feat(web): check for low-probability exact + exact-key correction matches 📚

common/models/templates/src/trie-model.ts

@@ -109,7 +109,7 @@ class Traversal implements LexiconTraversal {
 
     // Split into individual code units.
     let steps = char.split('');
-    let traversal: ReturnType<Traversal["_child"]> = this;
+    let traversal: Traversal | undefined = this;
 
     while(steps.length > 0 && traversal) {
       const step: string = steps.shift()!;
@@ -148,6 +148,10 @@ class Traversal implements LexiconTraversal {
 
   *children(): Generator<{char: USVString, traversal: () => LexiconTraversal}> {
     let root = this.root;
+
+    // We refer to the field multiple times in this method, and it doesn't change.
+    // This also assists minification a bit, since we can't minify when re-accessing
+    // through `this.`.
     const totalWeight = this.totalWeight;
 
     if(root.type == 'internal') {
@@ -220,11 +224,10 @@ class Traversal implements LexiconTraversal {
   }
 
   get entries() {
-    const totalWeight = this.totalWeight;
-    const entryMapper = function(value: Entry) {
+    const entryMapper = (value: Entry) => {
       return {
         text: value.content,
-        p: value.weight / totalWeight
+        p: value.weight / this.totalWeight
       }
     }
 
@@ -425,7 +428,7 @@ interface Entry {
  * Wrapper class for the trie and its nodes.
  */
 class Trie {
-  private root: Node;
+  public readonly root: Node;
   /** The total weight of the entire trie. */
   readonly totalWeight: number;
   /**
@@ -451,9 +454,26 @@ class Trie {
    * @param prefix
    */
   lookup(prefix: string): TextWithProbability[] {
-    let searchKey = this.toKey(prefix);
-    let rootTraversal = this.traverseFromRoot().child(searchKey);
-    return rootTraversal ? getSortedResults(rootTraversal) : [];
+    const searchKey = this.toKey(prefix);
+    const rootTraversal = this.traverseFromRoot().child(searchKey);
+
+    if(!rootTraversal) {
+      return [];
+    }
+
+    const directEntries = rootTraversal.entries;
+    // `Set` requires Chrome 38+, which is more recent than Chrome 35.
+    const directSet: Record<string, string> = {};
+    for(const entry of directEntries) {
+      directSet[entry.text] = entry.text;
+    }
+
+    const bestEntries = getSortedResults(rootTraversal);
+    const deduplicated = bestEntries.filter((entry) => !directSet[entry.text]);
+
+    // Any entries directly hosted on the current node should get full display
+    // priority over anything from its descendants.
+    return directEntries.concat(deduplicated);
   }
 
   /**

common/models/templates/test/test-trie-traversal.js

@@ -13,6 +13,12 @@ var smpForUnicode = function(code){
   return String.fromCharCode(H, L);
 }
 
+// Prob:  entry weight / total weight
+// "the" is the highest-weighted word in the fixture.
+const PROB_OF_THE     = 1000 / 500500;
+const PROB_OF_TRUE    =  607 / 500500;
+const PROB_OF_TROUBLE =  267 / 500500;
+
 describe('Trie traversal abstractions', function() {
   it('root-level iteration over child nodes', function() {
     var model = new TrieModel(jsonFixture('tries/english-1000'));
@@ -39,10 +45,6 @@ describe('Trie traversal abstractions', function() {
   it('traversal with simple internal nodes', function() {
     var model = new TrieModel(jsonFixture('tries/english-1000'));
 
-    // Prob:  entry weight / total weight
-    // "the" is the highest-weighted word in the fixture.
-    const PROB_OF_THE = 1000 / 500500;
-
     let rootTraversal = model.traverseFromRoot();
     assert.isDefined(rootTraversal);
 
@@ -139,7 +141,6 @@ describe('Trie traversal abstractions', function() {
 
   it('traversal over compact leaf node', function() {
     var model = new TrieModel(jsonFixture('tries/english-1000'));
-    const PROB_OF_TROUBLE = 267 / 500500;
 
     let rootTraversal = model.traverseFromRoot();
     assert.isDefined(rootTraversal);
@@ -153,15 +154,15 @@ describe('Trie traversal abstractions', function() {
         assert.isDefined(traversalInner1);
         assert.isArray(child.traversal().entries);
         assert.isEmpty(child.traversal().entries);
-        assert.equal(traversalInner1.p, 1000 / 500500 /* prob of 'the' */);
+        assert.equal(traversalInner1.p, PROB_OF_THE);
 
