update to version 0.1.3

- Fix Issue #2, use IOError instead of FileNotFoundError in Python 2
- Use another private obj to represent end of list and dict
- InvalidTorrentFileException -> InvalidTorrentDataException
- Now UnicodeDecodeError is warpped in InvalidTorrentDataException

test.py

@@ -26,24 +26,18 @@ class Test(unittest.TestCase):
 
     def test_parse_correctness(self):
         data = parse_torrent_file(self.TEST_FILENAME)
-        self.assertIn(['udp://p4p.arenabg.ch:1337/announce'],
+        self.assertIn(['udp://tracker.publicbt.com:80/announce'],
                       data['announce-list'])
-        self.assertEqual(data['comment'],
-                         'Torrent downloaded from https://rarbg.to')
-        self.assertEqual(data['creation date'], 1472762993)
+        self.assertEqual(data['creation date'], 1409254242)
 
     def test_parse_two_times(self):
         with open(self.TEST_FILENAME, 'rb') as fp:
             parser = TorrentFileParser(fp)
             data = parser.parse()
-            self.assertIn(['udp://p4p.arenabg.ch:1337/announce'],
+            self.assertIn(['udp://tracker.publicbt.com:80/announce'],
                           data['announce-list'])
-            self.assertEqual(data['comment'],
-                             'Torrent downloaded from https://rarbg.to')
-            self.assertEqual(data['creation date'], 1472762993)
+            self.assertEqual(data['creation date'], 1409254242)
             data = parser.parse()
-            self.assertIn(['udp://p4p.arenabg.ch:1337/announce'],
+            self.assertIn(['udp://tracker.publicbt.com:80/announce'],
                           data['announce-list'])
-            self.assertEqual(data['comment'],
-                             'Torrent downloaded from https://rarbg.to')
-            self.assertEqual(data['creation date'], 1472762993)
+            self.assertEqual(data['creation date'], 1409254242)

torrent_parser.py

@@ -22,20 +22,33 @@ import io
 import json
 import sys
 
+try:
+    FileNotFoundError
+except NameError:
+    # Python 2 do not have FileNotFoundError, use IOError instead
+    # noinspection PyShadowingBuiltins
+    FileNotFoundError = IOError
 
 __all__ = [
-    'InvalidTorrentFileException',
+    'InvalidTorrentDataException',
     'parse_torrent_file',
     'TorrentFileParser',
 ]
 
-__version__ = '0.1.0'
+__version__ = '0.1.3'
 
 
-class InvalidTorrentFileException(Exception):
+class InvalidTorrentDataException(Exception):
     def __init__(self, pos, msg=None):
-        msg = msg or "Invalid torrent format when reading at pos " + str(pos)
-        super(InvalidTorrentFileException, self).__init__(msg)
+        msg = msg or "Invalid torrent format when read at pos {pos}"
+        msg = msg.format(pos=pos)
+        super(InvalidTorrentDataException, self).__init__(msg)
+
+
+class __EndCls(object):
+    pass
+
+_END = __EndCls()
 
 
 class TorrentFileParser(object):
@@ -88,7 +101,7 @@ class TorrentFileParser(object):
 
         try:
             c = self._read_byte(1, True)
-            raise InvalidTorrentFileException(
+            raise InvalidTorrentDataException(
                 0, 'Expect EOF, but get [{}] at pos {}'.format(c, self._pos)
             )
         except EOFError:  # expect EOF
@@ -97,7 +110,7 @@ class TorrentFileParser(object):
         if isinstance(data, dict):
             return data
 
-        raise InvalidTorrentFileException('Outermost element is not a dict')
+        raise InvalidTorrentDataException('Outermost element is not a dict')
 
     def _read_byte(self, count=1, raise_eof=False):
         assert count >= 0
@@ -105,7 +118,7 @@ class TorrentFileParser(object):
         if count != 0 and len(gotten) == 0:
             if raise_eof:
                 raise EOFError()
-            raise InvalidTorrentFileException(
+            raise InvalidTorrentDataException(
                 self._pos,
                 'Unexpected EOF when reading torrent file'
             )
@@ -121,12 +134,15 @@ class TorrentFileParser(object):
 
