[pypy-svn] r75426 - in pypy/trunk/pypy/module/_codecs: . test

[arigo at codespeak.net]

Author: arigo
Date: Wed Jun 16 17:08:03 2010
New Revision: 75426

Modified:
   pypy/trunk/pypy/module/_codecs/app_codecs.py
   pypy/trunk/pypy/module/_codecs/test/test_codecs.py
Log:
Some random fixes until test_codecs passes on Mac,
with a 16-bit unicode CPython and emulating a 32-bit
PyPy.


Modified: pypy/trunk/pypy/module/_codecs/app_codecs.py
==============================================================================
--- pypy/trunk/pypy/module/_codecs/app_codecs.py	(original)
+++ pypy/trunk/pypy/module/_codecs/app_codecs.py	Wed Jun 16 17:08:03 2010
@@ -39,8 +39,6 @@
 # XXX move some of these functions to RPython (like charmap_encode,
 # charmap_build) to make them faster
 
-import sys
-
 def escape_encode( obj, errors='strict'):
     """None
     """
@@ -84,36 +82,36 @@
     res = ''.join(res)
     return res, len(res)
 
-if sys.maxunicode == 65535:
-    unicode_bytes = 2
-else:
-    unicode_bytes = 4
-
 def unicode_internal_encode( obj, errors='strict'):
     """None
     """
-    if type(obj) == unicode:
-        p = []
-        t = [ord(x) for x in obj]
-        for i in t:
-            bytes = []
-            for j in xrange(unicode_bytes):
-                bytes += chr(i%256)
-                i >>= 8
-            if sys.byteorder == "big":
-                bytes.reverse()
-            p += bytes
-        res = ''.join(p)
-        return res, len(res)
+    import sys
+    if sys.maxunicode == 65535:
+        unicode_bytes = 2
     else:
-        res = "You can do better than this" # XXX make this right
-        return res, len(res)
+        unicode_bytes = 4
+    p = []
+    for x in obj:
+        i = ord(x)
+        bytes = []
+        for j in xrange(unicode_bytes):
+            bytes += chr(i%256)
+            i >>= 8
+        if sys.byteorder == "big":
+            bytes.reverse()
+        p += bytes
+    res = ''.join(p)
+    return res, len(res)
 
 def unicode_internal_decode( unistr, errors='strict'):
-    import sys
     if type(unistr) == unicode:
         return unistr, len(unistr)
     else:
+        import sys
+        if sys.maxunicode == 65535:
+            unicode_bytes = 2
+        else:
+            unicode_bytes = 4
         p = []
         i = 0
         if sys.byteorder == "big":
@@ -541,6 +539,7 @@
 hexdigits = [hex(i)[-1] for i in range(16)]+[hex(i)[-1].upper() for i in range(10, 16)]
 
 def hexescape(s, pos, digits, message, errors):
+    import sys
     chr = 0
     p = []
     if (pos+digits>len(s)):
@@ -580,6 +579,7 @@
     return res, pos
 
 def PyUnicode_DecodeUnicodeEscape(s, size, errors):
+    import sys
 
     if (size == 0):
         return u''
@@ -762,6 +762,7 @@
 
 
 def PyUnicode_DecodeRawUnicodeEscape(s, size, errors):
+    import sys
 
     if (size == 0):
         return u''

Modified: pypy/trunk/pypy/module/_codecs/test/test_codecs.py
==============================================================================
--- pypy/trunk/pypy/module/_codecs/test/test_codecs.py	(original)
+++ pypy/trunk/pypy/module/_codecs/test/test_codecs.py	Wed Jun 16 17:08:03 2010
@@ -84,7 +84,7 @@
         assert str(UnicodeTranslateError(
             u"g\uffffrk", 1, 2, "ouch"))== "can't translate character u'\\uffff' in position 1: ouch"
         
-        if sys.maxunicode > 0xffff:
+        if sys.maxunicode > 0xffff and len(unichr(0x10000)) == 1:
             assert str(UnicodeTranslateError(
                 u"g\U00010000rk", 1, 2, "ouch"))== "can't translate character u'\\U00010000' in position 1: ouch"
             
@@ -107,7 +107,7 @@
        
         assert str(UnicodeEncodeError(
             "ascii", u"\uffffx", 0, 1, "ouch"))=="'ascii' codec can't encode character u'\\uffff' in position 0: ouch"
-        if sys.maxunicode > 0xffff:
+        if sys.maxunicode > 0xffff and len(unichr(0x10000)) == 1:
             assert str(UnicodeEncodeError(
                 "ascii", u"\U00010000x", 0, 1, "ouch")) =="'ascii' codec can't encode character u'\\U00010000' in position 0: ouch"
     
@@ -229,7 +229,9 @@
 
     def test_unicode_internal_encode(self):
         import sys
-        enc = u"a".encode("unicode_internal")
+        class U(unicode):
+            pass
+        enc = U(u"a").encode("unicode_internal")
         if sys.maxunicode == 65535: # UCS2 build
             if sys.byteorder == "big":
                 assert enc == "\x00a"