Make the python3 compliant. This doesn't deal w/ unicode yet

bencode.py

@@ -28,7 +28,6 @@
 # Software.
 
 
-from types import IntType, LongType, StringType, ListType, TupleType, DictType
 try:
     from types import BooleanType
 except ImportError:
@@ -37,29 +36,23 @@ try:
     from types import UnicodeType
 except ImportError:
     UnicodeType = None
-from cStringIO import StringIO
+from io import StringIO
 
 def decode_int(x, f):
     f += 1
-    newf = x.index('e', f)
-    try:
-        n = int(x[f:newf])
-    except:
-        n = long(x[f:newf])
-    if x[f] == '-':
-        if x[f + 1] == '0':
+    newf = x.index(b'e', f)
+    n = int(x[f:newf])
+    if x[f] == b'-'[0]:
+        if x[f + 1] == b'0'[0]:
             raise ValueError
-    elif x[f] == '0' and newf != f+1:
+    elif x[f] == b'0'[0] and newf != f+1:
         raise ValueError
     return (n, newf+1)
-  
+
 def decode_string(x, f):
-    colon = x.index(':', f)
-    try:
-        n = int(x[f:colon])
-    except (OverflowError, ValueError):
-        n = long(x[f:colon])
-    if x[f] == '0' and colon != f+1:
+    colon = x.index(b':', f)
+    n = int(x[f:colon])
+    if x[f] == b'0'[0] and colon != f+1:
         raise ValueError
     colon += 1
     return (x[colon:colon+n], colon+n)
@@ -70,15 +63,15 @@ def decode_unicode(x, f):
 
 def decode_list(x, f):
     r, f = [], f+1
-    while x[f] != 'e':
+    while x[f] != b'e'[0]:
         v, f = decode_func[x[f]](x, f)
         r.append(v)
     return (r, f + 1)
 
 def decode_dict(x, f):
     r, f = {}, f+1
-    lastkey = None
-    while x[f] != 'e':
+    lastkey = b''
+    while x[f] != b'e'[0]:
         k, f = decode_string(x, f)
         if lastkey >= k:
             raise ValueError
@@ -87,58 +80,63 @@ def decode_dict(x, f):
     return (r, f + 1)
 
 decode_func = {}
-decode_func['l'] = decode_list
-decode_func['d'] = decode_dict
-decode_func['i'] = decode_int
-decode_func['0'] = decode_string
-decode_func['1'] = decode_string
-decode_func['2'] = decode_string
-decode_func['3'] = decode_string
-decode_func['4'] = decode_string
-decode_func['5'] = decode_string
-decode_func['6'] = decode_string
-decode_func['7'] = decode_string
-decode_func['8'] = decode_string
-decode_func['9'] = decode_string
-#decode_func['u'] = decode_unicode
-  
+decode_func[b'l'[0]] = decode_list
+decode_func[b'd'[0]] = decode_dict
+decode_func[b'i'[0]] = decode_int
+decode_func[b'0'[0]] = decode_string
+decode_func[b'1'[0]] = decode_string
+decode_func[b'2'[0]] = decode_string
+decode_func[b'3'[0]] = decode_string
+decode_func[b'4'[0]] = decode_string
+decode_func[b'5'[0]] = decode_string
+decode_func[b'6'[0]] = decode_string
+decode_func[b'7'[0]] = decode_string
+decode_func[b'8'[0]] = decode_string
+decode_func[b'9'[0]] = decode_string
+#decode_func['u'[0]] = decode_unicode
+
 def bdecode(x, sloppy = 0):
     try:
         r, l = decode_func[x[0]](x, 0)
 #    except (IndexError, KeyError):
     except (IndexError, KeyError, ValueError):
-        raise ValueError, "bad bencoded data"
+        raise ValueError("bad bencoded data")
     if not sloppy and l != len(x):
-        raise ValueError, "bad bencoded data"
+        raise ValueError("bad bencoded data")
     return r
 
