-# objects.py -- Acces to base git objects
+# objects.py -- Access to base git objects
# Copyright (C) 2007 James Westby <jw+debian@jameswestby.net>
-# Copyright (C) 2008 Jelmer Vernooij <jelmer@samba.org>
-# The header parsing code is based on that from git itself, which is
-# Copyright (C) 2005 Linus Torvalds
-# and licensed under v2 of the GPL.
-#
+# Copyright (C) 2008-2009 Jelmer Vernooij <jelmer@samba.org>
+#
# This program is free software; you can redistribute it and/or
# modify it under the terms of the GNU General Public License
# as published by the Free Software Foundation; version 2
# of the License or (at your option) a later version of the License.
-#
+#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
-#
+#
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
# MA 02110-1301, USA.
+
+"""Access to base git objects."""
+
+
+import binascii
+from cStringIO import (
+ StringIO,
+ )
import mmap
import os
-import sha
+import stat
+import time
import zlib
-from errors import (NotCommitError,
- NotTreeError,
- NotBlobError,
- )
+from dulwich.errors import (
+ NotBlobError,
+ NotCommitError,
+ NotTreeError,
+ )
+from dulwich.file import GitFile
+from dulwich.misc import (
+ make_sha,
+ )
BLOB_ID = "blob"
TAG_ID = "tag"
OBJECT_ID = "object"
TYPE_ID = "type"
TAGGER_ID = "tagger"
+ENCODING_ID = "encoding"
+
+S_IFGITLINK = 0160000
+
+def S_ISGITLINK(m):
+ return (stat.S_IFMT(m) == S_IFGITLINK)
def _decompress(string):
dcomp = zlib.decompressobj()
dcomped += dcomp.flush()
return dcomped
+
def sha_to_hex(sha):
- """Takes a string and returns the hex of the sha within"""
- hexsha = ''
- for c in sha:
- hexsha += "%02x" % ord(c)
- assert len(hexsha) == 40, "Incorrect length of sha1 string: %d" % \
- len(hexsha)
- return hexsha
+ """Takes a string and returns the hex of the sha within"""
+ hexsha = binascii.hexlify(sha)
+ assert len(hexsha) == 40, "Incorrect length of sha1 string: %d" % hexsha
+ return hexsha
+
def hex_to_sha(hex):
- """Takes a hex sha and returns a binary sha"""
- sha = ''
- for i in range(0, len(hex), 2):
- sha += chr(int(hex[i:i+2], 16))
- assert len(sha) == 20, "Incorrent length of sha1: %d" % len(sha)
- return sha
+ """Takes a hex sha and returns a binary sha"""
+ assert len(hex) == 40, "Incorrent length of hexsha: %s" % hex
+ return binascii.unhexlify(hex)
+
+
+def serializable_property(name, docstring=None):
+ def set(obj, value):
+ obj._ensure_parsed()
+ setattr(obj, "_"+name, value)
+ obj._needs_serialization = True
+ def get(obj):
+ obj._ensure_parsed()
+ return getattr(obj, "_"+name)
+ return property(get, set, doc=docstring)
+
class ShaFile(object):
- """A git SHA file."""
-
- @classmethod
- def _parse_legacy_object(cls, map):
- """Parse a legacy object, creating it and setting object._text"""
- text = _decompress(map)
- object = None
- for posstype in type_map.keys():
- if text.startswith(posstype):
- object = type_map[posstype]()
- text = text[len(posstype):]
- break
- assert object is not None, "%s is not a known object type" % text[:9]
- assert text[0] == ' ', "%s is not a space" % text[0]
- text = text[1:]
- size = 0
- i = 0
- while text[0] >= '0' and text[0] <= '9':
- if i > 0 and size == 0:
- assert False, "Size is not in canonical format"
- size = (size * 10) + int(text[0])
- text = text[1:]
- i += 1
- object._size = size
- assert text[0] == "\0", "Size not followed by null"
- text = text[1:]
- object._text = text
- return object
-
- def as_raw_string(self):
- return self._num_type, self._text
-
- @classmethod
- def _parse_object(cls, map):
- """Parse a new style object , creating it and setting object._text"""
- used = 0
- byte = ord(map[used])
- used += 1
- num_type = (byte >> 4) & 7
- try:
- object = num_type_map[num_type]()
- except KeyError:
- assert False, "Not a known type: %d" % num_type
- while((byte & 0x80) != 0):
- byte = ord(map[used])
- used += 1
- raw = map[used:]
- object._text = _decompress(raw)
- return object
-
- @classmethod
- def _parse_file(cls, map):
- word = (ord(map[0]) << 8) + ord(map[1])
- if ord(map[0]) == 0x78 and (word % 31) == 0:
- return cls._parse_legacy_object(map)
- else:
- return cls._parse_object(map)
-
- def __init__(self):
- """Don't call this directly"""
-
- def _parse_text(self):
- """For subclasses to do initialisation time parsing"""
-
- @classmethod
- def from_file(cls, filename):
- """Get the contents of a SHA file on disk"""
- size = os.path.getsize(filename)
- f = open(filename, 'rb')
- try:
- map = mmap.mmap(f.fileno(), size, access=mmap.ACCESS_READ)
- shafile = cls._parse_file(map)
- shafile._parse_text()
- return shafile
- finally:
- f.close()
-
- @classmethod
- def from_raw_string(cls, type, string):
- """Creates an object of the indicated type from the raw string given.
