Makefile

BABELON_FILES := $(wildcard babelon/hp-*.babelon.tsv)

# Extract the variable part (e.g., pt)
TRANSLATIONS := $(patsubst babelon/hp-%.babelon.tsv, %, $(BABELON_FILES))

list-translations:
	@echo $(TRANSLATIONS)

##### Process translations

sort-%: babelon/hp-%.babelon.tsv babelon/hp-%.synonyms.tsv
	mkdir -p tmp
	{ head -n 2 babelon/hp-$*.synonyms.tsv && tail -n +3 babelon/hp-$*.synonyms.tsv | sort; } > tmp/$*_synonyms_sorted.tsv
	{ head -n 1 babelon/hp-$*.babelon.tsv && tail -n +2 babelon/hp-$*.babelon.tsv | sort; } > tmp/$*_babelon_sorted.tsv
	mv tmp/$*_babelon_sorted.tsv babelon/hp-$*.babelon.tsv
	mv tmp/$*_synonyms_sorted.tsv babelon/hp-$*.synonyms.tsv

clean-%: babelon/hp-%.babelon.tsv babelon/hp-%.synonyms.tsv
	sed -E 's/\t[ ]+/\t/g; s/[ ]+\t/\t/g' babelon/hp-$*.babelon.tsv > temp.tsv && mv temp.tsv babelon/hp-$*.babelon.tsv
	sed -E 's/\t[ ]+/\t/g; s/[ ]+\t/\t/g' babelon/hp-$*.synonyms.tsv > temp.tsv && mv temp.tsv babelon/hp-$*.synonyms.tsv

clean-all:
	$(MAKE) clean-pt clean-de clean-fr clean-pt clean-zh

sort-all:
	$(MAKE) sort-pt sort-de sort-fr sort-pt sort-zh

validate-%: babelon/hp-%.babelon.tsv babelon/hp-%.synonyms.tsv
	@output=$$(tsvalid babelon/hp-$*.synonyms.tsv --skip "W1"); \
	if echo "$$output" | grep -Eq 'E[0-9]+:[ ]'; then \
		echo "Error detected in hp-$*.synonyms.tsv: $$output"; \
		exit 1; \
	fi
	@output=$$(tsvalid babelon/hp-$*.babelon.tsv --skip "W1"); \
	if echo "$$output" | grep -Eq 'E[0-9]+:[ ]'; then \
		echo "Error detected in hp-$*.babelon.tsv: $$output"; \
		exit 1; \
	fi
	babelon convert babelon/hp-$*.babelon.tsv --output-format owl -o tmp/$*-babelon.owl
	robot template --template babelon/hp-$*.synonyms.tsv --output tmp/$*-synonyms.owl

validate-all:
	$(MAKE) $(foreach lang, $(TRANSLATIONS), validate-$(lang))

qc: validate-all
	@echo "All translations are valid"

update: babelon/hp-ja.babelon.tsv
	$(MAKE) sort-all
	$(MAKE) clean-all
	$(MAKE) validate-all

install:
	pip install -U babelon==0.3.4 --break-system-packages

###
prepare_data:
	rm -rf crowdin_data/
	unzip -d crowdin_data/ all_translations.zip

babelon/hp-%.babelon.tsv:
	mkdir -p babelon/
	babelon parse crowdin_data/$*/hpo_notes.xliff -o $@

babelon/hp-ja.babelon.tsv:
	wget "https://raw.githubusercontent.com/ogishima/HPO-Japanese/master/HPO-japanese.alpha.21Jul2023.tsv" -O $@

LANGUAGES=cs nl tr
HP_TRANSLATIONS=$(patsubst %, babelon/hp-%.babelon.tsv, $(LANGUAGES))

# Recipe: Go to https://crowdin.com/project/hpo-translation/translations# and click "Build and Download"
# Safe the downloaded file as tests/data/all_translations.zip
# Run "make prepare_data" to unpack the translations
# Run "make languages" to process them
# Copy the results into hpo/src/ontology/translations
translations: $(HP_TRANSLATIONS)



.PHONY: help
help:
	@echo "$$data"

define data
Usage: make command

----------------------------------------
	Command reference
----------------------------------------

Core commands:
* prepare_data:	Unpack XLIFF files from ZIP container
* translations:	Compute all Babelon translations from the XLIFF files

endef
export data