout)/${L2}-${L1}.A3.final.gz
$(gizout)/${L1}.txt.gz: | $(gizout)/${L2}-${L1}.A3.final.gz
	$(lock)
	(zcat $(a3file) | perl -ne 'print if ++$$ctr%3 == 2;' | gzip > $@_) && mv $@_ $@
	$(unlock)

$(gizout)/${L2}.txt.gz: a3file = $(gizout)/${L1}-${L2}.A3.final.gz
$(gizout)/${L2}.txt.gz: | $(gizout)/${L1}-${L2}.A3.final.gz
	$(lock)
	(zcat $(a3file) | perl -ne 'print if ++$$ctr%3 == 2;' | gzip > $@_) && mv $@_ $@
	$(unlock)

$(gizout)/${L1}-${L2}.symal.gz: A3fwd = $(gizout)/${L1}-${L2}.A3.final.gz 
$(gizout)/${L1}-${L2}.symal.gz: A3bwd = $(gizout)/${L2}-${L1}.A3.final.gz 
$(gizout)/${L1}-${L2}.symal.gz: | $(gizout)/${L1}-${L2}.A3.final.gz 
$(gizout)/${L1}-${L2}.symal.gz: | $(gizout)/${L2}-${L1}.A3.final.gz $(giza2bal.pl)
	$(lock)
	$(giza2bal.pl) -d 'gunzip -c ${A3fwd}' -i 'gunzip -c ${A3bwd}' \
	| $(symal) $(symal_args) | perl -pe 's/^.*{##}\s+//' | gzip > $@_ && mv $@_ $@
	$(unlock)

# merge alignments produced by mgiza
$(gizout)/%.A3.final.gz: | $(giztmp)/%/mgiza.DONE 
	mkdir -p ${@D}
	$(lock)
	$(mgiza.merge) $(shell ls $(giztmp)/$*/$*.A3.final.part* 2>/dev/null) | gzip > $@_
	mv $@_ $@ 
	$(unlock)

# run mgiza:
%/mgiza.DONE: | %/mgiza.cfg
	$(lock)
	$(mgiza) $| && touch $@
	$(unlock)

$(giztmp) $(gizout):
	mkdir -p $@

# --- run mkcls (to get word classes) ------------------------------------------
# ATTENTION: mkcls does not accept spaces between option markers and 
# option arguments mkcls -n 10 -c 50 ... will make it fail!
# mkcls_classes and mkcls_iterations should be set in parameters.mak
stream = find -L $(1) -type f -name "*.$(2)" -or -name "*.$(2).gz" | sort | xargs zcat -f
mkcls_cmd = $(call stream,$(1),$(2)) | ${mkcls} $(mkcls_args) -p/dev/stdin -V$(3) opt
# mkcls: -n: iterations 
# mkcls: -c: classes 
# mkcls: -p: input data 
# mkcls: -V: word classes (output)

$(giztmp)/%.vcb.classes: ${mkcls}
$(giztmp)/%.vcb.classes: | $(gizaln.in)
	@echo CREATING $@
	$(lock)
	mkdir -p $(@D)
	@$(call mkcls_cmd,$|,$*,$@_) && mv $@_ $@
	@mv $@_.cats $@.cats
	$(unlock)

# NUMBERIZED CORPUS FOR GIZA
$(giztmp)/${L1}.vcb:       | $(giztmp)/${L1}-${L2}.snt
$(giztmp)/${L1}.vcb:       | $(giztmp)/${L1}-${L2}.snt
$(giztmp)/${L2}.vcb:       | $(giztmp)/${L1}-${L2}.snt
$(giztmp)/${L2}-${L1}.snt: | $(giztmp)/${L1}-${L2}.snt
#$(info $(addprefix $(pll-clean), .${L1}.gz))
$(giztmp)/${L1}-${L2}.snt: L1files = $(addsuffix .${L1}.gz, $(pll-clean))
$(giztmp)/${L1}-${L2}.snt: L2files = $(addsuffix .${L2}.gz, $(pll-clean))
$(giztmp)/${L1}-${L2}.snt: | $(giztmp) 
$(giztmp)/${L1}-${L2}.snt: | $(addsuffix .${L1}.gz, $(pll-clean))
$(giztmp)/${L1}-${L2}.snt: | $(addsuffix .${L2}.gz, $(pll-clean))
	$(lock)
	$(plain2snt) \
		<(ls $(L1files) | xargs zcat -f) \
		<(ls $(L2files) | xargs zcat -f) \
		-vcb1 $(giztmp)/${L1}.vcb_ -vcb2 $(giztmp)/${L2}.vcb_ \
		-snt1 $(giztmp)/${L1}-${L2}.snt_ -snt2 $(giztmp)/${L2}-${L1}.snt_
	mv $(giztmp)/${L1}.vcb_ $(giztmp)/${L1}.vcb
	mv $(giztmp)/${L2}.vcb_ $(giztmp)/${L2}.vcb
	mv $(giztmp)/${L1}-${L2}.snt_ $(giztmp)/${L1}-${L2}.snt
	mv $(giztmp)/${L2}-${L1}.snt_ $(giztmp)/${L2}-${L1}.snt
	$(unlock)

