Fix space omission when translating into CJK languages (#956)

This PR fixes an issue where we are checking for equality on the
language tag for CJK languages, when we really need to be checking
if the tag starts with the language tag.

Before this PR `zh-Hans` would not match. After this PR it will match.

inference/src/translator/annotation.cpp

@@ -34,6 +34,13 @@ void AnnotatedText::appendSentence(string_view prefix, std::vector<string_view>:
   annotation.token_begin_.push_back(offset);
 }
 
+/// A simple helper function to check if a string starts with a prefix.
+/// The std::string object only has a starts_with() method in C++20, which
+/// is not what we are currently compiling with.
+bool startsWith(string_view prefix, string_view str) {
+  return str.size() >= prefix.size() && prefix == str.substr(0, prefix.size());
+}
+
 bool AnnotatedText::shouldOmitSpaceBetweenSentences() const {
   if (targetLanguage_.empty()) {
     // The target language is not specified, so we should not make assumptions about
@@ -45,11 +52,11 @@ bool AnnotatedText::shouldOmitSpaceBetweenSentences() const {
   // More robustly handle which language tags should omit whitespace between sentences.
   return (
     // Japanese does not use space between sentences.
-    targetLanguage_ == "ja" ||
+    startsWith("ja", targetLanguage_) ||
     // Korean does not use space between sentences.
-    targetLanguage_ == "ko" ||
+    startsWith("ko", targetLanguage_) ||
     // Chinese does not use space between sentences.
-    targetLanguage_ == "zh"
+    startsWith("zh", targetLanguage_)
   );
 }