Issue #1285086: Get rid of the refcounting hack and speed up urllib.unquote().

diff --git a/Lib/urllib.py b/Lib/urllib.py
index 33641a5700125f7f251210dc917b48f31adb618e..f9655f9e88a47c021562ec660d9eeb3825b7ef06 100644 (file)
--- a/Lib/urllib.py
+++ b/Lib/urllib.py
@@ -28,6 +28,7 @@ import os
 import time
 import sys
 import base64
+import re
 
 from urlparse import urljoin as basejoin
 
@@ -1198,22 +1199,35 @@ def splitvalue(attr):
 _hexdig = '0123456789ABCDEFabcdef'
 _hextochr = dict((a + b, chr(int(a + b, 16)))
                  for a in _hexdig for b in _hexdig)
+_asciire = re.compile('([\x00-\x7f]+)')
 
 def unquote(s):
     """unquote('abc%20def') -> 'abc def'."""
-    res = s.split('%')
+    if _is_unicode(s):
+        if '%' not in s:
+            return s
+        bits = _asciire.split(s)
+        res = [bits[0]]
+        append = res.append
+        for i in range(1, len(bits), 2):
+            append(unquote(str(bits[i])).decode('latin1'))
+            append(bits[i + 1])
+        return ''.join(res)
+
+    bits = s.split('%')
     # fastpath
-    if len(res) == 1:
+    if len(bits) == 1:
         return s
-    s = res[0]
-    for item in res[1:]:
+    res = [bits[0]]
+    append = res.append
+    for item in bits[1:]:
         try:
-            s += _hextochr[item[:2]] + item[2:]
+            append(_hextochr[item[:2]])
+            append(item[2:])
         except KeyError:
-            s += '%' + item
-        except UnicodeDecodeError:
-            s += unichr(int(item[:2], 16)) + item[2:]
-    return s
+            append('%')
+            append(item)
+    return ''.join(res)
 
 def unquote_plus(s):
     """unquote('%7e/abc+def') -> '~/abc def'"""

diff --git a/Lib/urlparse.py b/Lib/urlparse.py
index f370ce3bdcc17b11e446a1e83a7df74d50c6b7cb..4ce982e8fdd4e52e403a41520ec566f17f0d1f5a 100644 (file)
--- a/Lib/urlparse.py
+++ b/Lib/urlparse.py
@@ -28,6 +28,8 @@ test_urlparse.py provides a good indicator of parsing behavior.
 
 """
 
+import re
+
 __all__ = ["urlparse", "urlunparse", "urljoin", "urldefrag",
            "urlsplit", "urlunsplit", "parse_qs", "parse_qsl"]
 
@@ -311,6 +313,15 @@ def urldefrag(url):
     else:
         return url, ''
 
+try:
+    unicode
+except NameError:
+    def _is_unicode(x):
+        return 0
+else:
+    def _is_unicode(x):
+        return isinstance(x, unicode)
+
 # unquote method for parse_qs and parse_qsl
 # Cannot use directly from urllib as it would create a circular reference
 # because urllib uses urlparse methods (urljoin).  If you update this function,
@@ -319,22 +330,35 @@ def urldefrag(url):
 _hexdig = '0123456789ABCDEFabcdef'
 _hextochr = dict((a+b, chr(int(a+b,16)))
                  for a in _hexdig for b in _hexdig)
+_asciire = re.compile('([\x00-\x7f]+)')
 
 def unquote(s):
     """unquote('abc%20def') -> 'abc def'."""
-    res = s.split('%')
+    if _is_unicode(s):
+        if '%' not in s:
+            return s
+        bits = _asciire.split(s)
+        res = [bits[0]]
+        append = res.append
+        for i in range(1, len(bits), 2):
+            append(unquote(str(bits[i])).decode('latin1'))
+            append(bits[i + 1])
+        return ''.join(res)
+
+    bits = s.split('%')
     # fastpath
-    if len(res) == 1:
+    if len(bits) == 1:
         return s
-    s = res[0]
-    for item in res[1:]:
+    res = [bits[0]]
+    append = res.append
+    for item in bits[1:]:
         try:
-            s += _hextochr[item[:2]] + item[2:]
+            append(_hextochr[item[:2]])
+            append(item[2:])
         except KeyError:
-            s += '%' + item
-        except UnicodeDecodeError:
-            s += unichr(int(item[:2], 16)) + item[2:]
-    return s
+            append('%')
+            append(item)
+    return ''.join(res)
 
 def parse_qs(qs, keep_blank_values=0, strict_parsing=0):
     """Parse a query given as a string argument.

diff --git a/Misc/NEWS b/Misc/NEWS
index ae0402b2484870089062a707a7a1492b9ca54410..10c3bd0f829506ad9f6ac68c15f5b061322b3021 100644 (file)
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -214,6 +214,8 @@ Core and Builtins
 Library
 -------
 
+- Issue #1285086: Get rid of the refcounting hack and speed up urllib.unquote().
+
 - Issue #17368: Fix an off-by-one error in the Python JSON decoder that caused
   a failure while decoding empty object literals when object_pairs_hook was
   specified.