From 286593607e8a9d43627a00a82cfcde55183b0818 Mon Sep 17 00:00:00 2001
From: SuHun Han <ssut@ssut.me>
Date: Fri, 7 Apr 2017 22:43:48 +0900
Subject: [PATCH] Fix invalid tokenizer due to the complexity of the obfuscated
 code

This commit solves #14 and bumps the version to 2.1.2
---
 googletrans/__init__.py |  2 +-
 googletrans/gtoken.py   | 26 +++++++++++++-------------
 tests/test_client.py    |  4 ++--
 3 files changed, 16 insertions(+), 16 deletions(-)

diff --git a/googletrans/__init__.py b/googletrans/__init__.py
index 9f818df..c5dba57 100644
--- a/googletrans/__init__.py
+++ b/googletrans/__init__.py
@@ -1,6 +1,6 @@
 """Free Google Translate API for Python. Translates totally free of charge."""
 __all__ = 'Translator',
-__version_info__ = 2, 1, 1
+__version_info__ = 2, 1, 2
 __version__ = '.'.join(str(v) for v in __version_info__)
 
 
diff --git a/googletrans/gtoken.py b/googletrans/gtoken.py
index 34294a0..cdf29de 100644
--- a/googletrans/gtoken.py
+++ b/googletrans/gtoken.py
@@ -148,21 +148,21 @@ class TokenAcquirer(object):
             if l < 128:
                 e.append(l)
             # append calculated value if l is less than 2048
-            elif l < 2048:
-                e.append(l >> 6 | 192)
-                e.append(l)
-            # append calculated value if l matches special condition
-            elif (l & 64512) == 55296 and g + 1 < size and \
-                    ord(text[g + 1]) & 64512 == 56320:
-                g += 1
-                l = 65536 + ((l & 1023) << 10) + ord(text[g]) & 1023
-                e.append(l >> 18 | 240)
-                e.append(l >> 12 & 63 | 128)
             else:
-                e.append(l >> 12 | 224)
-                e.append(l >> 6 & 63 | 128)
+                if l < 2048:
+                    e.append(l >> 6 | 192)
+                else:
+                    # append calculated value if l matches special condition
+                    if (l & 64512) == 55296 and g + 1 < size and \
+                            ord(text[g + 1]) & 64512 == 56320:
+                        g += 1
+                        l = 65536 + ((l & 1023) << 10) + ord(text[g]) & 1023
+                        e.append(l >> 18 | 240)
+                        e.append(l >> 12 & 63 | 128)
+                    else:
+                        e.append(l >> 12 | 224)
+                        e.append(l >> 6 & 63 | 128)
                 e.append(l & 63 | 128)
-
         a = b
         for i, value in enumerate(e):
             a += value
diff --git a/tests/test_client.py b/tests/test_client.py
index 2407ab0..ccbfae6 100644
--- a/tests/test_client.py
+++ b/tests/test_client.py
@@ -28,9 +28,9 @@ def test_unicode(translator):
 
 
 def test_special_chars(translator):
-    text = u"Copyright © Google"
+    text = u"©×《》"
 
-    result = translator.translate(text, src='en', dest='fr')
+    result = translator.translate(text, src='en', dest='en')
     assert result.text == text