diff --git a/scripts/translations/export_translations.py b/scripts/translations/export_translations.py
index 5c10b58e20..ffb96bcdc0 100644
--- a/scripts/translations/export_translations.py
+++ b/scripts/translations/export_translations.py
@@ -1,5 +1,6 @@
 # coding=UTF-8
 
+import math
 import os
 import re
 import sys
@@ -9,6 +10,54 @@
 PREFIXED_NAME_START = 'name="fui_'
 UNPREFIXED_NAME_START = 'name="'
 
+CHAR_LIMIT_PATTERN = re.compile(r'\[CHAR_LIMIT=\d+\]')
+TRANSLATION_DESC_PATTERN = re.compile(r'(translation_description="[^"]*?)(")')
+XML_TAG_PATTERN = re.compile(r'<[^>]+>')
+STRING_VALUE_PATTERN = re.compile(r'>([^<]*(?:<xliff:g[^>]*>[^<]*</xliff:g>[^<]*)*)</string>')
+ITEM_VALUE_PATTERN = re.compile(r'>([^<]*(?:<xliff:g[^>]*>[^<]*</xliff:g>[^<]*)*)</item>')
+
+XML_ENTITIES = {'&amp;': '&', '&lt;': '<', '&gt;': '>', '&quot;': '"', '&apos;': "'"}
+ESCAPE_PATTERN = re.compile(r'\\(u[0-9A-Fa-f]{4}|n|t|\'|")')
+
+
+def _decode_entities(text):
+    """Decode XML entities and Android escape sequences to get true visible length."""
+    for entity, char in XML_ENTITIES.items():
+        text = text.replace(entity, char)
+    text = ESCAPE_PATTERN.sub('X', text)
+    return text
+
+
+def _extract_visible_text(value):
+    """Strip XML tags and decode entities to get the visible text length."""
+    text = XML_TAG_PATTERN.sub('', value).strip()
+    return _decode_entities(text)
+
+
+def _calculate_char_limit(english_text):
+    """Return ~1.5x the English text length, rounded up to the nearest 5."""
+    length = len(english_text)
+    limit = math.ceil(length * 1.5)
+    return int(math.ceil(limit / 5.0) * 5)
+
+
+def _add_char_limit(text, value_pattern):
+    """Add [CHAR_LIMIT=xxx] to translation_description if missing."""
+    if 'translation_description=' not in text:
+        return text
+    if CHAR_LIMIT_PATTERN.search(text):
+        return text
+    match = value_pattern.search(text)
+    if not match:
+        return text
+    visible = _extract_visible_text(match.group(1))
+    if not visible:
+        return text
+    limit = _calculate_char_limit(visible)
+    return TRANSLATION_DESC_PATTERN.sub(
+        r'\1 [CHAR_LIMIT=%d]\2' % limit, text, count=1)
+
+
 class ExportTranslationsScript(BaseStringScript):
 
     def ProcessTag(self, line, type):
@@ -16,13 +65,21 @@ def ProcessTag(self, line, type):
 
         if PREFIXED_NAME_START in joined:
             joined = joined.replace(PREFIXED_NAME_START, UNPREFIXED_NAME_START)
-            return joined.split('\n')
-        else:
-            return line
+
+        if type == self.TYPE_STR:
+            joined = _add_char_limit(joined, STRING_VALUE_PATTERN)
+        elif type == self.TYPE_PLUR:
+            result_lines = []
+            for l in joined.split('\n'):
+                if '<item ' in l and '</item>' in l:
+                    l = _add_char_limit(l, ITEM_VALUE_PATTERN)
+                result_lines.append(l)
+            joined = '\n'.join(result_lines)
+
+        return joined.split('\n')
 
     def WriteFile(self, file_name, file_contents):
-        # Override to just print the contents
-        print file_contents
+        print(file_contents)
 
 if __name__ == '__main__':
     ets = ExportTranslationsScript()