-
- Type is the numeric type of an object. String is the raw uncompressed
- contents.
- """
- real_class = num_type_map[type]
- obj = real_class()
- obj._num_type = type
- obj._text = string
- obj._parse_text()
- return obj
-
- def _header(self):
- return "%s %lu\0" % (self._type, len(self._text))
-
- def crc32(self):
- return zlib.crc32(self._text)
-
- def sha(self):
- """The SHA1 object that is the name of this object."""
- ressha = sha.new()
- ressha.update(self._header())
- ressha.update(self._text)
- return ressha
-
- @property
- def id(self):
- return self.sha().hexdigest()
-
- @property
- def type(self):
- return self._num_type
-
- def __repr__(self):
- return "<%s %s>" % (self.__class__.__name__, self.id)
-
- def __eq__(self, other):
- """Return true id the sha of the two objects match.
-
- The __le__ etc methods aren't overriden as they make no sense,
- certainly at this level.
- """
- return self.sha().digest() == other.sha().digest()
+ """A git SHA file."""
+
+ @classmethod
+ def _parse_legacy_object(cls, map):
+ """Parse a legacy object, creating it and setting object._text"""
+ text = _decompress(map)
+ object = None
+ for posstype in type_map.keys():
+ if text.startswith(posstype):
+ object = type_map[posstype]()
+ text = text[len(posstype):]
+ break
+ assert object is not None, "%s is not a known object type" % text[:9]
+ assert text[0] == ' ', "%s is not a space" % text[0]
+ text = text[1:]
+ size = 0
+ i = 0
+ while text[0] >= '0' and text[0] <= '9':
+ if i > 0 and size == 0:
+ raise AssertionError("Size is not in canonical format")
+ size = (size * 10) + int(text[0])
+ text = text[1:]
+ i += 1
+ object._size = size
+ assert text[0] == "\0", "Size not followed by null"
+ text = text[1:]
+ object.set_raw_string(text)
+ return object
+
+ def as_legacy_object(self):
+ text = self.as_raw_string()
+ return zlib.compress("%s %d\0%s" % (self._type, len(text), text))
+
+ def as_raw_string(self):
+ if self._needs_serialization:
+ self.serialize()
+ return self._text
+
+ def __str__(self):
+ return self.as_raw_string()
+
+ def __hash__(self):
+ return hash(self.id)
+
+ def as_pretty_string(self):
+ return self.as_raw_string()
+
+ def _ensure_parsed(self):
+ if self._needs_parsing:
+ self._parse_text()
+
+ def set_raw_string(self, text):
+ if type(text) != str:
+ raise TypeError(text)
+ self._text = text
+ self._sha = None
+ self._needs_parsing = True
+ self._needs_serialization = False
+
+ @classmethod
+ def _parse_object(cls, map):
+ """Parse a new style object , creating it and setting object._text"""
+ used = 0
+ byte = ord(map[used])
+ used += 1
+ num_type = (byte >> 4) & 7
+ try:
+ object = num_type_map[num_type]()
+ except KeyError:
+ raise AssertionError("Not a known type: %d" % num_type)
+ while (byte & 0x80) != 0:
+ byte = ord(map[used])
+ used += 1
+ raw = map[used:]
+ object.set_raw_string(_decompress(raw))
+ return object
+
+ @classmethod
+ def _parse_file(cls, map):
+ word = (ord(map[0]) << 8) + ord(map[1])
+ if ord(map[0]) == 0x78 and (word % 31) == 0:
+ return cls._parse_legacy_object(map)
+ else:
+ return cls._parse_object(map)
+
+ def __init__(self):
+ """Don't call this directly"""
+ self._sha = None
+
+ def _parse_text(self):
+ """For subclasses to do initialisation time parsing"""
+
+ @classmethod
+ def from_file(cls, filename):
+ """Get the contents of a SHA file on disk"""
+ size = os.path.getsize(filename)
+ f = GitFile(filename, 'rb')
+ try:
+ map = mmap.mmap(f.fileno(), size, access=mmap.ACCESS_READ)
+ shafile = cls._parse_file(map)
+ return shafile
+ finally:
+ f.close()
+
+ @classmethod
+ def from_raw_string(cls, type, string):
+ """Creates an object of the indicated type from the raw string given.
