Support for Python3

doc/source/conf.py

@@ -37,8 +37,8 @@
 master_doc = 'index'
 
 # General information about the project.
-project = u'GitDB'
-copyright = u'2011, Sebastian Thiel'
+project = 'GitDB'
+copyright = '2011, Sebastian Thiel'
 
 # The version info for the project you're documenting, acts as replacement for
 # |version| and |release|, also used in various other places throughout the
@@ -171,8 +171,8 @@
 # Grouping the document tree into LaTeX files. List of tuples
 # (source start file, target name, title, author, documentclass [howto/manual]).
 latex_documents = [
-  ('index', 'GitDB.tex', u'GitDB Documentation',
-   u'Sebastian Thiel', 'manual'),
+  ('index', 'GitDB.tex', 'GitDB Documentation',
+   'Sebastian Thiel', 'manual'),
 ]
 
 # The name of an image file (relative to this directory) to place at the top of

gitdb/__init__.py

@@ -32,7 +32,7 @@ def _init_externals():
 
 
 # default imports
-from db import *
-from base import *
-from stream import *
+from .db import *
+from .base import *
+from .stream import *
 

gitdb/base.py

@@ -3,12 +3,12 @@
 # This module is part of GitDB and is released under
 # the New BSD License: http://www.opensource.org/licenses/bsd-license.php
 """Module with basic data structures - they are designed to be lightweight and fast"""
-from util import (
+from .util import (
         bin_to_hex,
         zlib
     )
 
-from fun import (
+from .fun import (
                     type_id_to_type_map,
                     type_to_type_id_map
                 )

gitdb/db/__init__.py

@@ -3,10 +3,10 @@
 # This module is part of GitDB and is released under
 # the New BSD License: http://www.opensource.org/licenses/bsd-license.php
 
-from base import *
-from loose import *
-from mem import *
-from pack import *
-from git import *
-from ref import *
+from .base import *
+from .loose import *
+from .mem import *
+from .pack import *
+from .git import *
+from .ref import *
 

gitdb/db/base.py

@@ -20,6 +20,7 @@
     )
 
 from itertools import chain
+from functools import reduce
 
 
 __all__ = ('ObjectDBR', 'ObjectDBW', 'FileDBBase', 'CompoundDB', 'CachingDB')

gitdb/db/git.py

@@ -2,15 +2,15 @@
 #
 # This module is part of GitDB and is released under
 # the New BSD License: http://www.opensource.org/licenses/bsd-license.php
-from base import (
+from .base import (
                         CompoundDB, 
                         ObjectDBW, 
                         FileDBBase
                     )
 
-from loose import LooseObjectDB
-from pack import PackedDB
-from ref import ReferenceDB
+from .loose import LooseObjectDB
+from .pack import PackedDB
+from .ref import ReferenceDB
 
 from gitdb.util import LazyMixin
 from gitdb.exc import (

gitdb/db/loose.py

@@ -2,7 +2,7 @@
 #
 # This module is part of GitDB and is released under
 # the New BSD License: http://www.opensource.org/licenses/bsd-license.php
-from base import (
+from .base import (
                         FileDBBase, 
                         ObjectDBR, 
                         ObjectDBW
@@ -69,9 +69,9 @@ class LooseObjectDB(FileDBBase, ObjectDBR, ObjectDBW):
 
     # On windows we need to keep it writable, otherwise it cannot be removed
     # either
-    new_objects_mode = 0444
+    new_objects_mode = 0o444
     if os.name == 'nt':
-        new_objects_mode = 0644
+        new_objects_mode = 0o644
 
 
     def __init__(self, root_path):
@@ -133,7 +133,7 @@ def _map_loose_object(self, sha):
         db_path = self.db_path(self.object_path(bin_to_hex(sha)))
         try:
             return file_contents_ro_filepath(db_path, flags=self._fd_open_flags)
-        except OSError,e:
+        except OSError as e:
             if e.errno != ENOENT:
                 # try again without noatime
                 try:
@@ -200,7 +200,7 @@ def store(self, istream):
                 if istream.binsha is not None:
                     # copy as much as possible, the actual uncompressed item size might
                     # be smaller than the compressed version
-                    stream_copy(istream.read, writer.write, sys.maxint, self.stream_chunk_size)
+                    stream_copy(istream.read, writer.write, sys.maxsize, self.stream_chunk_size)
                 else:
                     # write object with header, we have to make a new one
                     write_object(istream.type, istream.size, istream.read, writer.write,

