Mnemonic Language Detection Utils

packages/primitives/src/mnemonic.ts

@@ -1,9 +1,19 @@
 import {
-  bip39WordListEnglish,
   decodeBip39MnemonicNonStandard,
   deriveSeedFromBip39Mnemonic,
   encodeBip39MnemonicNonStandard,
   generateBip39MnemonicNonStandard,
+  // Word lists.
+  bip39WordListChineseSimplified,
+  bip39WordListChineseTraditional,
+  bip39WordListCzech,
+  bip39WordListEnglish,
+  bip39WordListFrench,
+  bip39WordListItalian,
+  bip39WordListJapanese,
+  bip39WordListKorean,
+  bip39WordListPortuguese,
+  bip39WordListSpanish,
 } from '@bitauth/libauth';
 import { Bytes } from './bytes.js';
 
@@ -253,15 +263,79 @@ export class Mnemonic {
    * Checks if a word exists in the provided dictionary.
    *
    * @param word Word to check.
-   * @param dictionary Optional wordlist (defaults to English BIP-39 wordlist).
+   * @param wordList Optional wordlist (defaults to English BIP-39 wordlist).
    *
    * @returns boolean indicating if the word is valid.
    */
   public static isValidWord(
     word: string,
-    dictionary: string[] = bip39WordListEnglish,
+    wordList: string[] = bip39WordListEnglish,
   ): boolean {
-    return dictionary.includes(word.toLowerCase());
+    return wordList.includes(word.toLowerCase());
+  }
+
+  /**
+   * Detect which languages (word-lists) the given mnemonic matches.
+   *
+   * @param words The mnemonic words to check
+   * @param wordLists The BIP39 word lists to check against, keyed by language name
+   * @param allowPartialMatches If true, treats incomplete words as valid matches (useful for autocomplete)
+   *
+   * @returns Object mapping language names to their word lists for all matching languages
+   */
+  public static detectLanguages(
+    words: Array<string>,
+    wordLists: { [name: string]: Array<string> },
+    allowPartialMatches = false,
+  ): { [name: string]: Array<string> } {
+    // Define our match function.
+    const isWordMatch = (word: string, wordList: Array<string>): boolean => {
+      if (allowPartialMatches && word.length === 0) return true;
+      if (wordList.includes(word)) return true;
+      return (
+        allowPartialMatches &&
+        wordList.some((fullWord) => fullWord.startsWith(word))
+      );
+    };
+
+    // Iterate over each word-list and check whether it's a match.
+    return Object.fromEntries(
+      Object.entries(wordLists).filter(([, wordList]) =>
+        words.every((word) => isWordMatch(word, wordList)),
+      ),
+    );
+  }
+
+  /**
+   * Get all valid BIP39 words from languages that match the given mnemonic words.
+   *
+   * Uses partial matching to determine eligible languages, then returns all words
+   * from those languages (useful for autocomplete and word validation).
+   *
+   * @param words The mnemonic words to check (can be partial/incomplete)
+   * @param wordLists The BIP39 word lists to check against, keyed by language name
+   * @returns Array of all unique words from matching languages, sorted alphabetically
+   *
+   * @example
+   * ```typescript
+   * // Get all valid words for languages that match partial input
+   * const eligibleWords = Mnemonic.getEligibleWords(['aban', 'abil'], wordLists);
+   * // Returns all words from languages where 'aban*' and 'abil*' are valid prefixes
+   *
+   * // Filter for autocomplete
+   * const completions = eligibleWords.filter(word => word.startsWith('acc'));
+   * ```
+   */
+  public static getEligibleWords(
+    words: Array<string>,
+    wordLists: { [name: string]: Array<string> },
+  ): Array<string> {
+    // Given the words provided, find eligible languages/word-lists.
+    const matchingLanguages = this.detectLanguages(words, wordLists, true);
+
+    // Merge all eligible word completions into an Array.
+    // NOTE: We use a Set to de-deuplicate words that might appear in multiple word-lists.
+    return [...new Set<string>(Object.values(matchingLanguages).flat())];
   }
 
   /**
@@ -347,3 +421,24 @@ export class Mnemonic {
     return this.mnemonic.passphrase;
   }
 }
+
+// NOTE: This is handled exceptionally due to tree-shaking considerations.
+//       Word lists incur a large cost on the bundle size and LibAuth considers non-English word-lists as non-standard:
+//       https://github.com/bitauth/libauth/blob/master/src/lib/key/bip39.ts#L178
+//       And the BIP39 spec warns against their usage:
+//       https://github.com/bitcoin/bips/blob/master/bip-0039.mediawiki#wordlists
+//       If we include this in the class, it cannot be tree-shaken by most tooling, so we define a stand-alone function instead.
+export function getAllBIP39WordLists() {
+  return {
+    'Chinese Simplified': bip39WordListChineseSimplified,
+    'Chinese Traditional': bip39WordListChineseTraditional,
+    Czech: bip39WordListCzech,
+    English: bip39WordListEnglish,
+    French: bip39WordListFrench,
+    Italian: bip39WordListItalian,
+    Japanese: bip39WordListJapanese,
+    Korean: bip39WordListKorean,
+    Portuguese: bip39WordListPortuguese,
+    Spanish: bip39WordListSpanish,
+  };
+}