lingdocs
/
pashto-inflector
mirror of https://github.com/lingdocs/pashto-inflector.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

more

This commit is contained in:
adueck 2023-08-14 01:01:34 +04:00
parent 7f031c4a71
commit 7fab8c1851
7 changed files with 192 additions and 207 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
src/components/src/vp-explorer/AllTensesDisplay.tsx
View File

@ -151,6 +151,7 @@ function AllTensesDisplay({
<ChartDisplay
verb={verb}
objectNP={
VS.isCompound === "dynamic" &&
objectNP?.selection.type === "noun"
? {
...objectNP,

25
src/lib/src/new-verb-engine/render-verb.ts
View File

@ -7,6 +7,7 @@ import {
} from "../misc-helpers";
import { fmapSingleOrLengthOpts } from "../fp-ps";
import { concatPsString, getLength } from "../p-text-helpers";
import { zipWith } from "rambda";
import {
presentEndings,
pastEndings,
@ -362,16 +363,20 @@ function ensure3rdPast(
];
}
if (verb.entry.tppp && verb.entry.tppf) {
const tip = removeAccents(
verb.entry.separationAtP !== undefined
? makePsString(
verb.entry.tppp.slice(verb.entry.separationAtP),
verb.entry.tppf.slice(verb.entry.separationAtF)
)
: makePsString(verb.entry.tppp, verb.entry.tppf)
);
const aTip = aspect === "imperfective" ? accentOnNFromEnd(tip, 0) : tip;
return [aTip];
const tppp = verb.entry.tppp.split(",").map((x) => x.trim());
const tppf = verb.entry.tppf.split(",").map((x) => x.trim());
const tpps = zipWith((p, f) => ({ p, f }), tppp, tppf);
return tpps.map(({ p, f }) => {
const tip = removeAccents(
verb.entry.separationAtP !== undefined
? makePsString(
p.slice(verb.entry.separationAtP),
f.slice(verb.entry.separationAtF)
)
: makePsString(p, f)
);
return aspect === "imperfective" ? accentOnNFromEnd(tip, 0) : tip;
});
// if it ends in a consonant, the special form will also have another
// variation ending with a ه - u
// const endsInAConsonant = (pashtoConsonants.includes(tip.p.slice(-1)) || tip.f.slice(-1) === "w");

18
src/lib/src/parsing/lookup.tsx
View File

@ -31,9 +31,23 @@ export function lookup(s: Partial<T.DictionaryEntry>): T.DictionaryEntry[] {
return nounsAdjs.filter((e) => e[key] === value) as T.DictionaryEntry[];
}
export function shouldCheckTpp(s: string): boolean {
return (
["د", "ړ", "ت", "ځ", "و", "ډ", "ڼ", "ن", "ه"].includes(s.slice(-1)) ||
["ست", "ښت"].includes(s.slice(-2)) ||
["ښود"].includes(s.slice(-3))
);
}
export function verbLookup(input: string): T.VerbEntry[] {
const s = input.slice(0, -1);
// check endings TODO: ONLY LOOKUP THE VERB POSSIBILITIES IF IT HAS A LEGITIMATE ENDING
// if theres no legit verb ending and no tpp possibilities, just return an empty array
const sWoutOo = s.startsWith("و") ? s.slice(1) : undefined;
const checkTpp = shouldCheckTpp(input);
const inputWoutOo =
checkTpp && input.startsWith("و") ? input.slice(1) : undefined;
// TODO: don't do the slice of and checking for useless things when you have a NON verb ending (like with the tpp)
if (s.endsWith("ېږ")) {
return verbs.filter(
sWoutOo
@ -43,12 +57,14 @@ export function verbLookup(input: string): T.VerbEntry[] {
entry.p
) ||
[s, sWoutOo].includes(entry.p) ||
(checkTpp && [input, inputWoutOo].includes(entry.tppp)) ||
(entry.psp && [s, sWoutOo].includes(entry.psp)) ||
entry.prp === s ||
entry.ssp === s
: ({ entry }) =>
entry.p.slice(0, -1) === s ||
entry.p === s.slice(0, -1) + "دل" ||
(checkTpp && [input, inputWoutOo].includes(entry.tppp)) ||
entry.p === s ||
entry.psp === s ||
entry.prp === s ||
@ -63,6 +79,7 @@ export function verbLookup(input: string): T.VerbEntry[] {
[s, sWoutOo].includes(entry.p.slice(0, -3)) ||
[s, sWoutOo].includes(entry.p) ||
(entry.psp && [s, sWoutOo].includes(entry.psp)) ||
(checkTpp && [input, inputWoutOo].includes(entry.tppp)) ||
entry.prp === s ||
entry.ssp === s ||
(entry.separationAtP &&
@ -73,6 +90,7 @@ export function verbLookup(input: string): T.VerbEntry[] {
// for short intransitive forms
entry.p.slice(0, -3) === s ||
entry.p === s ||
(checkTpp && [input, inputWoutOo].includes(entry.tppp)) ||
entry.psp === s ||
entry.prp === s ||
entry.ssp === s ||

