str.translate now process also non ASCII code

Author: ppisl

graalpython/com.oracle.graal.python.test/src/tests/test_string.py

@@ -907,6 +907,19 @@ def test_translate_from_byte_table():
     assert "ahoj".translate(table) == "AHOJ"
     assert "ahoj".translate(bytearray(table)) == "AHOJ"
     assert "ahoj".translate(memoryview(table)) == "AHOJ"
+    
+def test_tranlslate_from_short_table():
+    table = b'\x00\x01\x02\x03\x04\x05\x06\x07\x08\t\n\x0b\x0c\r\x0e\x0f\x10\x11\x12\x13\x14\x15\x16\x17\x18\x19\x1a\x1b\x1c\x1d\x1e\x1f !"#$%&\'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_`ABCDEFGH'
+    assert "ahoj".translate(table) == "AHoj"
+
+def test_translate_nonascii_from_byte_table():
+    table = bytes.maketrans(bytes(string.ascii_lowercase, 'ascii'), bytes(string.ascii_uppercase, 'ascii'))
+    assert "ačhřožj".translate(table) == "AčHřOžJ"
+
+def test_translate_from_long_byte_table():
+    table = bytes.maketrans(bytes(string.ascii_lowercase, 'ascii'), bytes(string.ascii_uppercase, 'ascii'))
+    table *= 30
+    assert 'ahoj453875287ščřžýáí'.translate(table) == 'AHOJ453875287A\rY~ýáí'
 
 def test_splitlines():
     assert len(str.splitlines("\n\n")) == 2

graalpython/com.oracle.graal.python/src/com/oracle/graal/python/builtins/objects/str/StringBuiltins.java

@@ -107,6 +107,7 @@
 import com.oracle.truffle.api.dsl.TypeSystemReference;
 import com.oracle.truffle.api.profiles.BranchProfile;
 import com.oracle.truffle.api.profiles.ConditionProfile;
+import java.nio.charset.StandardCharsets;
 
 @CoreFunctions(extendClasses = PythonBuiltinClassType.PString)
 public final class StringBuiltins extends PythonBuiltins {
@@ -757,10 +758,15 @@ public String translate(String self, PDict table,
 
         @TruffleBoundary
         private static String translateFromByteTable(String text, byte[] table) {
-            byte[] translatedChars = text.getBytes();
-            for (int i = 0; i < translatedChars.length; i++) {
-                byte original = translatedChars[i];
-                translatedChars[i] = table[original];
+            char[] translatedChars = new char[text.length()];
+            // convert only ascii or up to the lenght of table
+            for (int i = 0; i < text.length(); i++) {
+                char code = text.charAt(i);
+                if (code < table.length) {
+                    translatedChars[i] = (char) (table[code] & 0xFF);
+                } else {
+                    translatedChars[i] = code;
+                }
             }
             return new String(translatedChars);
         }