         for(let tChild of traversalInner1.children()) {
           if(tChild.char == 'r') {
             let traversalInner2 = tChild.traversal();
             assert.isDefined(traversalInner2);
             assert.isArray(tChild.traversal().entries);
             assert.isEmpty(tChild.traversal().entries);
-            assert.equal(traversalInner2.p, 607 / 500500 /* prob of 'true', the best 'tr-' entry */);
+            assert.equal(traversalInner2.p, PROB_OF_TRUE);
 
             for(let rChild of traversalInner2.children()) {
               if(rChild.char == 'o') {

common/models/types/index.d.ts

@@ -32,8 +32,10 @@ type TextWithProbability = {
 
   /**
    * The probability of the lexical entry, directly based upon its frequency.
+   *
+   * A real-number weight, from 0 to 1.
    */
-  p: number; // real-number weight, from 0 to 1
+  p: number;
 }
 
 /**
@@ -68,10 +70,9 @@ declare interface LexiconTraversal {
   children(): Generator<{char: USVString, traversal: () => LexiconTraversal}>;
 
   /**
-   * Allows direct access to the traversal state that results when appending a
-   * `char` representing one or more individual UTF-16 codepoints to the
-   * current traversal state's prefix.  This bypasses the need to iterate
-   * among all legal child Traversals.
+   * Allows direct access to the traversal state that results when appending one
+   * or more codepoints encoded in UTF-16 to the current traversal state's prefix.
+   * This allows bypassing iteration among all legal child Traversals.
    *
    * If such a traversal state is not supported, returns `undefined`.
    *

common/test/resources/json/models/future_suggestions/i_got_distracted_by_hazel.json

@@ -48,7 +48,7 @@
   [
     {
       "transform": {
-        "insert": "distracted ",
+        "insert": "distracted by ",
         "deleteLeft": 0
       },
       "displayAs": "distracted by"

common/test/resources/models/simple-dummy.js

@@ -9,31 +9,29 @@
 
     Model.punctuation = {
       quotesForKeepSuggestion: { open: '“', close: '”'},
-      // Important! Set this, or else the model compositor will
-      // insert something for us!
-      insertAfterWord: "",
+      insertAfterWord: " ",
     };
 
     // A direct import/copy from i_got_distracted_by_hazel.json.
     Model.futureSuggestions = [
       [
         {
           "transform": {
-            "insert": "I ",
+            "insert": "I",
             "deleteLeft": 0
           },
           "displayAs": "I"
         },
         {
           "transform": {
-            "insert": "I'm ",
+            "insert": "I'm",
             "deleteLeft": 0
           },
           "displayAs": "I'm"
         },
         {
           "transform": {
-            "insert": "Oh ",
+            "insert": "Oh",
             "deleteLeft": 0
           },
           "displayAs": "Oh"
@@ -42,21 +40,21 @@
       [
         {
           "transform": {
-            "insert": "love ",
+            "insert": "love",
             "deleteLeft": 0
           },
           "displayAs": "love"
         },
         {
           "transform": {
-            "insert": "am ",
+            "insert": "am",
             "deleteLeft": 0
           },
           "displayAs": "am"
         },
         {
           "transform": {
-            "insert": "got ",
+            "insert": "got",
             "deleteLeft": 0
           },
           "displayAs": "got"
@@ -65,21 +63,21 @@
       [
         {
           "transform": {
-            "insert": "distracted ",
+            "insert": "distracted by",
             "deleteLeft": 0
           },
           "displayAs": "distracted by"
         },
         {
           "transform": {
-            "insert": "distracted ",
+            "insert": "distracted",
             "deleteLeft": 0
           },
           "displayAs": "distracted"
         },
         {
           "transform": {
-            "insert": "a ",
+            "insert": "a",
             "deleteLeft": 0
           },
           "displayAs": "a"
@@ -88,27 +86,27 @@
       [
         {
           "transform": {
-            "insert": "Hazel ",
+            "insert": "Hazel",
             "deleteLeft": 0
           },
           "displayAs": "Hazel"
         },
         {
           "transform": {
-            "insert": "the ",
+            "insert": "the",
             "deleteLeft": 0
           },
           "displayAs": "the"
         },
         {
           "transform": {
-            "insert": "a ",
+            "insert": "a",
             "deleteLeft": 0
           },
           "displayAs": "a"
         }
       ]
-    ];    
+    ];
 
     return Model;
   }());

common/web/input-processor/tests/cases/predictionContext.js

@@ -36,7 +36,7 @@ const appleDummySuggestionSets = [[
 ], [
   // Set 2:
   {
-    transform: { insert: '', deleteLeft: 0},
+    transform: { insert: 'e', deleteLeft: 0},
     displayAs: 'apple',
     tag: 'keep'
   }, {