convert chinese character to pinyin

Cargo.lock

@@ -3343,6 +3343,7 @@ dependencies = [
  "eyre",
  "k",
  "ndarray 0.15.6",
+ "pinyin",
  "pyo3",
  "rand 0.9.1",
  "rerun",
@@ -8420,6 +8421,12 @@ version = "0.1.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "8b870d8c151b6f2fb93e84a13146138f05d02ed11c7e7c54f8826aaaf7c9f184"
 
+[[package]]
+name = "pinyin"
+version = "0.10.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "16f2611cd06a1ac239a0cea4521de9eb068a6ca110324ee00631aa68daa74fc0"
+
 [[package]]
 name = "piper"
 version = "0.2.4"

examples/speech-to-text/whisper-dev.yml

@@ -28,7 +28,7 @@ nodes:
       # USE_MODELSCOPE_HUB: true
 
   - id: dora-rerun
-    build: cargo build -p dora-rerun --release
+    build: pip install -e ../../node-hub/dora-rerun
     path: dora-rerun
     inputs:
       original_text: dora-distil-whisper/text

node-hub/dora-rerun/Cargo.toml

@@ -28,6 +28,7 @@ pyo3 = { workspace = true, features = [
 ], optional = true }
 bytemuck = "1.20.0"
 rand = "0.9.1"
+pinyin = "0.10.0"
 
 
 [lib]

node-hub/dora-rerun/src/lib.rs

@@ -12,6 +12,7 @@ use dora_node_api::{
 };
 use eyre::{bail, eyre, Context, Result};
 
+use pinyin::ToPinyin;
 use rerun::{
     components::ImageBuffer, external::log::warn, ImageFormat, Points2D, Points3D, SpawnOptions,
 };
@@ -317,7 +318,24 @@ pub fn lib_main() -> Result<()> {
                 let buffer: StringArray = data.to_data().into();
                 buffer.iter().try_for_each(|string| -> Result<()> {
                     if let Some(str) = string {
-                        rec.log(id.as_str(), &rerun::TextLog::new(str))
+                        let chars = str.chars().collect::<Vec<_>>();
+                        let mut new_string = vec![];
+                        for char in chars {
+                            // Check if the character is a Chinese character
+                            if char.is_ascii() || char.is_control() {
+                                new_string.push(char);
+                                continue;
+                            }
+                            // If it is a Chinese character, replace it with its pinyin
+                            if let Some(pinyin) = char.to_pinyin() {
+                                for char in pinyin.with_tone().chars() {
+                                    new_string.push(char);
+                                }
+                                new_string.push(' ');
+                            }
+                        }
+                        let pinyined_str = new_string.iter().collect::<String>();
+                        rec.log(id.as_str(), &rerun::TextLog::new(pinyined_str))
                             .wrap_err("Could not log text")
                     } else {
                         Ok(())