 def test_bdecode():
     try:
-        bdecode('0:0:')
+        bdecode(b'0:0:')
+        assert 0
+    except ValueError:
+        pass
+    try:
+        bdecode(b'ie')
         assert 0
     except ValueError:
         pass
     try:
-        bdecode('ie')
+        bdecode(b'i341foo382e')
         assert 0
     except ValueError:
         pass
+    assert bdecode(b'i4e') == 4
+    assert bdecode(b'i0e') == 0
+    assert bdecode(b'i123456789e') == 123456789
+    assert bdecode(b'i-10e') == -10
     try:
-        bdecode('i341foo382e')
+        bdecode(b'i-0e')
         assert 0
     except ValueError:
         pass
-    assert bdecode('i4e') == 4L
-    assert bdecode('i0e') == 0L
-    assert bdecode('i123456789e') == 123456789L
-    assert bdecode('i-10e') == -10L
     try:
-        bdecode('i-0e')
+        bdecode(b'i123')
         assert 0
     except ValueError:
         pass
     try:
-        bdecode('i123')
+        bdecode(b'')
         assert 0
     except ValueError:
         pass
@@ -148,108 +146,108 @@ def test_bdecode():
     except ValueError:
         pass
     try:
-        bdecode('i6easd')
+        bdecode(b'i6easd')
         assert 0
     except ValueError:
         pass
     try:
-        bdecode('35208734823ljdahflajhdf')
+        bdecode(b'35208734823ljdahflajhdf')
         assert 0
     except ValueError:
         pass
     try:
-        bdecode('2:abfdjslhfld')
+        bdecode(b'2:abfdjslhfld')
         assert 0
     except ValueError:
         pass
-    assert bdecode('0:') == ''
-    assert bdecode('3:abc') == 'abc'
-    assert bdecode('10:1234567890') == '1234567890'
+    assert bdecode(b'0:') == b''
+    assert bdecode(b'3:abc') == b'abc'
+    assert bdecode(b'10:1234567890') == b'1234567890'
     try:
-        bdecode('02:xy')
+        bdecode(b'02:xy')
         assert 0
     except ValueError:
         pass
     try:
-        bdecode('l')
+        bdecode(b'l')
         assert 0
     except ValueError:
         pass
-    assert bdecode('le') == []
+    assert bdecode(b'le') == []
     try:
-        bdecode('leanfdldjfh')
+        bdecode(b'leanfdldjfh')
         assert 0
     except ValueError:
         pass
-    assert bdecode('l0:0:0:e') == ['', '', '']
+    assert bdecode(b'l0:0:0:e') == [b'', b'', b'']
     try:
-        bdecode('relwjhrlewjh')
+        bdecode(b'relwjhrlewjh')
         assert 0
     except ValueError:
         pass
-    assert bdecode('li1ei2ei3ee') == [1, 2, 3]
-    assert bdecode('l3:asd2:xye') == ['asd', 'xy']
-    assert bdecode('ll5:Alice3:Bobeli2ei3eee') == [['Alice', 'Bob'], [2, 3]]
+    assert bdecode(b'li1ei2ei3ee') == [1, 2, 3]
+    assert bdecode(b'l3:asd2:xye') == [b'asd', b'xy']
+    assert bdecode(b'll5:Alice3:Bobeli2ei3eee') == [[b'Alice', b'Bob'], [2, 3]]
     try:
-        bdecode('d')
+        bdecode(b'd')
         assert 0
     except ValueError:
         pass
     try:
-        bdecode('defoobar')
+        bdecode(b'defoobar')
         assert 0
     except ValueError:
         pass
-    assert bdecode('de') == {}
-    assert bdecode('d3:agei25e4:eyes4:bluee') == {'age': 25, 'eyes': 'blue'}
-    assert bdecode('d8:spam.mp3d6:author5:Alice6:lengthi100000eee') == {'spam.mp3': {'author': 'Alice', 'length': 100000}}
+    assert bdecode(b'de') == {}
+    assert bdecode(b'd3:agei25e4:eyes4:bluee') == {b'age': 25, b'eyes': b'blue'}
+    assert bdecode(b'd8:spam.mp3d6:author5:Alice6:lengthi100000eee') == {b'spam.mp3': {b'author': b'Alice', b'length': 100000}}
     try:
-        bdecode('d3:fooe')
+        bdecode(b'd3:fooe')
         assert 0
     except ValueError:
         pass
     try:
-        bdecode('di1e0:e')
+        bdecode(b'di1e0:e')
         assert 0
     except ValueError:
         pass
     try:
-        bdecode('d1:b0:1:a0:e')
+        bdecode(b'd1:b0:1:a0:e')
         assert 0
     except ValueError:
         pass
     try:
-        bdecode('d1:a0:1:a0:e')
+        bdecode(b'd1:a0:1:a0:e')
         assert 0
     except ValueError:
         pass
     try:
-        bdecode('i03e')
+        bdecode(b'i03e')
         assert 0
     except ValueError:
         pass
     try:
-        bdecode('l01:ae')
+        bdecode(b'l01:ae')
         assert 0
     except ValueError:
         pass
     try:
-        bdecode('9999:x')
+        bdecode(b'9999:x')
         assert 0
     except ValueError:
         pass
     try:
-        bdecode('l0:')
+        bdecode(b'l0:')
         assert 0
     except ValueError:
         pass
     try:
-        bdecode('d0:0:')
+        bdecode(b'd0:0:')
         assert 0
     except ValueError:
         pass
     try:
-        bdecode('d0:')
+        bdecode(b'd0:')
         assert 0
     except ValueError:
         pass
@@ -261,83 +259,79 @@ class Bencached:
         self.marker = bencached_marker
         self.bencoded = s
 