+
+ Type is the numeric type of an object. String is the raw uncompressed
+ contents.
+ """
+ real_class = num_type_map[type]
+ obj = real_class()
+ obj.type = type
+ obj.set_raw_string(string)
+ return obj
+
+ @classmethod
+ def from_string(cls, string):
+ """Create a blob from a string."""
+ shafile = cls()
+ shafile.set_raw_string(string)
+ return shafile
+
+ def _raw_length(self):
+ """Returns the length of the raw string of this object."""
+ return len(self.as_raw_string())
+
+ def _header(self):
+ return "%s %lu\0" % (self._type, self._raw_length())
+
+ def _make_sha(self):
+ ret = make_sha()
+ ret.update(self._header())
+ ret.update(self.as_raw_string())
+ return ret
+
+ def sha(self):
+ """The SHA1 object that is the name of this object."""
+ if self._needs_serialization or self._sha is None:
+ self._sha = self._make_sha()
+ return self._sha
+
+ @property
+ def id(self):
+ return self.sha().hexdigest()
+
+ def get_type(self):
+ return self._num_type
+
+ def set_type(self, type):
+ self._num_type = type
+
+ type = property(get_type, set_type)
+
+ def __repr__(self):
+ return "<%s %s>" % (self.__class__.__name__, self.id)
+
+ def __ne__(self, other):
+ return self.id != other.id
+
+ def __eq__(self, other):
+ """Return true id the sha of the two objects match.
+
+ The __le__ etc methods aren't overriden as they make no sense,
+ certainly at this level.
+ """
+ return self.id == other.id
class Blob(ShaFile):
- """A Git Blob object."""
+ """A Git Blob object."""
- _type = BLOB_ID
- _num_type = 3
+ _type = BLOB_ID
+ _num_type = 3
- @property
- def data(self):
- """The text contained within the blob object."""
- return self._text
+ def __init__(self):
+ super(Blob, self).__init__()
+ self._chunked = []
+ self._text = ""
+ self._needs_parsing = False
+ self._needs_serialization = False
- @classmethod
- def from_file(cls, filename):
- blob = ShaFile.from_file(filename)
- if blob._type != cls._type:
- raise NotBlobError(filename)
- return blob
+ def _get_data(self):
+ if self._needs_serialization:
+ self.serialize()
+ return self._text
- @classmethod
- def from_string(cls, string):
- """Create a blob from a string."""
- shafile = cls()
- shafile._text = string
- return shafile
+ def _set_data(self, data):
+ self._text = data
+ self._needs_parsing = True
+ self._needs_serialization = False
+ data = property(_get_data, _set_data,
+ "The text contained within the blob object.")
-class Tag(ShaFile):
- """A Git Tag object."""
-
- _type = TAG_ID
-
- @classmethod
- def from_file(cls, filename):
- blob = ShaFile.from_file(filename)
- if blob._type != cls._type:
- raise NotBlobError(filename)
- return blob
-
- @classmethod
- def from_string(cls, string):
- """Create a blob from a string."""
