From b282512ffb68b697217d6a36f21417136f631ace Mon Sep 17 00:00:00 2001
From: dhowe <dhowe@cims.nyu.edu>
Date: Sun, 9 Jun 2024 12:33:48 +0100
Subject: [PATCH] remove async call in Conjugator constructor

---
 src/conjugator.js        | 15 +++++++++------
 src/lexicon.js           |  1 +
 src/randgen.js           | 15 ++++++++-------
 src/rita_dict.js         | 15 ++++++---------
 test/conjugator-tests.js |  8 ++++++++
 test/lexicon-tests.js    |  5 +++++
 6 files changed, 37 insertions(+), 22 deletions(-)

diff --git a/src/conjugator.js b/src/conjugator.js
index 2cb2614..7cacb9d 100644
--- a/src/conjugator.js
+++ b/src/conjugator.js
@@ -1,3 +1,5 @@
+import { log } from "@tensorflow/tfjs";
+
 /**
  * @memberof module:rita
  */
@@ -39,12 +41,13 @@ class Conjugator {
   constructor(parent) {
     this.RiTa = parent;
     this._reset();
-    this.RiTa.search({ pos: 'v', limit: -1, minLength: -1 })
-      .then(res => {
-        this.allVerbs = res;
-        this.verbsEndingInE = res.filter(v => v.endsWith("e"));
-        this.verbsEndingInDouble = res.filter(v => /([^])\1$/.test(v));
-      });
+    let { data } = parent.lexicon;
+    this.allVerbs = Object.keys(data).filter(word => data[word][1].split(' ').includes('vb'));
+    //console.log('Conjugator', 'allVerbs', this.allVerbs.slice(0, 100) + '...');
+    // this.allVerbs.slice(0,100).forEach(v => console.log(v, data[v]));
+    // console.log(this.RiTa.searchSync({ pos: 'v', limit: -1, minLength: -1 }).length, 'vs', this.allVerbs.length);
+    this.verbsEndingInE = this.allVerbs.filter(v => v.endsWith("e"));
+    this.verbsEndingInDouble = this.allVerbs.filter(v => /([^])\1$/.test(v));
   }
 
   // TODO: add handling of past tense modals.
diff --git a/src/lexicon.js b/src/lexicon.js
index b34b577..87c0f7e 100644
--- a/src/lexicon.js
+++ b/src/lexicon.js
@@ -424,6 +424,7 @@ class Lexicon {
     if (tpos && tpos.length) {
       opts.pluralize = (tpos === "nns");
       opts.conjugate = (tpos[0] === "v" && tpos.length > 2);
+      // simple pos conversion
       if (tpos[0] === "n") tpos = "nn";
       else if (tpos[0] === "v") tpos = "vb";
       else if (tpos === "r") tpos = "rb";
diff --git a/src/randgen.js b/src/randgen.js
index 5af2ead..0bc9fb4 100644
--- a/src/randgen.js
+++ b/src/randgen.js
@@ -55,7 +55,7 @@ class SeededRandom {
   */
   pselect(probs) {
     if (!probs || !probs.length) throw Error('arg required');
-    let point = this._rndf(), cutoff = 0;
+    let point = Math.random(), cutoff = 0;
     for (let i = 0; i < probs.length - 1; ++i) {
       cutoff += probs[i];
       if (point < cutoff) return i;
@@ -75,12 +75,13 @@ class SeededRandom {
   }
 
