From 5ae6875e8899c7907ce524b5f43dbfe12b3e5c44 Mon Sep 17 00:00:00 2001
From: Yigong Wang <yigwang@cs.stonybrook.edu>
Date: Tue, 27 Mar 2018 20:04:43 -0400
Subject: [PATCH] Fix `mangle` for Pythons compiled with UCS-2

---
 hy/_compat.py    |  4 ++++
 hy/lex/parser.py | 22 ++++++++++++++++++----
 2 files changed, 22 insertions(+), 4 deletions(-)

diff --git a/hy/_compat.py b/hy/_compat.py
index 3c5b7d9..c40e44d 100644
--- a/hy/_compat.py
+++ b/hy/_compat.py
@@ -25,6 +25,10 @@ PY35 = sys.version_info >= (3, 5)
 PY36 = sys.version_info >= (3, 6)
 PY37 = sys.version_info >= (3, 7)
 
+# The value of UCS4 indicates whether Unicode strings are stored as UCS-4.
+# It is always true on Pythons >= 3.3, which use USC-4 on all systems.
+UCS4 = sys.maxunicode == 0x10FFFF
+
 str_type     = str   if PY3 else unicode      # NOQA
 bytes_type   = bytes if PY3 else str          # NOQA
 long_type    = int   if PY3 else long         # NOQA
diff --git a/hy/lex/parser.py b/hy/lex/parser.py
index 238f57a..1287d91 100755
--- a/hy/lex/parser.py
+++ b/hy/lex/parser.py
@@ -10,7 +10,7 @@ import string, re, unicodedata
 
 from rply import ParserGenerator
 
-from hy._compat import PY3, str_type, isidentifier
+from hy._compat import PY3, str_type, isidentifier, UCS4
 from hy.models import (HyBytes, HyComplex, HyCons, HyDict, HyExpression,
                        HyFloat, HyInteger, HyKeyword, HyList, HySet, HyString,
                        HySymbol)
@@ -25,14 +25,28 @@ pg = ParserGenerator(
 
 mangle_delim = 'Δ' if PY3 else 'X'
 
+def unicode_to_ucs4iter(ustr):
+    # Covert a unicode string to an iterable object,
+    # elements in the object are single USC-4 unicode characters
+    if UCS4:
+        return ustr
+    ucs4_list = list(ustr)
+    for i, u in enumerate(ucs4_list):
+        if 0xD7FF < ord(u) < 0xDC00:
+            ucs4_list[i] += ucs4_list[i + 1]
+            del ucs4_list[i + 1]
+    return ucs4_list
+
 def mangle(s):
     """Stringify the argument and convert it to a valid Python identifier
     according to Hy's mangling rules."""
+    def unicode_char_to_hex(uchr):
+        # Covert a unicode char to hex string, without prefix
+        return uchr.encode('unicode-escape').decode('utf-8').lstrip('\\U').lstrip('\\u').lstrip('0')
 
     assert s
 
     s = str_type(s)
-
     s = s.replace("-", "_")
     s2 = s.lstrip('_')
     leading_underscores = '_' * (len(s) - len(s2))
@@ -50,8 +64,8 @@ def mangle(s):
                  # allowed at the start of an identifier.
                else '{0}{1}{0}'.format(mangle_delim,
                    unicodedata.name(c, '').lower().replace('-', 'H').replace(' ', '_')
-                   or 'U{:x}'.format(ord(c)))
-            for c in s)
+                   or 'U{}'.format(unicode_char_to_hex(c)))
+            for c in unicode_to_ucs4iter(s))
 
     s = leading_underscores + s
     assert isidentifier(s)