- shafile = cls()
- shafile._text = string
- return shafile
-
- def _parse_text(self):
- """Grab the metadata attached to the tag"""
- text = self._text
- count = 0
- assert text.startswith(OBJECT_ID), "Invalid tag object, " \
- "must start with %s" % OBJECT_ID
- count += len(OBJECT_ID)
- assert text[count] == ' ', "Invalid tag object, " \
- "%s must be followed by space not %s" % (OBJECT_ID, text[count])
- count += 1
- self._object_sha = text[count:count+40]
- count += 40
- assert text[count] == '\n', "Invalid tag object, " \
- "%s sha must be followed by newline" % OBJECT_ID
- count += 1
- assert text[count:].startswith(TYPE_ID), "Invalid tag object, " \
- "%s sha must be followed by %s" % (OBJECT_ID, TYPE_ID)
- count += len(TYPE_ID)
- assert text[count] == ' ', "Invalid tag object, " \
- "%s must be followed by space not %s" % (TAG_ID, text[count])
- count += 1
- self._object_type = ""
- while text[count] != '\n':
- self._object_type += text[count]
- count += 1
- count += 1
- assert self._object_type in (COMMIT_ID, BLOB_ID, TREE_ID, TAG_ID), "Invalid tag object, " \
- "unexpected object type %s" % self._object_type
- self._object_type = type_map[self._object_type]
-
- assert text[count:].startswith(TAG_ID), "Invalid tag object, " \
- "object type must be followed by %s" % (TAG_ID)
- count += len(TAG_ID)
- assert text[count] == ' ', "Invalid tag object, " \
- "%s must be followed by space not %s" % (TAG_ID, text[count])
- count += 1
- self._name = ""
- while text[count] != '\n':
- self._name += text[count]
- count += 1
- count += 1
-
- assert text[count:].startswith(TAGGER_ID), "Invalid tag object, " \
- "%s must be followed by %s" % (TAG_ID, TAGGER_ID)
- count += len(TAGGER_ID)
- assert text[count] == ' ', "Invalid tag object, " \
- "%s must be followed by space not %s" % (TAGGER_ID, text[count])
- count += 1
- self._tagger = ""
- while text[count] != '>':
- assert text[count] != '\n', "Malformed tagger information"
- self._tagger += text[count]
- count += 1
- self._tagger += text[count]
- count += 1
- assert text[count] == ' ', "Invalid tag object, " \
- "tagger information must be followed by space not %s" % text[count]
- count += 1
- self._tag_time = int(text[count:count+10])
- while text[count] != '\n':
- count += 1
- count += 1
- assert text[count] == '\n', "There must be a new line after the headers"
- count += 1
- self._message = text[count:]
-
- @property
- def object(self):
- """Returns the object pointed by this tag, represented as a tuple(type, sha)"""
- return (self._object_type, self._object_sha)
-
- @property
- def name(self):
- """Returns the name of this tag"""
- return self._name
-
- @property
- def tagger(self):
- """Returns the name of the person who created this tag"""
- return self._tagger
-
- @property
- def tag_time(self):
- """Returns the creation timestamp of the tag.
-
- Returns it as the number of seconds since the epoch"""
- return self._tag_time
-
- @property
- def message(self):
- """Returns the message attached to this tag"""
- return self._message
+ def _get_chunked(self):
+ self._ensure_parsed()
+ return self._chunked
+ def _set_chunked(self, chunks):
+ self._chunked = chunks
+ self._needs_serialization = True
-class Tree(ShaFile):
- """A Git tree object"""
+ chunked = property(_get_chunked, _set_chunked,
+ "The text within the blob object, as chunks (not necessarily lines).")
- _type = TREE_ID
- _num_type = 2
+ def _parse_text(self):
+ self._chunked = [self._text]
- def __init__(self):
- self._entries = []
+ def serialize(self):
+ self._text = "".join(self._chunked)
- @classmethod
- def from_file(cls, filename):
- tree = ShaFile.from_file(filename)
- if tree._type != cls._type:
- raise NotTreeError(filename)
- return tree
+ def _raw_length(self):
+ ret = 0
+ for chunk in self.chunked:
+ ret += len(chunk)
+ return ret
- def add(self, mode, name, hexsha):
- self._entries.append((mode, name, hexsha))
+ def _make_sha(self):
+ ret = make_sha()
+ ret.update(self._header())
+ for chunk in self._chunked:
+ ret.update(chunk)
+ return ret
- def entries(self):
- """Return a list of tuples describing the tree entries"""
- return self._entries
+ @classmethod
+ def from_file(cls, filename):
+ blob = ShaFile.from_file(filename)
+ if blob._type != cls._type:
+ raise NotBlobError(filename)
+ return blob
- def _parse_text(self):
- """Grab the entries in the tree"""
+
+class Tag(ShaFile):
+ """A Git Tag object."""
