A free, open source, and extensible speech-to-text application that works completely offline.
refactor(transcription): optimize word correction performance
Vlad Gerasimov6404ce74878f
diff --git a/src-tauri/src/managers/transcription.rs b/src-tauri/src/managers/transcription.rsindex 13fc4c20..9f70501c 100644--- a/src-tauri/src/managers/transcription.rs+++ b/src-tauri/src/managers/transcription.rs@@ -32,6 +32,11 @@ fn correct_words(text: &str, correct_words: &[String]) -> String {return text.to_string();}+ // Pre-compute lowercase versions to avoid repeated allocations+ let correct_words_lower: Vec<String> = correct_words.iter()+ .map(|w| w.to_lowercase())+ .collect();+let words: Vec<&str> = text.split_whitespace().collect();let mut corrected_words = Vec::new();@@ -43,19 +48,29 @@ fn correct_words(text: &str, correct_words: &[String]) -> String {continue;}+ // Skip extremely long words to avoid performance issues+ if cleaned_word.len() > 50 {+ corrected_words.push(word.to_string());+ continue;+ }+let mut best_match: Option<&String> = None;let mut best_score = f64::MAX;- for correct_word in correct_words {- let correct_word_lower = correct_word.to_lowercase();+ for (i, correct_word_lower) in correct_words_lower.iter().enumerate() {+ // Skip if lengths are too different (optimization)+ let len_diff = (cleaned_word.len() as i32 - correct_word_lower.len() as i32).abs();+ if len_diff > 5 {+ continue;+ }// Calculate Levenshtein distance (normalized by length)- let levenshtein_dist = levenshtein(&cleaned_word, &correct_word_lower);+ let levenshtein_dist = levenshtein(&cleaned_word, correct_word_lower);let max_len = cleaned_word.len().max(correct_word_lower.len()) as f64;let levenshtein_score = if max_len > 0.0 { levenshtein_dist as f64 / max_len } else { 1.0 };// Calculate phonetic similarity using Soundex- let phonetic_match = soundex(&cleaned_word, &correct_word_lower);+ let phonetic_match = soundex(&cleaned_word, correct_word_lower);// Combine scores: favor phonetic matches, but also consider string similaritylet combined_score = if phonetic_match {@@ -66,7 +81,7 @@ fn correct_words(text: &str, correct_words: &[String]) -> String {// Accept if the score is good enough (threshold: 0.4 for good matches)if combined_score < 0.4 && combined_score < best_score {- best_match = Some(correct_word);+ best_match = Some(&correct_words[i]);best_score = combined_score;}}@@ -85,14 +100,12 @@ fn correct_words(text: &str, correct_words: &[String]) -> String {replacement.clone()};- // Preserve punctuation from original word- let original_suffix: String = word.chars().rev()- .take_while(|c| !c.is_alphabetic())- .collect::<String>()- .chars().rev().collect();- let original_prefix: String = word.chars()- .take_while(|c| !c.is_alphabetic())- .collect();+ // Preserve punctuation from original word - optimized version+ let prefix_end = word.chars().take_while(|c| !c.is_alphabetic()).count();+ let suffix_start = word.char_indices().rev().take_while(|(_, c)| !c.is_alphabetic()).count();++ let original_prefix = if prefix_end > 0 { &word[..prefix_end] } else { "" };+ let original_suffix = if suffix_start > 0 { &word[word.len() - suffix_start..] } else { "" };corrected_words.push(format!("{}{}{}", original_prefix, corrected, original_suffix));} else {diff --git a/src/components/settings/CorrectWords.tsx b/src/components/settings/CorrectWords.tsxindex f0908245..e0115c35 100644--- a/src/components/settings/CorrectWords.tsx+++ b/src/components/settings/CorrectWords.tsx@@ -1,5 +1,4 @@import React, { useState } from "react";-import { SettingContainer } from "../ui/SettingContainer";import { useSettings } from "../../hooks/useSettings";interface CorrectWordsProps {