gitdb/db/mem.py

@@ -3,8 +3,8 @@
 # This module is part of GitDB and is released under
 # the New BSD License: http://www.opensource.org/licenses/bsd-license.php
 """Contains the MemoryDatabase implementation"""
-from loose import LooseObjectDB
-from base import (
+from .loose import LooseObjectDB
+from .base import (
                         ObjectDBR, 
                         ObjectDBW
                     )
@@ -23,7 +23,7 @@
                             DecompressMemMapReader,
                         )
 
-from cStringIO import StringIO
+from io import StringIO
 
 __all__ = ("MemoryDB", )
 
@@ -85,7 +85,7 @@ def size(self):
         return len(self._cache)
 
     def sha_iter(self):
-        return self._cache.iterkeys()
+        return iter(self._cache.keys())
 
 
     #{ Interface 

gitdb/db/pack.py

@@ -3,7 +3,7 @@
 # This module is part of GitDB and is released under
 # the New BSD License: http://www.opensource.org/licenses/bsd-license.php
 """Module containing a database to deal with packs"""
-from base import (
+from .base import (
                         FileDBBase, 
                         ObjectDBR, 
                         CachingDB
@@ -21,6 +21,7 @@
 
 import os
 import glob
+from functools import reduce
 
 __all__ = ('PackedDB', )
 
@@ -104,7 +105,7 @@ def sha_iter(self):
         for entity in self.entities():
             index = entity.index()
             sha_by_index = index.sha
-            for index in xrange(index.size()):
+            for index in range(index.size()):
                 yield sha_by_index(index)
             # END for each index
         # END for each entity

gitdb/db/ref.py

@@ -2,7 +2,7 @@
 #
 # This module is part of GitDB and is released under
 # the New BSD License: http://www.opensource.org/licenses/bsd-license.php
-from base import (
+from .base import (
                     CompoundDB,
                 )
 
@@ -33,7 +33,7 @@ def _update_dbs_from_ref_file(self):
         dbcls = self.ObjectDBCls
         if dbcls is None:
             # late import
-            from git import GitDB
+            from .git import GitDB
             dbcls = GitDB
         # END get db type
 
@@ -68,7 +68,7 @@ def _update_dbs_from_ref_file(self):
                     db.databases()
                 # END verification
                 self._dbs.append(db)
-            except Exception, e:
+            except Exception as e:
                 # ignore invalid paths or issues
                 pass
         # END for each path to add

gitdb/exc.py

@@ -3,7 +3,7 @@
 # This module is part of GitDB and is released under
 # the New BSD License: http://www.opensource.org/licenses/bsd-license.php
 """Module with common exceptions"""
-from util import to_hex_sha
+from .util import to_hex_sha
 
 class ODBError(Exception):
     """All errors thrown by the object database"""

gitdb/fun.py

@@ -6,17 +6,18 @@
 Keeping this code separate from the beginning makes it easier to out-source
 it into c later, if required"""
 
-from exc import (
+from .exc import (
     BadObjectType
     )
 
-from util import zlib
+from .util import zlib
+from functools import reduce
 decompressobj = zlib.decompressobj
 
 import mmap
-from itertools import islice, izip
+from itertools import islice
 
-from cStringIO import StringIO
+from io import StringIO
 
 # INVARIANTS
 OFS_DELTA = 6
@@ -249,7 +250,7 @@ def compress(self):
                 #if first_data_index is not None:
                     nd = StringIO()                     # new data
                     so = self[first_data_index].to      # start offset in target buffer
-                    for x in xrange(first_data_index, i-1):
+                    for x in range(first_data_index, i-1):
                         xdc = self[x]
                         nd.write(xdc.data[:xdc.ts])
                     # END collect data
@@ -296,10 +297,10 @@ def check_integrity(self, target_size=-1):
 