+
+ _type = TAG_ID
+ _num_type = 4
+
+ def __init__(self):
+ super(Tag, self).__init__()
+ self._needs_parsing = False
+ self._needs_serialization = True
+
+ @classmethod
+ def from_file(cls, filename):
+ blob = ShaFile.from_file(filename)
+ if blob._type != cls._type:
+ raise NotBlobError(filename)
+ return blob
+
+ @classmethod
+ def from_string(cls, string):
+ """Create a blob from a string."""
+ shafile = cls()
+ shafile.set_raw_string(string)
+ return shafile
+
+ def serialize(self):
+ f = StringIO()
+ f.write("%s %s\n" % (OBJECT_ID, self._object_sha))
+ f.write("%s %s\n" % (TYPE_ID, num_type_map[self._object_type]._type))
+ f.write("%s %s\n" % (TAG_ID, self._name))
+ if self._tagger:
+ if self._tag_time is None:
+ f.write("%s %s\n" % (TAGGER_ID, self._tagger))
+ else:
+ f.write("%s %s %d %s\n" % (TAGGER_ID, self._tagger, self._tag_time, format_timezone(self._tag_timezone)))
+ f.write("\n") # To close headers
+ f.write(self._message)
+ self._text = f.getvalue()
+ self._needs_serialization = False
+
+ def _parse_text(self):
+ """Grab the metadata attached to the tag"""
+ self._tagger = None
+ f = StringIO(self._text)
+ for l in f:
+ l = l.rstrip("\n")
+ if l == "":
+ break # empty line indicates end of headers
+ (field, value) = l.split(" ", 1)
+ if field == OBJECT_ID:
+ self._object_sha = value
+ elif field == TYPE_ID:
+ self._object_type = type_map[value]
+ elif field == TAG_ID:
+ self._name = value
+ elif field == TAGGER_ID:
+ try:
+ sep = value.index("> ")
+ except ValueError:
+ self._tagger = value
+ self._tag_time = None
+ self._tag_timezone = None
+ else:
+ self._tagger = value[0:sep+1]
+ (timetext, timezonetext) = value[sep+2:].rsplit(" ", 1)
+ try:
+ self._tag_time = int(timetext)
+ except ValueError: #Not a unix timestamp
+ self._tag_time = time.strptime(timetext)
+ self._tag_timezone = parse_timezone(timezonetext)
+ else:
+ raise AssertionError("Unknown field %s" % field)
+ self._message = f.read()
+ self._needs_parsing = False
+
+ def _get_object(self):
+ """Returns the object pointed by this tag, represented as a tuple(type, sha)"""
+ self._ensure_parsed()
+ return (self._object_type, self._object_sha)
+
+ def _set_object(self, value):
+ self._ensure_parsed()
+ (self._object_type, self._object_sha) = value
+ self._needs_serialization = True
+
+ object = property(_get_object, _set_object)
+
+ name = serializable_property("name", "The name of this tag")
+ tagger = serializable_property("tagger",
+ "Returns the name of the person who created this tag")
+ tag_time = serializable_property("tag_time",
+ "The creation timestamp of the tag. As the number of seconds since the epoch")
+ tag_timezone = serializable_property("tag_timezone",
+ "The timezone that tag_time is in.")
+ message = serializable_property("message", "The message attached to this tag")
+
+
+def parse_tree(text):
+ """Parse a tree text.
+
+ :param text: Serialized text to parse
+ :return: Dictionary with names as keys, (mode, sha) tuples as values
+ """
+ ret = {}
count = 0
- while count < len(self._text):
- mode = 0
- chr = self._text[count]
- while chr != ' ':
- assert chr >= '0' and chr <= '7', "%s is not a valid mode char" % chr
- mode = (mode << 3) + (ord(chr) - ord('0'))
- count += 1
- chr = self._text[count]
- count += 1
- chr = self._text[count]
- name = ''
- while chr != '\0':
- name += chr
- count += 1
- chr = self._text[count]
- count += 1
- chr = self._text[count]
- sha = self._text[count:count+20]
- hexsha = sha_to_hex(sha)
- self.add(mode, name, hexsha)
- count = count + 20
-
- def serialize(self):
- self._text = ""
- for mode, name, hexsha in self._entries:
- self._text += "%04o %s\0%s" % (mode, name, hex_to_sha(hexsha))
+ l = len(text)
+ while count < l:
+ mode_end = text.index(' ', count)
+ mode = int(text[count:mode_end], 8)
+ name_end = text.index('\0', mode_end)
+ name = text[mode_end+1:name_end]
+ count = name_end+21
+ sha = text[name_end+1:count]
+ ret[name] = (mode, sha_to_hex(sha))
+ return ret
+
+
+def serialize_tree(items):
+ """Serialize the items in a tree to a text.