21
src/lib/src/parsing/parse-ba.ts Normal file
View File

@ -0,0 +1,21 @@
import * as T from "../../../types";
export function parseBa(
tokens: Readonly<T.Token[]>
): T.ParseResult<{ type: "ba" }>[] {
if (!tokens.length) {
return [];
}
const [first, ...rest] = tokens;
if (first.s === "به") {
return [
{
body: {
type: "ba",
},
errors: [],
tokens: rest,
},
];
} else return [];
}

221
src/lib/src/parsing/parse-verb.test.ts
View File

@ -15,6 +15,7 @@ const kenaastul = wordQuery("کېناستل", "verb");
const prexodul = wordQuery("پرېښودل", "verb");
const xodul = wordQuery("ښودل", "verb");
const kexodul = wordQuery("کېښودل", "verb");
const katul = wordQuery("کتل", "verb");
const tests: {
label: string;
@ -376,6 +377,7 @@ const tests: {
},
],
},
// TODO!! THESE COULD ALSO BE MALE
{
input: "لیده",
output: [
@ -402,6 +404,7 @@ const tests: {
},
],
},
// BUT NOT THIS ONE
{
input: "ولیدله",
output: [
@ -471,6 +474,33 @@ const tests: {
},
],
},
{
input: "خوړ",
output: [
{
ph: undefined,
root: {
persons: [T.Person.ThirdSingMale],
aspects: ["imperfective", "perfective"],
},
verb: khorul,
},
],
},
// TODO: should do کاته as well... what to do, have multiple tpp forms ? YES!
{
input: "وکوت",
output: [
{
ph: "و",
root: {
persons: [T.Person.ThirdSingMale],
aspects: ["perfective"],
},
verb: katul,
},
],
},
],
},
{
@ -633,194 +663,3 @@ tests.forEach(({ label, cases }) => {
});
});
});
const b = [
[
undefined,
{
info: {
aspect: "imperfective",
base: "stem",
type: "verb",
verb: {
entry: {
c: "v. trans.",
e: "to put, to put down, to set in place",
ec: "put,puts,putting,put,put",
f: "kexodul",
g: "kexodul",
i: 11193,
noOo: true,
p: "کېښودل",
psf: "Gd",
psp: "ږد",
r: 4,
separationAtF: 2,
separationAtP: 2,
ssf: "kéGd",
ssp: "کېږد",
ts: 1527812284,
},
},
},
person: 6,
type: "VB",
},
],
[
undefined,
{
info: {
aspect: "imperfective",
base: "stem",
type: "verb",
verb: {
entry: {
c: "v. trans.",
e: "to put, to put down, to set in place",
ec: "put,puts,putting,put,put",
f: "kexodul",
g: "kexodul",
i: 11193,
noOo: true,
p: "کېښودل",
psf: "Gd",
psp: "ږد",
r: 4,
separationAtF: 2,
separationAtP: 2,
ssf: "kéGd",
ssp: "کېږد",
ts: 1527812284,
},
},
},
person: 7,
type: "VB",
},
],
[
undefined,
{
info: {
aspect: "perfective",
base: "stem",
type: "verb",
verb: {
entry: {
c: "v. trans.",
e: "to leave, abandon, forsake, let go, allow",
ec: "abandon",
f: "prexodúl",
g: "prexodul",
i: 2516,
noOo: true,
p: "پرېښودل",
psf: "preGd",
psp: "پرېږد",
r: 4,
separationAtF: 3,
separationAtP: 3,
ts: 1527815190,
},
},
},
person: 6,
type: "VB",
},
],
[
undefined,
{
info: {
aspect: "perfective",