         left = islice(self, 0, len(self)-1)
         right = iter(self)
-        right.next()
+        next(right)
         # this is very pythonic - we might have just use index based access here, 
         # but this could actually be faster
-        for lft,rgt in izip(left, right):
+        for lft,rgt in zip(left, right):
             assert lft.rbound() == rgt.to
             assert lft.to + lft.ts == rgt.to
         # END for each pair
@@ -380,7 +381,7 @@ def is_loose_object(m):
     """
     :return: True the file contained in memory map m appears to be a loose object.
         Only the first two bytes are needed"""
-    b0, b1 = map(ord, m[:2])
+    b0, b1 = list(map(ord, m[:2]))
     word = (b0 << 8) + b1
     return b0 == 0x78 and (word % 31) == 0
 

gitdb/pack.py

@@ -8,15 +8,15 @@
                         UnsupportedOperation,
                         ParseError
                         )
-from util import (
+from .util import (
                     zlib,
                     mman,
                     LazyMixin,
                     unpack_from,
                     bin_to_hex,
                     )
 
-from fun import (
+from .fun import (
                     create_pack_object_header,
                     pack_object_header_info,
                     is_equal_canonical_sha,
@@ -36,7 +36,7 @@
     pass
 # END try c module
 
-from base import (      # Amazing !
+from .base import (      # Amazing !
                         OInfo,
                         OStream,
                         OPackInfo,
@@ -45,7 +45,7 @@
                         ODeltaPackInfo,
                         ODeltaPackStream,
                     )
-from stream import (
+from .stream import (
                         DecompressMemMapReader,
                         DeltaApplyReader,
                         Sha1Writer,
@@ -60,7 +60,7 @@
 
 from binascii import crc32
 
-from itertools import izip
+
 import tempfile
 import array
 import os
@@ -200,7 +200,7 @@ def write(self, pack_sha, write):
         for t in self._objs:
             tmplist[ord(t[0][0])] += 1
         #END prepare fanout
-        for i in xrange(255):
+        for i in range(255):
             v = tmplist[i]
             sha_write(pack('>L', v))
             tmplist[i+1] += v
@@ -407,7 +407,7 @@ def offsets(self):
                 a.byteswap()
             return a
         else:
-            return tuple(self.offset(index) for index in xrange(self.size()))
+            return tuple(self.offset(index) for index in range(self.size()))
         # END handle version
 
     def sha_to_index(self, sha):
@@ -673,8 +673,8 @@ def _set_cache_(self, attr):
         else:
             iter_offsets = iter(offsets_sorted)
             iter_offsets_plus_one = iter(offsets_sorted)
-            iter_offsets_plus_one.next()
-            consecutive = izip(iter_offsets, iter_offsets_plus_one)
+            next(iter_offsets_plus_one)
+            consecutive = zip(iter_offsets, iter_offsets_plus_one)
 
             offset_map = dict(consecutive)
 
@@ -694,7 +694,7 @@ def _iter_objects(self, as_stream):
         """Iterate over all objects in our index and yield their OInfo or OStream instences"""
         _sha = self._index.sha
         _object = self._object
-        for index in xrange(self._index.size()):
+        for index in range(self._index.size()):
             yield _object(_sha(index), as_stream, index)
         # END for each index
 

gitdb/stream.py

@@ -3,12 +3,12 @@
 # This module is part of GitDB and is released under
 # the New BSD License: http://www.opensource.org/licenses/bsd-license.php
 
-from cStringIO import StringIO
+from io import StringIO
 import errno
 import mmap
 import os
 
-from fun import (
+from .fun import (
                     msb_size,
                     stream_copy, 
                     apply_delta_data,
@@ -17,7 +17,7 @@
                     delta_types
                 )
 
-from util import (
+from .util import (
         allocate_memory,
         LazyMixin,
         make_sha,
@@ -417,7 +417,7 @@ def _set_cache_brute_(self, attr):
         # For the actual copying, we use a seek and write pattern of buffer
         # slices.
         final_target_size = None
-        for (dbuf, offset, src_size, target_size), dstream in reversed(zip(buffer_info_list, self._dstreams)):
+        for (dbuf, offset, src_size, target_size), dstream in reversed(list(zip(buffer_info_list, self._dstreams))):
             # allocate a buffer to hold all delta data - fill in the data for 
             # fast access. We do this as we know that reading individual bytes
             # from our stream would be slower than necessary ( although possible )