+
+ :param items: Sorted iterable over (name, mode, sha) tuples
+ :return: Serialized tree text
+ """
+ f = StringIO()
+ for name, mode, hexsha in items:
+ f.write("%04o %s\0%s" % (mode, name, hex_to_sha(hexsha)))
+ return f.getvalue()
-class Commit(ShaFile):
- """A git commit object"""
+def sorted_tree_items(entries):
+ """Iterate over a tree entries dictionary in the order in which
+ the items would be serialized.
- _type = COMMIT_ID
- _num_type = 1
+ :param entries: Dictionary mapping names to (mode, sha) tuples
+ :return: Iterator over (name, mode, sha)
+ """
+ def cmp_entry((name1, value1), (name2, value2)):
+ if stat.S_ISDIR(value1[0]):
+ name1 += "/"
+ if stat.S_ISDIR(value2[0]):
+ name2 += "/"
+ return cmp(name1, name2)
+ for name, entry in sorted(entries.iteritems(), cmp=cmp_entry):
+ yield name, entry[0], entry[1]
- def __init__(self):
- self._parents = []
- @classmethod
- def from_file(cls, filename):
- commit = ShaFile.from_file(filename)
- if commit._type != cls._type:
- raise NotCommitError(filename)
- return commit
+class Tree(ShaFile):
+ """A Git tree object"""
+
+ _type = TREE_ID
+ _num_type = 2
+
+ def __init__(self):
+ super(Tree, self).__init__()
+ self._entries = {}
+ self._needs_parsing = False
+ self._needs_serialization = True
+
+ @classmethod
+ def from_file(cls, filename):
+ tree = ShaFile.from_file(filename)
+ if tree._type != cls._type:
+ raise NotTreeError(filename)
+ return tree
+
+ def __contains__(self, name):
+ self._ensure_parsed()
+ return name in self._entries
+
+ def __getitem__(self, name):
+ self._ensure_parsed()
+ return self._entries[name]
+
+ def __setitem__(self, name, value):
+ assert isinstance(value, tuple)
+ assert len(value) == 2
+ self._ensure_parsed()
+ self._entries[name] = value
+ self._needs_serialization = True
+
+ def __delitem__(self, name):
+ self._ensure_parsed()
+ del self._entries[name]
+ self._needs_serialization = True
+
+ def __len__(self):
+ self._ensure_parsed()
+ return len(self._entries)
+
+ def add(self, mode, name, hexsha):
+ assert type(mode) == int
+ assert type(name) == str
+ assert type(hexsha) == str
+ self._ensure_parsed()
+ self._entries[name] = mode, hexsha
+ self._needs_serialization = True
+
+ def entries(self):
+ """Return a list of tuples describing the tree entries"""
+ self._ensure_parsed()
+ # The order of this is different from iteritems() for historical
+ # reasons
+ return [
+ (mode, name, hexsha) for (name, mode, hexsha) in self.iteritems()]
+
+ def iteritems(self):
+ """Iterate over all entries in the order in which they would be
+ serialized.
+
+ :return: Iterator over (name, mode, sha) tuples
+ """
+ self._ensure_parsed()
+ return sorted_tree_items(self._entries)
+
+ def _parse_text(self):
+ """Grab the entries in the tree"""
+ self._entries = parse_tree(self._text)
+ self._needs_parsing = False
+
+ def serialize(self):
+ self._text = serialize_tree(self.iteritems())
+ self._needs_serialization = False
+
+ def as_pretty_string(self):
+ text = ""
+ for name, mode, hexsha in self.iteritems():
+ if mode & stat.S_IFDIR:
+ kind = "tree"
+ else:
+ kind = "blob"
+ text += "%04o %s %s\t%s\n" % (mode, kind, hexsha, name)
+ return text
+
+
+def parse_timezone(text):
+ offset = int(text)
+ signum = (offset < 0) and -1 or 1
+ offset = abs(offset)
+ hours = int(offset / 100)
+ minutes = (offset % 100)
+ return signum * (hours * 3600 + minutes * 60)
+
+
+def format_timezone(offset):
+ if offset % 60 != 0:
+ raise ValueError("Unable to handle non-minute offset.")