-BencachedType = type(Bencached('')) # insufficient, but good as a filter
+BencachedType = type(Bencached(b'')) # insufficient, but good as a filter
 
 def encode_bencached(x,r):
     assert x.marker == bencached_marker
     r.append(x.bencoded)
 
 def encode_int(x,r):
-    r.extend(('i',str(x),'e'))
+    r.append(b'i%de' % x)
 
 def encode_bool(x,r):
     encode_int(int(x),r)
 
-def encode_string(x,r):    
-    r.extend((str(len(x)),':',x))
+def encode_bytes(x,r):
+    r.extend((b'%d:' % len(x),x))
 
-def encode_unicode(x,r):
+def encode_string(x,r):
     #r.append('u')
-    encode_string(x.encode('UTF-8'),r)
+    encode_bytes(x.encode('UTF-8'),r)
 
 def encode_list(x,r):
-        r.append('l')
+        r.append(b'l')
         for e in x:
             encode_func[type(e)](e, r)
-        r.append('e')
+        r.append(b'e')
 
 def encode_dict(x,r):
-    r.append('d')
-    ilist = x.items()
-    ilist.sort()
-    for k,v in ilist:
-        r.extend((str(len(k)),':',k))
+    r.append(b'd')
+    for k,v in sorted(x.items()):
+        r.extend((b'%d:' % len(k),k.encode('UTF-8')))
         encode_func[type(v)](v, r)
-    r.append('e')
+    r.append(b'e')
 
 encode_func = {}
 encode_func[BencachedType] = encode_bencached
-encode_func[IntType] = encode_int
-encode_func[LongType] = encode_int
-encode_func[StringType] = encode_string
-encode_func[ListType] = encode_list
-encode_func[TupleType] = encode_list
-encode_func[DictType] = encode_dict
+encode_func[int] = encode_int
+encode_func[str] = encode_string
+encode_func[list] = encode_list
+encode_func[tuple] = encode_list
+encode_func[type({})] = encode_dict
 if BooleanType:
     encode_func[BooleanType] = encode_bool
 if UnicodeType:
     encode_func[UnicodeType] = encode_unicode
-    
+
 def bencode(x):
     r = []
     try:
         encode_func[type(x)](x, r)
     except:
-        print "*** error *** could not encode type %s (value: %s)" % (type(x), x)
-        assert 0
-    return ''.join(r)
+        raise ValueError("could not encode type %s (value: %s)" % (type(x), x))
+    return b''.join(r)
 
 def test_bencode():
-    assert bencode(4) == 'i4e'
-    assert bencode(0) == 'i0e'
-    assert bencode(-10) == 'i-10e'
-    assert bencode(12345678901234567890L) == 'i12345678901234567890e'
-    assert bencode('') == '0:'
-    assert bencode('abc') == '3:abc'
-    assert bencode('1234567890') == '10:1234567890'
-    assert bencode([]) == 'le'
-    assert bencode([1, 2, 3]) == 'li1ei2ei3ee'
-    assert bencode([['Alice', 'Bob'], [2, 3]]) == 'll5:Alice3:Bobeli2ei3eee'
-    assert bencode({}) == 'de'
-    assert bencode({'age': 25, 'eyes': 'blue'}) == 'd3:agei25e4:eyes4:bluee'
-    assert bencode({'spam.mp3': {'author': 'Alice', 'length': 100000}}) == 'd8:spam.mp3d6:author5:Alice6:lengthi100000eee'
+    assert bencode(4) == b'i4e'
+    assert bencode(0) == b'i0e'
+    assert bencode(-10) == b'i-10e'
+    assert bencode(12345678901234567890) == b'i12345678901234567890e'
+    assert bencode('') == b'0:'
+    assert bencode('abc') == b'3:abc'
+    assert bencode('1234567890') == b'10:1234567890'
+    assert bencode([]) == b'le'
+    assert bencode([1, 2, 3]) == b'li1ei2ei3ee'
+    assert bencode([['Alice', 'Bob'], [2, 3]]) == b'll5:Alice3:Bobeli2ei3eee'
+    assert bencode({}) == b'de'
+    assert bencode({'age': 25, 'eyes': 'blue'}) == b'd3:agei25e4:eyes4:bluee'
+    assert bencode({'spam.mp3': {'author': 'Alice', 'length': 100000}}) == b'd8:spam.mp3d6:author5:Alice6:lengthi100000eee'
     try:
         bencode({1: 'foo'})
         assert 0
-    except AssertionError:
+    except (ValueError, AssertionError):
         pass
 
-  
+
 try:
     import psyco
     psyco.bind(bdecode)