base: "stem",
type: "verb",
verb: {
entry: {
c: "v. trans.",
e: "to leave, abandon, forsake, let go, allow",
ec: "abandon",
f: "prexodúl",
g: "prexodul",
i: 2516,
noOo: true,
p: "پرېښودل",
psf: "preGd",
psp: "پرېږد",
r: 4,
separationAtF: 3,
separationAtP: 3,
ts: 1527815190,
},
},
},
person: 7,
type: "VB",
},
],
[
{ s: "کې", type: "PH" },
{
info: {
aspect: "perfective",
base: "stem",
type: "verb",
verb: {
entry: {
c: "v. trans.",
e: "to put, to put down, to set in place",
ec: "put,puts,putting,put,put",
f: "kexodul",
g: "kexodul",
i: 11193,
noOo: true,
p: "کېښودل",
psf: "Gd",
psp: "ږد",
r: 4,
separationAtF: 2,
separationAtP: 2,
ssf: "kéGd",
ssp: "کېږد",
ts: 1527812284,
},
},
},
person: 6,
type: "VB",
},
],
[
{ s: "کې", type: "PH" },
{
info: {
aspect: "perfective",
base: "stem",
type: "verb",
verb: {
entry: {
c: "v. trans.",
e: "to put, to put down, to set in place",
ec: "put,puts,putting,put,put",
f: "kexodul",
g: "kexodul",
i: 11193,
noOo: true,
p: "کېښودل",
psf: "Gd",
psp: "ږد",
r: 4,
separationAtF: 2,
separationAtP: 2,
ssf: "kéGd",
ssp: "کېږد",
ts: 1527812284,
},
},
},
person: 7,
type: "VB",
},
],
];

68
src/lib/src/parsing/parse-verb.ts
View File

@ -1,5 +1,27 @@
import * as T from "../../../types";
// third persion idosyncratic
// if it ends in a dental or ه - look for tttp
//
// if not having tttp
// automatic things: (with blank or u)
// ېد ست ښت
// ښود
//
// ول - اوه
// وېشه ?
// test ګالو ❌ vs ګاللو ✅
// واخیست / واخیسته / واخیستلو
// ولید // ولیده // ولیدو
//
// ووت / واته
//
// also write the rules for the third pers sing endings in the grammar
// multiple third pers sing options
export function parseVerb(
tokens: Readonly<T.Token[]>,
verbLookup: (s: string) => T.VerbEntry[]
@ -258,6 +280,52 @@ function matchVerbs(
});
});
}
const tppMatches = {
imperfective: entries.filter(
({ entry: e }) => !e.c.includes("comp") && s === e.tppp
),
perfective: entries.reduce<
{ ph: string | undefined; entry: T.VerbEntry }[]
>((acc, entry) => {
const e = entry.entry;
const sNoOo = s.startsWith("و") && s.slice(1);
if (sNoOo && sNoOo === e.tppp) {
return [
...acc,
{
ph: "و",
entry,
},
];
} else if (s === e.tppp) {
return [
...acc,
{
ph: undefined,
entry,
},
];
}
return acc;
}, []),
};
Object.entries(tppMatches).forEach(([aspect, entries]) => {
entries.forEach((verb) => {
w.push([
"ph" in verb && verb.ph ? { type: "PH", s: verb.ph } : undefined,
{
type: "VB",
person: T.Person.ThirdSingMale,
info: {
type: "verb",
aspect: aspect as T.Aspect,
base: "root",
verb: "ph" in verb ? verb.entry : verb,
},
},
]);
});
});
return w;
}