+ sign = (offset < 0) and '-' or '+'
+ offset = abs(offset)
+ return '%c%02d%02d' % (sign, offset / 3600, (offset / 60) % 60)
- def _parse_text(self):
- text = self._text
- count = 0
- assert text.startswith(TREE_ID), "Invalid commit object, " \
- "must start with %s" % TREE_ID
- count += len(TREE_ID)
- assert text[count] == ' ', "Invalid commit object, " \
- "%s must be followed by space not %s" % (TREE_ID, text[count])
- count += 1
- self._tree = text[count:count+40]
- count = count + 40
- assert text[count] == "\n", "Invalid commit object, " \
- "tree sha must be followed by newline"
- count += 1
- self._parents = []
- while text[count:].startswith(PARENT_ID):
- count += len(PARENT_ID)
- assert text[count] == ' ', "Invalid commit object, " \
- "%s must be followed by space not %s" % (PARENT_ID, text[count])
- count += 1
- self._parents.append(text[count:count+40])
- count += 40
- assert text[count] == "\n", "Invalid commit object, " \
- "parent sha must be followed by newline"
- count += 1
- self._author = None
- if text[count:].startswith(AUTHOR_ID):
- count += len(AUTHOR_ID)
- assert text[count] == ' ', "Invalid commit object, " \
- "%s must be followed by space not %s" % (AUTHOR_ID, text[count])
- count += 1
- self._author = ''
- while text[count] != '>':
- assert text[count] != '\n', "Malformed author information"
- self._author += text[count]
- count += 1
- self._author += text[count]
- count += 1
- while text[count] != '\n':
- count += 1
- count += 1
- self._committer = None
- if text[count:].startswith(COMMITTER_ID):
- count += len(COMMITTER_ID)
- assert text[count] == ' ', "Invalid commit object, " \
- "%s must be followed by space not %s" % (COMMITTER_ID, text[count])
- count += 1
- self._committer = ''
- while text[count] != '>':
- assert text[count] != '\n', "Malformed committer information"
- self._committer += text[count]
- count += 1
- self._committer += text[count]
- count += 1
- assert text[count] == ' ', "Invalid commit object, " \
- "commiter information must be followed by space not %s" % text[count]
- count += 1
- self._commit_time = int(text[count:count+10])
- while text[count] != '\n':
- count += 1
- count += 1
- assert text[count] == '\n', "There must be a new line after the headers"
- count += 1
- # XXX: There can be an encoding field.
- self._message = text[count:]
-
- def serialize(self):
- self._text = ""
- self._text += "%s %s\n" % (TREE_ID, self._tree)
- for p in self._parents:
- self._text += "%s %s\n" % (PARENT_ID, p)
- self._text += "%s %s %s +0000\n" % (AUTHOR_ID, self._author, str(self._commit_time))
- self._text += "%s %s %s +0000\n" % (COMMITTER_ID, self._committer, str(self._commit_time))
- self._text += "\n" # There must be a new line after the headers
- self._text += self._message
-
- @property
- def tree(self):
- """Returns the tree that is the state of this commit"""
- return self._tree
-
- @property
- def parents(self):
- """Return a list of parents of this commit."""
- return self._parents
-
- @property
- def author(self):
- """Returns the name of the author of the commit"""
- return self._author
-
- @property
- def committer(self):
- """Returns the name of the committer of the commit"""
- return self._committer
-
- @property
- def message(self):
- """Returns the commit message"""
- return self._message
-
- @property
- def commit_time(self):
- """Returns the timestamp of the commit.
-
- Returns it as the number of seconds since the epoch.
