Merge branch 'main' of github.com:toaq/kuna

src/dictionary.ts

@@ -34,7 +34,11 @@ export const nonVerbTypes = [
 	'modality with complement',
 	'plural coordinator',
 	'polarity',
-	'prefix',
+	'prefix', // verb-to-verb
+	'prefix aspect',
+	'prefix conjunctionizer', // na-
+	'prefix pronoun', // hu-
+	'prefix tense',
 	'preposition',
 	'pronoun',
 	'retroactive cleft',
@@ -114,7 +118,17 @@ export function initializeDictionary(): void {
 				});
 			}
 		}
+
+		// We'll assume "prefix" is a verb-to-verb prefix, and make some
+		// sub-types for special prefixes.
+		if (e.toaq == 'hu-') {
+			e.type = 'prefix pronoun';
+		}
+		if (e.toaq == 'na-') {
+			e.type = 'prefix conjunctionizer';
+		}
 		dictionary.set(e.toaq.toLowerCase(), e);
+
 		if (e.type === 'determiner') {
 			const oid = inTone(e.toaq, Tone.T4);
 			dictionary.set(oid, {
@@ -161,6 +175,26 @@ export function initializeDictionary(): void {
 				type: 'modality with complement',
 			});
 		}
+
+		if (e.type === 'aspect') {
+			const prefix = e.toaq + '-';
+			dictionary.set(prefix, {
+				toaq: prefix,
+				english: e.english,
+				gloss: e.gloss,
+				type: 'prefix aspect',
+			});
+		}
+
+		if (e.type === 'tense') {
+			const prefix = e.toaq + '-';
+			dictionary.set(prefix, {
+				toaq: prefix,
+				english: e.english,
+				gloss: e.gloss,
+				type: 'prefix tense',
+			});
+		}
 	}
 
 	dictionary.set('◌́', {

src/english.ts

@@ -21,11 +21,21 @@ function leafToEnglish(leaf: Tree): string {
 	return new Glosser(true).glossWord(leafText(leaf));
 }
 
+function verbToEnglish(tree: Tree): string {
+	if ('word' in tree) {
+		return leafToEnglish(tree);
+	} else if ('left' in tree) {
+		return verbToEnglish(tree.left) + verbToEnglish(tree.right);
+	} else {
+		throw new Error('weird verb');
+	}
+}
+
 function serialToEnglish(serial: Tree): string {
 	if ('word' in serial && serial.word === 'covert') return '';
 	if (serial.label !== '*Serial') throw new Error('non-*Serial serial');
 	if (!('children' in serial)) throw new Error('non-Rose serial');
-	return serial.children.map(x => leafToEnglish(x)).join('-');
+	return serial.children.map(x => verbToEnglish(x)).join('-');
 }
 
 class ClauseTranslator {
@@ -37,7 +47,9 @@ class ClauseTranslator {
 	toaqAspect: string = 'tam';
 	negative: boolean = false;
 	subject?: string = undefined;
+	earlyAdjuncts: string[] = [];
 	objects: string[] = [];
+	lateAdjuncts: string[] = [];
 	modals: string[] = [];
 	constructor(toaqSpeechAct?: string) {
 		this.toaqSpeechAct = toaqSpeechAct;
@@ -56,11 +68,24 @@ class ClauseTranslator {
 			if ('children' in node) {
 				if (node.label !== '*𝘷P') throw new Error('non-*𝘷P Rose');
 				this.verb = serialToEnglish(node.children[0]);
-				if (node.children[1]) {
-					this.subject = treeToEnglish(node.children[1]);
-				}
-				for (let i = 2; i < node.children.length; i++) {
-					this.objects.push(treeToEnglish(node.children[i]));
+				let late = false;
+				for (let i = 1; i < node.children.length; i++) {
+					const child = node.children[i];
+					const english = treeToEnglish(child);
+					if (child.label === 'AdjunctP') {
+						if (late) {
+							this.lateAdjuncts.push(english);
+						} else {
+							this.earlyAdjuncts.push(english);
+						}
+					} else {
+						if (this.subject) {
+							this.objects.push(english);
+						} else {
+							this.subject = english;
+						}
+						late = true;
+					}
 				}
 				break;
 			} else if ('left' in node) {
@@ -179,19 +204,23 @@ class ClauseTranslator {
 				tense,
 				aspect,
 				auxiliary,
+				...this.earlyAdjuncts,
 				this.subject ?? '',
 				this.verb ?? '',
 				...this.objects,
+				...this.lateAdjuncts,
 			];
 		} else {
 			order = [
 				complementizer,
+				...this.earlyAdjuncts,
 				this.subject ?? '',
 				tense,
 				aspect,
 				auxiliary ?? '',
 				this.verb ?? '',
 				...this.objects,
+				...this.lateAdjuncts,
 			];
 		}
 

