TesseractOCR: remove call to setlocale().

This was a workaround for JVM crashes caused by Tesseract for years.
This is not needed anymore.

src/main/org/audiveris/omr/text/tesseract/TesseractOCR.java

@@ -103,11 +103,6 @@ public Set<String> getLanguages ()
             final Path ocrFolder = getOcrFolder();
             final TreeSet<String> set = new TreeSet<>();
 
-            // we need to call this to avoid crash
-            // in native code on non-english systems
-            // https://github.com/bytedeco/javacpp-presets/issues/694
-            setlocale(LC_ALL(), "C");
-
             try {
                 final TessBaseAPI api = new TessBaseAPI();
 

src/main/org/audiveris/omr/text/tesseract/TesseractOrder.java

@@ -186,9 +186,6 @@ public List<TextLine> process ()
                 }
             }
 
-            // prevent JVM crashes on non-english systems
-            setlocale(LC_ALL(), "C");
-
             if (api.Init(ocrFolder.toString(), lang, OEM_TESSERACT_ONLY) != 0) {
                 logger.warn("Could not initialize Tesseract with lang {}", lang);