- """
- return self._commit_time
+
+class Commit(ShaFile):
+ """A git commit object"""
+
+ _type = COMMIT_ID
+ _num_type = 1
+
+ def __init__(self):
+ super(Commit, self).__init__()
+ self._parents = []
+ self._encoding = None
+ self._needs_parsing = False
+ self._needs_serialization = True
+ self._extra = {}
+
+ @classmethod
+ def from_file(cls, filename):
+ commit = ShaFile.from_file(filename)
+ if commit._type != cls._type:
+ raise NotCommitError(filename)
+ return commit
+
+ def _parse_text(self):
+ self._parents = []
+ self._extra = []
+ self._author = None
+ f = StringIO(self._text)
+ for l in f:
+ l = l.rstrip("\n")
+ if l == "":
+ # Empty line indicates end of headers
+ break
+ (field, value) = l.split(" ", 1)
+ if field == TREE_ID:
+ self._tree = value
+ elif field == PARENT_ID:
+ self._parents.append(value)
+ elif field == AUTHOR_ID:
+ self._author, timetext, timezonetext = value.rsplit(" ", 2)
+ self._author_time = int(timetext)
+ self._author_timezone = parse_timezone(timezonetext)
+ elif field == COMMITTER_ID:
+ self._committer, timetext, timezonetext = value.rsplit(" ", 2)
+ self._commit_time = int(timetext)
+ self._commit_timezone = parse_timezone(timezonetext)
+ elif field == ENCODING_ID:
+ self._encoding = value
+ else:
+ self._extra.append((field, value))
+ self._message = f.read()
+ self._needs_parsing = False
+
+ def serialize(self):
+ f = StringIO()
+ f.write("%s %s\n" % (TREE_ID, self._tree))
+ for p in self._parents:
+ f.write("%s %s\n" % (PARENT_ID, p))
+ f.write("%s %s %s %s\n" % (AUTHOR_ID, self._author, str(self._author_time), format_timezone(self._author_timezone)))
+ f.write("%s %s %s %s\n" % (COMMITTER_ID, self._committer, str(self._commit_time), format_timezone(self._commit_timezone)))
+ if self.encoding:
+ f.write("%s %s\n" % (ENCODING_ID, self.encoding))
+ for k, v in self.extra:
+ if "\n" in k or "\n" in v:
+ raise AssertionError("newline in extra data: %r -> %r" % (k, v))
+ f.write("%s %s\n" % (k, v))
+ f.write("\n") # There must be a new line after the headers
+ f.write(self._message)
+ self._text = f.getvalue()
+ self._needs_serialization = False
+
+ tree = serializable_property("tree", "Tree that is the state of this commit")
+
+ def _get_parents(self):
+ """Return a list of parents of this commit."""
+ self._ensure_parsed()
+ return self._parents
+
+ def _set_parents(self, value):
+ """Set a list of parents of this commit."""
+ self._ensure_parsed()
+ self._needs_serialization = True
+ self._parents = value
+
+ parents = property(_get_parents, _set_parents)
+
+ def _get_extra(self):
+ """Return extra settings of this commit."""
+ self._ensure_parsed()
+ return self._extra
+
+ extra = property(_get_extra)
+
+ author = serializable_property("author",
+ "The name of the author of the commit")
+
+ committer = serializable_property("committer",
+ "The name of the committer of the commit")
+
+ message = serializable_property("message",
+ "The commit message")
+
+ commit_time = serializable_property("commit_time",
+ "The timestamp of the commit. As the number of seconds since the epoch.")
+
+ commit_timezone = serializable_property("commit_timezone",
+ "The zone the commit time is in")
+
+ author_time = serializable_property("author_time",
+ "The timestamp the commit was written. as the number of seconds since the epoch.")
+
+ author_timezone = serializable_property("author_timezone",
+ "Returns the zone the author time is in.")
+
+ encoding = serializable_property("encoding",
+ "Encoding of the commit message.")
type_map = {
- BLOB_ID : Blob,
- TREE_ID : Tree,
- COMMIT_ID : Commit,
- TAG_ID: Tag,
+ BLOB_ID : Blob,
+ TREE_ID : Tree,
+ COMMIT_ID : Commit,
+ TAG_ID: Tag,
}
num_type_map = {
- 0: None,
- 1: Commit,
- 2: Tree,
- 3: Blob,
- 4: Tag,
- # 5 Is reserved for further expansion
+ 0: None,
+ 1: Commit,
+ 2: Tree,
+ 3: Blob,
+ 4: Tag,
+ # 5 Is reserved for further expansion
}
+try:
+ # Try to import C versions
+ from dulwich._objects import parse_tree, sorted_tree_items
+except ImportError:
+ pass
git.samba.org / jelmer / dulwich-libgit2.git / blobdiff