src/grammar.ts

@@ -10,6 +10,7 @@ declare var conjunction: any;
 declare var conjunction_in_t1: any;
 declare var conjunction_in_t4: any;
 declare var aspect: any;
+declare var prefix_aspect: any;
 declare var topic_marker: any;
 declare var complementizer: any;
 declare var subordinating_complementizer: any;
@@ -25,11 +26,13 @@ declare var text_quote: any;
 declare var modality: any;
 declare var modality_with_complement: any;
 declare var cleft_verb: any;
+declare var prefix: any;
 declare var plural_coordinator: any;
 declare var illocution: any;
 declare var polarity: any;
 declare var word_quote: any;
 declare var tense: any;
+declare var prefix_tense: any;
 declare var end_quote: any;
 declare var predicate: any;
 declare var object_incorporating_verb: any;
@@ -46,6 +49,8 @@ const {
     makeCovertLeaf,
 	makeLeaf,
 	makeOptLeaf,
+	makePrefixLeaf,
+	makePrefixP,
 	makeRose,
 	makeRose2,
 	makeSerial,
@@ -160,8 +165,10 @@ const grammar: Grammar = {
     {"name": "CPsub1", "symbols": ["CPsub"], "postprocess": id},
     {"name": "CPsub1", "symbols": ["CPsub", "Conjunction", "CPsub1"], "postprocess": makeConn},
     {"name": "T1", "symbols": ["T"], "postprocess": id},
+    {"name": "T1", "symbols": ["T_prefix"], "postprocess": id},
     {"name": "T1", "symbols": ["T", "Conjunction", "T1"], "postprocess": makeConn},
     {"name": "Asp1", "symbols": ["Asp"], "postprocess": id},
+    {"name": "Asp1", "symbols": ["Asp_prefix"], "postprocess": id},
     {"name": "Asp1", "symbols": ["Asp", "Conjunction", "Asp1"], "postprocess": makeConn},
     {"name": "AdjunctP1", "symbols": ["AdjunctP"], "postprocess": id},
     {"name": "AdjunctP1", "symbols": ["AdjunctP", "Conjunction", "AdjunctP1"], "postprocess": makeConn},
@@ -172,6 +179,7 @@ const grammar: Grammar = {
     {"name": "Vlast", "symbols": ["Verblike"], "postprocess": id},
     {"name": "V1", "symbols": ["Verblike"], "postprocess": id},
     {"name": "V1", "symbols": ["Verblike", "ConjunctionT1", "V1"], "postprocess": makeConn},
+    {"name": "Verblike", "symbols": ["Prefix", "Verblike"], "postprocess": makePrefixP},
     {"name": "Verblike", "symbols": ["V"], "postprocess": id},
     {"name": "Verblike", "symbols": ["ShuP"], "postprocess": id},
     {"name": "ShuP", "symbols": ["Shu", "Word"], "postprocess": makeBranch('shuP')},
@@ -185,6 +193,7 @@ const grammar: Grammar = {
     {"name": "ConjunctionT1", "symbols": [(lexer.has("conjunction_in_t1") ? {type: "conjunction_in_t1"} : conjunction_in_t1)], "postprocess": makeLeaf('&')},
     {"name": "ConjunctionT4", "symbols": [(lexer.has("conjunction_in_t4") ? {type: "conjunction_in_t4"} : conjunction_in_t4)], "postprocess": makeLeaf('&')},
     {"name": "Asp", "symbols": [(lexer.has("aspect") ? {type: "aspect"} : aspect)], "postprocess": makeLeaf('Asp')},
+    {"name": "Asp_prefix", "symbols": [(lexer.has("prefix_aspect") ? {type: "prefix_aspect"} : prefix_aspect)], "postprocess": makeLeaf('Asp')},
     {"name": "Bi", "symbols": [(lexer.has("topic_marker") ? {type: "topic_marker"} : topic_marker)], "postprocess": makeLeaf('Topic')},
     {"name": "C", "symbols": [(lexer.has("complementizer") ? {type: "complementizer"} : complementizer)], "postprocess": makeLeaf('C')},
     {"name": "Copt$ebnf$1", "symbols": ["C"], "postprocess": id},
@@ -206,6 +215,7 @@ const grammar: Grammar = {
     {"name": "Modal", "symbols": [(lexer.has("modality") ? {type: "modality"} : modality)], "postprocess": makeLeaf('Modal')},
     {"name": "ModalT4", "symbols": [(lexer.has("modality_with_complement") ? {type: "modality_with_complement"} : modality_with_complement)], "postprocess": makeLeaf('Modal')},
     {"name": "Na", "symbols": [(lexer.has("cleft_verb") ? {type: "cleft_verb"} : cleft_verb)], "postprocess": makeLeaf('𝘷')},
+    {"name": "Prefix", "symbols": [(lexer.has("prefix") ? {type: "prefix"} : prefix)], "postprocess": makePrefixLeaf},
     {"name": "Roi", "symbols": [(lexer.has("plural_coordinator") ? {type: "plural_coordinator"} : plural_coordinator)], "postprocess": makeLeaf('&')},
     {"name": "SA", "symbols": [(lexer.has("illocution") ? {type: "illocution"} : illocution)], "postprocess": makeLeaf('SA')},
     {"name": "SAopt$ebnf$1", "symbols": ["SA"], "postprocess": id},
@@ -214,6 +224,7 @@ const grammar: Grammar = {
     {"name": "Sigma", "symbols": [(lexer.has("polarity") ? {type: "polarity"} : polarity)], "postprocess": makeLeaf('Σ')},
     {"name": "Shu", "symbols": [(lexer.has("word_quote") ? {type: "word_quote"} : word_quote)], "postprocess": makeLeaf('shu')},
     {"name": "T", "symbols": [(lexer.has("tense") ? {type: "tense"} : tense)], "postprocess": makeLeaf('T')},
+    {"name": "T_prefix", "symbols": [(lexer.has("prefix_tense") ? {type: "prefix_tense"} : prefix_tense)], "postprocess": makeLeaf('T')},
     {"name": "Teo", "symbols": [(lexer.has("end_quote") ? {type: "end_quote"} : end_quote)], "postprocess": makeLeaf('teo')},
     {"name": "Text", "symbols": ["Fragment"], "postprocess": id},
     {"name": "V", "symbols": [(lexer.has("predicate") ? {type: "predicate"} : predicate)], "postprocess": makeLeaf('V')},