# .cooc files
$(giztmp)/${L1}-${L2}.cooc: V1 = $(giztmp)/${L1}.vcb 
$(giztmp)/${L1}-${L2}.cooc: V2 = $(giztmp)/${L2}.vcb 
$(giztmp)/${L2}-${L1}.cooc: V1 = $(giztmp)/${L2}.vcb 
$(giztmp)/${L2}-${L1}.cooc: V2 = $(giztmp)/${L1}.vcb 
$(giztmp)/%.cooc: | $(giztmp)/%.snt
	@echo CREATING $@
	$(lock)
	$(snt2cooc) $@_ ${V1} ${V2} $| && mv $@_ $@
	$(unlock)

################################################################################
# MGIZA CONFIG FILE:
#
$(giztmp)/%/mgiza.cfg: SHELL=bash
# --- CORPUS RESOURCES ---------------------------------------------------------
$(giztmp)/%/mgiza.cfg: V1   = $(giztmp)/${FROM}.vcb
$(giztmp)/%/mgiza.cfg: V2   = $(giztmp)/${TO}.vcb
$(giztmp)/%/mgiza.cfg: SNT  = $(giztmp)/${FROM}-${TO}.snt
$(giztmp)/%/mgiza.cfg: COOC = $(giztmp)/${FROM}-${TO}.cooc
$(giztmp)/%/mgiza.cfg: ODIR = $(giztmp)/${FROM}-${TO}/${FROM}-${TO}
#$(giztmp)/%/mgiza.cfg: | $(giztmp)/%.cooc $(giztmp)/%.snt 
$(giztmp)/${L1}-${L2}/mgiza.cfg: FROM = ${L1}
$(giztmp)/${L1}-${L2}/mgiza.cfg: TO   = ${L2}
$(giztmp)/${L2}-${L1}/mgiza.cfg: FROM = ${L2}
$(giztmp)/${L2}-${L1}/mgiza.cfg: TO   = ${L1}
$(giztmp)/${L1}-${L2}/mgiza.cfg: | $(giztmp)/${L1}-${L2}.cooc 
$(giztmp)/${L1}-${L2}/mgiza.cfg: | $(giztmp)/${L1}-${L2}.snt 
$(giztmp)/${L2}-${L1}/mgiza.cfg: | $(giztmp)/${L2}-${L1}.cooc 
$(giztmp)/${L2}-${L1}/mgiza.cfg: | $(giztmp)/${L2}-${L1}.snt 
$(giztmp)/${L1}-${L2}/mgiza.cfg \
$(giztmp)/${L2}-${L1}/mgiza.cfg: | \
	$(giztmp)/${L1}.vcb $(giztmp)/${L1}.vcb.classes \
	$(giztmp)/${L2}.vcb $(giztmp)/${L2}.vcb.classes 
	mkdir -p $(dir $@)
	touch $@
	@echo "s    ${V1}"   >> $@
	@echo "t    ${V2}"   >> $@
	@echo "c    ${SNT}"  >> $@
	@echo "cooc ${COOC}" >> $@
	@echo "m1   ${m1}"   >> $@
	@echo "m2   ${m2}"   >> $@
	@echo "mh   ${mh}"   >> $@
	@echo "m3   ${m3}"   >> $@
	@echo "m4   ${m4}"   >> $@
	@echo "t1   ${m1}"   >> $@
	@echo "t2   ${m2}"   >> $@
	@echo "th   ${mh}"   >> $@
	@echo "t3   ${m3}"   >> $@
	@echo "t4   ${m4}"   >> $@
	@echo "o    ${ODIR}" >> $@
	@echo "model4smoothfactor ${model4smoothfactor}"   >> $@
	@echo "onlyaldumps ${onlyaldumps}"                 >> $@
	@echo "nodumps ${nodumps}"                         >> $@
	@echo "nsmooth ${nsmooth}"                         >> $@
	@echo "NCPUS ${NCPUS}"                             >> $@
# # ------------------------------------------------------------------------------
# sanity checks
ifeq ($(gizaln),)
$(warning Giza base directory not defined)
endif
ifeq ($(gizaln.in),)
$(warning No directory for Giza++ training data specified!)
endif