45
src/lib/src/parsing/parse-vp.ts
View File

@ -8,6 +8,7 @@ import {
} from "../phrase-building/blocks-utils";
import { vEntry } from "../new-verb-engine/rs-helpers";
import { getPersonFromNP, isThirdPerson } from "../phrase-building/vp-tools";
import { parseBa } from "./parse-ba";
// to hide equatives type-doubling issue
const kedulStat = vEntry({
ts: 1581086654898,
@ -40,16 +41,44 @@ export function parseVP(
}
// how to make this into a nice pipeline... 🤔
const NP1 = parseNP(tokens, lookup).filter(({ errors }) => !errors.length);
const ba = bindParseResult(NP1, (tokens, np1) => {
const b = parseBa(tokens);
if (!b.length) {
return [
{
tokens,
body: {
np1,
ba: false,
},
errors: [],
},
];
} else {
return b.map(({ tokens, errors }) => ({
body: {
np1,
ba: true,
},
errors,
tokens,
}));
}
});
const NP2 = bindParseResult<
{
inflected: boolean;
selection: T.NPSelection;
np1: {
inflected: boolean;
selection: T.NPSelection;
};
ba: boolean;
},
{
np1: {
inflected: boolean;
selection: T.NPSelection;
};
ba: boolean;
np2:
| {
inflected: boolean;
@ -57,7 +86,7 @@ export function parseVP(
}
| undefined;
}
>(NP1, (tokens, np1) => {
>(ba, (tokens, { np1, ba }) => {
const np2s = parseNP(tokens, lookup);
if (!np2s.length) {
const r: T.ParseResult<{
@ -65,6 +94,7 @@ export function parseVP(
inflected: boolean;
selection: T.NPSelection;
};
ba: boolean;
np2: undefined;
}>[] = [
{
@ -72,6 +102,7 @@ export function parseVP(
body: {
np1,
np2: undefined,
ba,
},
errors: [],
},
@ -83,6 +114,7 @@ export function parseVP(
body: {
np1,
np2: p.body,
ba,
},
errors: p.errors,
}));
@ -96,12 +128,13 @@ export function parseVP(
np2: nps.np2,
v: p.body,
np1: nps.np1,
ba: nps.ba,
},
errors: p.errors,
}));
}).filter(({ errors }) => !errors.length);
// TODO: be able to bind mulitple vals
return bindParseResult(vb, (tokens, { np1, np2, v: [ph, v] }) => {
return bindParseResult(vb, (tokens, { np1, np2, v: [ph, v], ba }) => {
const w: T.ParseResult<T.VPSelectionComplete>[] = [];
if (v.info.type === "equative") {
throw new Error("not yet implemented");
@ -143,7 +176,7 @@ export function parseVP(
canChangeTransitivity: false,
canChangeStatDyn: false,
negative: false,
tense: getTenseFromRootsStems(false, v.info.base, v.info.aspect),
tense: getTenseFromRootsStems(ba, v.info.base, v.info.aspect),
canChangeVoice: true,
isCompound: false,
voice: "active",
@ -240,7 +273,7 @@ export function parseVP(
canChangeTransitivity: false,
canChangeStatDyn: false,
negative: false,
tense: getTenseFromRootsStems(false, v.info.base, v.info.aspect),
tense: getTenseFromRootsStems(ba, v.info.base, v.info.aspect),
canChangeVoice: true,
isCompound: false,
voice: "active",