   /*
-    Returns a normalised probability distribution (summing to 1) for arbitrary positive weights
-    If temperature is provided this is basically the softmax, otherwise it simple normalisation
-    Temperature parameter: range is between 0 and +Infinity (excluding both).
-    Lower values move the highest-weighted output toward a probability of 1.0.
-    Higher values tend to even out all the probabilities
-  */
+   * Returns a normalised probability distribution (summing to 1) for arbitrary positive weights
+   * If temperature is provided this is basically the softmax, otherwise it is simple normalisation
+   *
+   * Temperature parameter: range is between 0 and +Infinity (excluding both).
+   * Lower values move the highest-weighted output toward a probability of 1.0.
+   * Higher values tend to even out all the probabilities
+   */
   ndist(weights, temp) {
     let probs = [], sum = 0;
     if (!temp) { // no temp here
diff --git a/src/rita_dict.js b/src/rita_dict.js
index d3d5ebd..849fbed 100644
--- a/src/rita_dict.js
+++ b/src/rita_dict.js
@@ -1,4 +1,4 @@
-  
+
 /** @type {object} */
 export default {
   "a": ["ey1", "dt"],
@@ -9680,7 +9680,7 @@ export default {
   "imputation": ["ih-m p-y-ah t-ey1 sh-ah-n", "nn"],
   "impute": ["ih-m-p-y-uw1-t", "vb"],
   "imputed": ["ih-m p-y-uw1 t-ih-d", "vbn jj"],
-  "in": ["ih-n", "in nn rb rp rbr"],
+  "in": ["ih-n", "in rb rp rbr"],
   "inability": ["ih-n ah b-ih1 l-ih t-iy", "nn"],
   "inaccessible": ["ih-n ah-k s-eh1 s-ah b-ah-l", "jj"],
   "inaccuracy": ["ih-n ae1 k-y-er ah s-iy", "nn"],
@@ -13378,7 +13378,7 @@ export default {
   "oust": ["aw1-s-t", "vb"],
   "ousted": ["aw1 s-t-ih-d", "vbn vbd jj"],
   "ouster": ["aw1 s-t-er", "nn"],
-  "out": ["aw1-t", "in jj nn rb rp"],
+  "out": ["aw1-t", "in jj rb rp"],
   "outage": ["aw1 t-ah-jh", "nn"],
   "outback": ["aw1-t b-ae-k", "nn"],
   "outbid": ["aw1-t b-ih-d", "vb vbd vbn vbp"],
@@ -19905,9 +19905,8 @@ export default {
   "tourism": ["t-uh1 r-ih z-ah-m", "nn"],
   "tourist": ["t-uh1 r-ah-s-t", "nn"],
   "tournament": ["t-uh1-r n-ah m-ah-n-t", "nn"],
-  "tourney": ["t-er1 n-iy", "nn"],
   "tout": ["t-aw1-t", "vb nn"],
-  "tow": ["t-ow1", "nn vb"],
+  "tow": ["t-ow1", "vb"],
   "toward": ["t-ah w-ao1-r-d", "in"],
   "towards": ["t-ah w-ao1-r-d-z", "in"],
   "towel": ["t-aw1 ah-l", "nn"],
@@ -19928,7 +19927,6 @@ export default {
   "trachea": ["t-r-ey1 k-iy ah", "nn"],
   "tracing": ["t-r-ey1 s-ih-ng", "vbg jj nn"],
   "track": ["t-r-ae1-k", "nn vbp vb"],
-  "trackage": ["t-r-ae1 k-ih-jh", "nn"],
   "tracked": ["t-r-ae1-k-t", "vbn vbd jj"],
   "tract": ["t-r-ae1-k-t", "nn"],
   "tractor": ["t-r-ae1-k t-er", "nn"],
@@ -19948,7 +19946,7 @@ export default {
   "tragedy": ["t-r-ae1 jh-ah d-iy", "nn"],
   "tragic": ["t-r-ae1 jh-ih-k", "jj nn"],
   "tragically": ["t-r-ae1 jh-ih-k l-iy", "rb"],
-  "tragicomic": ["t-r-ae jh-ih k-aa1 m-ih-k", "jj nn"],
+  "tragicomic": ["t-r-ae jh-ih k-aa1 m-ih-k", "jj"],
   "trail": ["t-r-ey1-l", "nn vbp vb"],
   "trailer": ["t-r-ey1 l-er", "nn"],
   "train": ["t-r-ey1-n", "nn vb vbp"],
@@ -20106,11 +20104,10 @@ export default {
   "trim": ["t-r-ih1-m", "vb jj nn"],
   "trimester": ["t-r-ay m-eh1 s-t-er", "nn"],
   "trimmed": ["t-r-ih1-m-d", "vbn rbr vbd jj"],
-  "trimmer": ["t-r-ih1 m-er", "jjr"],
+  "trimmer": ["t-r-ih1 m-er", "nn jjr"],
   "trinket": ["t-r-ih1-ng k-ah-t", "nn"],
   "trio": ["t-r-iy1 ow", "nn"],
   "trip": ["t-r-ih1-p", "nn vb"],
-  "tripartite": ["t-r-ay p-aa1-r t-ay-t", "jj"],
   "tripe": ["t-r-ay1-p", "nn"],
   "triple": ["t-r-ih1 p-ah-l", "jj nn rb vb"],
   "triplet": ["t-r-ih1 p-l-ah-t", "nn"],
diff --git a/test/conjugator-tests.js b/test/conjugator-tests.js
index d4912e9..398553b 100644
--- a/test/conjugator-tests.js
+++ b/test/conjugator-tests.js
@@ -1,11 +1,18 @@
 import { expect } from 'chai';
 
 import { RiTa } from './index.js';
+import Conjugator from '../src/conjugator.js';
 
 let hasLex = true;
 
 describe('Conjugator', function () {
 
+  it('Should be constructed', function () {
+    let conjugator = new Conjugator(RiTa);
+    expect(conjugator).to.be.an('object');
+    expect(conjugator.verbsEndingInE).to.be.an('array');
+  });
+
   it('Should call pastPart', function () {
 
     expect(RiTa.pastPart("pen")).eq("penned"); // rita #150
@@ -528,6 +535,7 @@ describe('Conjugator', function () {
 
     if (!hasLex) this.skip();
 
+    expect(RiTa.conjugator.unconjugate("accessing")).eq("access");
     expect(RiTa.conjugator.unconjugate("trepanning")).eq("trepan");
 
     // 3rd person singular (regular)
diff --git a/test/lexicon-tests.js b/test/lexicon-tests.js
index fd76355..8b8c92f 100644
--- a/test/lexicon-tests.js
+++ b/test/lexicon-tests.js
@@ -9,6 +9,11 @@ describe('Lexicon', function () {
   this.timeout(10000);
 
   it('Should call hasWord', function () {
+    
+    expect(RiTa.hasWord("accessing")).to.be.true;
+    expect(RiTa.hasWord("Accessing")).to.be.true;
+    expect(RiTa.hasWord("accessed")).to.be.true;
+    expect(RiTa.hasWord("Accessed")).to.be.true;
 
     for (let i = 0; i < 10; i++) {
       expect(RiTa.hasWord(RiTa.randomWord())).to.be.true;