src/semantics/denote.ts

@@ -172,7 +172,7 @@ const fi = λ(['v', 't'], [], c =>
 );
 
 function denoteAspect(toaq: string): Expr {
-	switch (toaq) {
+	switch (toaq.replace(/-$/, '')) {
 		case 'tam':
 			return tam;
 		case 'chum':
@@ -233,7 +233,7 @@ const jela = λ(['i', 't'], [], c =>
 );
 
 function denoteTense(toaq: string): Expr {
-	switch (toaq) {
+	switch (toaq.replace(/-$/, '')) {
 		case 'naı':
 			return nai;
 		case 'pu':

src/toaq.ne

@@ -13,6 +13,8 @@ const {
     makeCovertLeaf,
 	makeLeaf,
 	makeOptLeaf,
+	makePrefixLeaf,
+	makePrefixP,
 	makeRose,
 	makeRose2,
 	makeSerial,
@@ -143,8 +145,10 @@ DP1 -> DP Roi DP1 {% makeConn %}
 CPsub1 -> CPsub {% id %}
 CPsub1 -> CPsub Conjunction CPsub1 {% makeConn %}
 T1 -> T {% id %}
+T1 -> T_prefix {% id %}
 T1 -> T Conjunction T1 {% makeConn %}
 Asp1 -> Asp {% id %}
+Asp1 -> Asp_prefix {% id %}
 Asp1 -> Asp Conjunction Asp1 {% makeConn %}
 AdjunctP1 -> AdjunctP {% id %}
 AdjunctP1 -> AdjunctP Conjunction AdjunctP1 {% makeConn %}
@@ -155,6 +159,7 @@ Vlast -> Verblike ConjunctionT1 Vlast {% makeConn %}
 Vlast -> Verblike {% id %}
 V1 -> Verblike {% id %}
 V1 -> Verblike ConjunctionT1 V1 {% makeConn %}
+Verblike -> Prefix Verblike {% makePrefixP %}
 Verblike -> V {% id %}
 Verblike -> ShuP {% id %}
 ShuP -> Shu Word {% makeBranch('shuP') %}
@@ -169,6 +174,7 @@ Conjunction -> %conjunction {% makeLeaf('&') %}
 ConjunctionT1 -> %conjunction_in_t1 {% makeLeaf('&') %}
 ConjunctionT4 -> %conjunction_in_t4 {% makeLeaf('&') %}
 Asp -> %aspect {% makeLeaf('Asp') %}
+Asp_prefix -> %prefix_aspect {% makeLeaf('Asp') %}
 Bi -> %topic_marker {% makeLeaf('Topic') %}
 C -> %complementizer {% makeLeaf('C') %}
 Copt -> C:? {% makeOptLeaf('C') %}
@@ -186,12 +192,14 @@ Mo -> %text_quote {% makeLeaf('mo') %}
 Modal -> %modality {% makeLeaf('Modal') %}
 ModalT4 -> %modality_with_complement {% makeLeaf('Modal') %}
 Na -> %cleft_verb {% makeLeaf('𝘷') %}
+Prefix -> %prefix {% makePrefixLeaf %}
 Roi -> %plural_coordinator {% makeLeaf('&') %}
 SA -> %illocution {% makeLeaf('SA') %}
 SAopt -> SA:? {% makeOptLeaf('SA') %}
 Sigma -> %polarity {% makeLeaf('Σ') %}
 Shu -> %word_quote {% makeLeaf('shu') %}
 T -> %tense {% makeLeaf('T') %}
+T_prefix -> %prefix_tense {% makeLeaf('T') %}
 Teo -> %end_quote {% makeLeaf('teo') %}
 # TODO: multiple-fragment quotes?
 Text -> Fragment {% id %}