     def _dict_items_generator(self):
         while True:
-            try:
-                k = self._next_element()
-            except InvalidTorrentFileException:
+            k = self._next_element()
+            if k is _END:
                 return
             if k == 'pieces':
-                v = self._pieces()
+                v = self._next_hash()
+            elif k == 'ed2k':
+                v = self._next_hash(16, False)
+            elif k == 'filehash':
+                v = self._next_hash(20, False)
             else:
                 v = self._next_element()
             if k == 'encoding':
@@ -141,9 +157,8 @@ class TorrentFileParser(object):
 
     def _list_items_generator(self):
         while True:
-            try:
-                element = self._next_element()
-            except InvalidTorrentFileException:
+            element = self._next_element()
+            if element is _END:
                 return
             yield element
 
@@ -156,7 +171,7 @@ class TorrentFileParser(object):
         while char != end:
             # noinspection PyTypeChecker
             if not b'0' <= char <= b'9':
-                raise InvalidTorrentFileException(self._pos)
+                raise InvalidTorrentDataException(self._pos - 1)
             value = value * 10 + int(char) - int(b'0')
             char = self._read_byte(1)
         return value
@@ -165,7 +180,13 @@ class TorrentFileParser(object):
         length = self._next_int(b':')
         raw = self._read_byte(length)
         if decode:
-            string = raw.decode(self._encoding)
+            try:
+                string = raw.decode(self._encoding)
+            except UnicodeDecodeError as e:
+                raise InvalidTorrentDataException(
+                    self._pos - length + e.start,
+                    "Fail to decode string at pos {pos} using " + e.encoding
+                )
             return string
         return raw
 
@@ -173,17 +194,25 @@ class TorrentFileParser(object):
     def __to_hex(v):
         return hex(ord(v) if isinstance(v, str) else v)[2:].rjust(2, str(0))
 
-    def _pieces(self):
+    def _next_hash(self, p_len=20, need_list=True):
         raw = self._next_string(decode=False)
-        if len(raw) % 20 != 0:
-            raise InvalidTorrentFileException(self._pos)
-        return [
+        if len(raw) % p_len != 0:
+            raise InvalidTorrentDataException(
+                self._pos - len(raw), "Hash bit length not match at pos {pos}"
+            )
+        res = [
             ''.join([self.__to_hex(c) for c in h])
-            for h in (raw[x:x+20] for x in range(0, len(raw), 20))
+            for h in (raw[x:x+p_len] for x in range(0, len(raw), p_len))
         ]
+        if len(res) == 0 and not need_list:
+            return ''
+        if len(res) == 1 and not need_list:
+            return res[0]
+        return res
 
-    def _next_end(self):
-        raise InvalidTorrentFileException(self._pos)
+    @staticmethod
+    def _next_end():
+        return _END
 
     def _next_type(self):
         for (element_type, indicator) in self.TYPES:
@@ -192,7 +221,7 @@ class TorrentFileParser(object):
             if indicator == char:
                 return element_type
             self._seek_back(indicator_length)
-        raise InvalidTorrentFileException(self._pos)
+        raise InvalidTorrentDataException(self._pos)
 
     def _type_to_func(self, t):
         return getattr(self, '_next_' + t)
@@ -229,6 +258,8 @@ def __main():
     parser.add_argument('--ascii', '-a', action='store_true', default=False,
                         help='ensure output json use ascii char, '
                              'escape other char use \\u')
+    parser.add_argument('--coding', '-c', default='utf-8',
+                        help='string encoding, default utf-8')
     parser.add_argument('--version', '-v', action='store_true', default=False,
                         help='print version and exit')
     args = parser.parse_args()
@@ -245,11 +276,11 @@ def __main():
         else:
             target_file = open(args.file, 'rb')
     except FileNotFoundError:
-        sys.stderr.write('Unable to find file {}\n'.format(args.file))
+        sys.stderr.write('File "{}" not exist\n'.format(args.file))
         exit(1)
 
     # noinspection PyUnboundLocalVariable
-    data = TorrentFileParser(target_file, not args.dict).parse()
+    data = TorrentFileParser(target_file, not args.dict, args.coding).parse()
 
     data = json.dumps(
         data, ensure_ascii=args.ascii,