regtest.json5

{
	steps: {
		fst: {
			cmd: '~/langtech/kal/tools/shellscripts/kal-tokenise ~/langtech/kal/tools/tokenisers/tokeniser-disamb-gt-desc.pmhfst | cg-sort',
		},
		pre1: {
			cmd: 'vislcg3 -g ~/langtech/kal/src/cg3/kal-pre1.cg3',
			type: 'cg',
		},
		hybrids: {
			cmd: '~/langtech/kal/tools/shellscripts/kal-hybrid-split ~/langtech/kal/src/fst/generator-gt-desc.hfstol | cg-sort',
		},
		heur: {
			cmd: '~/langtech/kal/tools/shellscripts/kal-spell-heur | cg-sort',
		},
		pre: {
			cmd: '~/langtech/kal/tools/shellscripts/kal-tokenise ~/langtech/kal/tools/tokenisers/tokeniser-disamb-gt-desc.pmhfst | vislcg3 -g ~/langtech/kal/src/cg3/kal-pre1.cg3 | ~/langtech/kal/tools/shellscripts/kal-hybrid-split ~/langtech/kal/src/fst/generator-gt-desc.hfstol | cg-sort',
		},
		'sems-l': {
			cmd: '~/langtech/katersat/apply-sems.py --last',
			type: 'cg',
		},
		'sems-f': {
			cmd: '~/langtech/katersat/apply-sems.py',
			type: 'cg',
		},
		pre2: {
			cmd: 'vislcg3 -g ~/langtech/kal/src/cg3/kal-pre2.cg3',
			type: 'cg',
		},
		morf: {
			cmd: 'vislcg3 -g ~/langtech/kal/src/cg3/disambiguator.cg3',
			type: 'cg',
		},
		'morf-c': {
			cmd: '~/langtech/kal/tools/shellscripts/kal-tokenise ~/langtech/kal/tools/tokenisers/tokeniser-disamb-gt-desc.pmhfst | vislcg3 -g ~/langtech/kal/src/cg3/kal-pre1.cg3 | ~/langtech/kal/tools/shellscripts/kal-hybrid-split ~/langtech/kal/src/fst/generator-gt-desc.hfstol | vislcg3 -g ~/langtech/kal/src/cg3/kal-pre2.cg3 | vislcg3 -g ~/langtech/kal/src/cg3/disambiguator.cg3 | cg-sort',
		},
		lu: {
			cmd: '~/langtech/kal/tools/shellscripts/kal-lu-prefix ~/langtech/kal/src/fst/generator-gt-desc.hfstol | cg-sort',
		},
		syntax: {
			cmd: 'vislcg3 -g ~/langtech/kal/src/cg3/functions.cg3',
			type: 'cg',
		},
		dep: {
			cmd: 'vislcg3 -g ~/langtech/kal/src/cg3/dependency.cg3',
			type: 'cg',
		},
		no2nd: {
			cmd: "perl -wpne 's~ [^/\\s]+/[^/\\s]+~~g; s~ i[A-Z\\d]\\w*~~g; s~\\x{E020}~\\x{20}~g; while(s~( DIRTALE[A-Z]+)\\1~$1~g){}' | cg-sort",
		},
		'no2nd-s': {
			cmd: "perl -wpne 's~ (?!i?Sem)[^/\\s]+/[^/\\s]+~~g; s~ i(?!Sem)[A-Z\\d]\\w*~~g; s~\\x{E020}~\\x{20}~g; while(s~( DIRTALE[A-Z]+)\\1~$1~g){}' | cg-sort",
		},
	},
	pipes: {
		kal: ['fst', 'pre1', 'hybrids', 'pre2', 'morf', 'syntax', 'dep', 'no2nd'],
		fst: ['fst'],
		'kal-sems-l': ['pre', 'sems-l', 'pre2', 'morf', 'syntax', 'dep', 'no2nd-s'],
		'kal-lu': ['morf-c', 'lu', 'syntax', 'dep', 'no2nd'],
		heur: ['fst', 'pre1', 'heur'],
		'gloss-k2d': '~/langtech/gloss/kal2dan/kal2dan.pl --regtest --cmd',
		'gloss-k2e': '~/langtech/gloss/kal2eng/kal2eng.pl --regtest --cmd',
	},
	corpora: {
		all: ['*'],
		gloss: ['KW', 'KWx'],
		heur: ['heur/*'],
		fst: ['fst/*'],
	},
	defaults: {
		pipe: 'kal',
		test: 'kal',
		corpora: ['all'],
		gold: false,
		git: true,
		env: [],
	},
	tests: {
		kal: {
			desc: 'Default Kalaalisut analysis pipe',
		},
		fst: {
			desc: 'FST only',
			pipe: 'fst',
			corpora: ['fst'],
		},
		'kal-lu': {
			desc: 'Split LU/LI/etc to their own token (dynamic corpus)',
			pipe: 'kal-lu',
			env: ['KAL_LU_PREFIX=1'],
			grep: ' (CONJ|ADV)-L',
		},
		'kal-sems-l': {
			desc: 'Semantic annotation, only final semantics',
			pipe: 'kal-sems-l',
		},
		'heur': {
			desc: 'Speller and heuristic analysis of unknown words',
			pipe: 'heur',
			corpora: ['heur'],
		},
		'gloss-k2d': {
			desc: 'Gloss kal2dan',
			pipe: 'gloss-k2d',
			corpora: ['gloss'],
			gold: true,
		},
		'gloss-k2e': {
			desc: 'Gloss kal2eng',
			pipe: 'gloss-k2e',
			corpora: ['gloss'],
			gold: true,
		},
	},
}