src/tokenize.ts

@@ -97,10 +97,13 @@ export class ToaqTokenizer {
 	reset(text: string, _info?: {}): void {
 		this.tokens = [];
 		this.pos = 0;
-		for (const m of [...text.matchAll(/[\p{L}\p{N}\p{Diacritic}]+-?/gu)]) {
+		for (const m of [...text.matchAll(/[\p{L}\p{N}\p{Diacritic}-]+/gu)]) {
 			const { prefixes, root } = splitPrefixes(m[0]);
 			for (const tokenText of [...prefixes.map(p => p + '-'), root]) {
 				const lemmaForm = clean(tokenText);
+				if (!lemmaForm) {
+					throw new Error('empty token at ' + m.index);
+				}
 				const exactEntry = dictionary.get(lemmaForm);
 
 				if (exactEntry) {

src/tree.ts

@@ -21,6 +21,12 @@ export type Label =
 	| 'AdjunctP'
 	| 'Asp'
 	| 'AspP'
+	| 'be'
+	| 'beP'
+	| 'bu'
+	| 'buP'
+	| 'buq'
+	| 'buqP'
 	| 'C'
 	| 'Crel'
 	| 'CP'
@@ -29,12 +35,16 @@ export type Label =
 	| 'DP'
 	| 'EvA'
 	| 'EvAP'
+	| 'ge'
+	| 'geP'
 	| 'Interjection'
 	| 'InterjectionP'
 	| 'mı'
 	| 'mıP'
 	| 'Modal'
 	| 'ModalP'
+	| 'mu'
+	| 'muP'
 	| 'n'
 	| 'nP'
 	| 'SA'
@@ -94,7 +104,11 @@ export function containsWords(
 }
 
 export function isQuestion(tree: Tree): boolean {
-	return containsWords(tree, ['hí', 'rí', 'rı', 'rî', 'ma', 'tıo'], ['CP']);
+	return containsWords(
+		tree,
+		['hí', 'rí', 'rı', 'rî', 'ma', 'tıo', 'hıa'],
+		['CP'],
+	);
 }
 
 export interface Leaf {
@@ -228,6 +242,8 @@ export function makeOptLeaf(label: Label) {
 	};
 }
 
+const arityPreservingVerbPrefixes: Label[] = ['buP', 'muP', 'buqP', 'geP'];
+
 function getFrame(verb: Tree): string {
 	if ('word' in verb) {
 		if (verb.word === 'covert') throw new Error('covert verb?');
@@ -246,6 +262,10 @@ function getFrame(verb: Tree): string {
 		return 'c';
 	} else if (verb.label === 'EvAP') {
 		return 'c';
+	} else if (verb.label === 'beP') {
+		return 'c';
+	} else if (arityPreservingVerbPrefixes.includes(verb.label)) {
+		return getFrame((verb as Branch<Tree>).right);
 	} else {
 		throw new Error('weird nonverb: ' + verb.label);
 	}
@@ -379,3 +399,18 @@ export function makeSigmaT1ModalvP([sigma, modal, tp]: [Tree, Tree, Tree]) {
 		right: makeT1ModalvP([modal, tp]),
 	};
 }
+
+export function makePrefixLeaf([token]: [ToaqToken]) {
+	return {
+		label: bare(token.value).replace(/-$/, ''),
+		word: makeWord([token]),
+	};
+}
+
+export function makePrefixP([prefix, verb]: [Tree, Tree]) {
+	return {
+		label: prefix.label + 'P',
+		left: prefix,
+		right: verb,
+	};
+}