summaryrefslogtreecommitdiff
path: root/git/util.py
diff options
context:
space:
mode:
authorSebastian Thiel <byronimo@gmail.com>2011-06-07 21:36:42 +0200
committerSebastian Thiel <byronimo@gmail.com>2011-06-07 21:36:42 +0200
commit58a930a632c867b65b9a3802e2f4190cf32e33ee (patch)
tree95b1311a3a4bfcdf4c2dba66f360e6985184013e /git/util.py
parenta98e0af511b728030c12bf8633b077866bb74e47 (diff)
parentf6897c78be5a5530129df50742cb6cabfb8609c9 (diff)
downloadgitpython-58a930a632c867b65b9a3802e2f4190cf32e33ee.tar.gz
Merge branch 'gitdbmerger'
Diffstat (limited to 'git/util.py')
-rw-r--r--git/util.py904
1 files changed, 577 insertions, 327 deletions
diff --git a/git/util.py b/git/util.py
index 7cbef07f..6009e158 100644
--- a/git/util.py
+++ b/git/util.py
@@ -4,28 +4,149 @@
# This module is part of GitPython and is released under
# the BSD License: http://www.opensource.org/licenses/bsd-license.php
+import platform
+import binascii
import os
-import re
+import mmap
import sys
+import errno
+import re
import time
import tempfile
-import platform
-
-from gitdb.util import (
- make_sha,
- LockedFD,
- file_contents_ro,
- LazyMixin,
- to_hex_sha,
- to_bin_sha
- )
__all__ = ( "stream_copy", "join_path", "to_native_path_windows", "to_native_path_linux",
"join_path_native", "Stats", "IndexFileSHA1Writer", "Iterable", "IterableList",
"BlockingLockFile", "LockFile", 'Actor', 'get_user_id', 'assure_directory_exists',
- 'RemoteProgress')
+ 'RepoAliasMixin', 'LockedFD', 'LazyMixin' )
+
+from cStringIO import StringIO
+
+# in py 2.4, StringIO is only StringI, without write support.
+# Hence we must use the python implementation for this
+if sys.version_info[1] < 5:
+ from StringIO import StringIO
+# END handle python 2.4
+
+try:
+ import async.mod.zlib as zlib
+except ImportError:
+ import zlib
+# END try async zlib
+
+from async import ThreadPool
+
+try:
+ import hashlib
+except ImportError:
+ import sha
+
+try:
+ from struct import unpack_from
+except ImportError:
+ from struct import unpack, calcsize
+ __calcsize_cache = dict()
+ def unpack_from(fmt, data, offset=0):
+ try:
+ size = __calcsize_cache[fmt]
+ except KeyError:
+ size = calcsize(fmt)
+ __calcsize_cache[fmt] = size
+ # END exception handling
+ return unpack(fmt, data[offset : offset + size])
+ # END own unpack_from implementation
+
+
+#{ Globals
+
+# A pool distributing tasks, initially with zero threads, hence everything
+# will be handled in the main thread
+pool = ThreadPool(0)
+
+#} END globals
+
+
+#{ Aliases
+
+hex_to_bin = binascii.a2b_hex
+bin_to_hex = binascii.b2a_hex
+
+# errors
+ENOENT = errno.ENOENT
+
+# os shortcuts
+exists = os.path.exists
+mkdir = os.mkdir
+chmod = os.chmod
+isdir = os.path.isdir
+isfile = os.path.isfile
+rename = os.rename
+remove = os.remove
+dirname = os.path.dirname
+basename = os.path.basename
+normpath = os.path.normpath
+expandvars = os.path.expandvars
+expanduser = os.path.expanduser
+abspath = os.path.abspath
+join = os.path.join
+read = os.read
+write = os.write
+close = os.close
+fsync = os.fsync
+
+# constants
+NULL_HEX_SHA = "0"*40
+NULL_BIN_SHA = "\0"*20
+
+#} END Aliases
+
+#{ compatibility stuff ...
+
+class _RandomAccessStringIO(object):
+ """Wrapper to provide required functionality in case memory maps cannot or may
+ not be used. This is only really required in python 2.4"""
+ __slots__ = '_sio'
+
+ def __init__(self, buf=''):
+ self._sio = StringIO(buf)
+
+ def __getattr__(self, attr):
+ return getattr(self._sio, attr)
+
+ def __len__(self):
+ return len(self.getvalue())
+
+ def __getitem__(self, i):
+ return self.getvalue()[i]
+
+ def __getslice__(self, start, end):
+ return self.getvalue()[start:end]
+
+#} END compatibility stuff ...
+
+#{ Routines
+
+def get_user_id():
+ """:return: string identifying the currently active system user as name@node
+ :note: user can be set with the 'USER' environment variable, usually set on windows"""
+ ukn = 'UNKNOWN'
+ username = os.environ.get('USER', os.environ.get('USERNAME', ukn))
+ if username == ukn and hasattr(os, 'getlogin'):
+ username = os.getlogin()
+ # END get username from login
+ return "%s@%s" % (username, platform.node())
+
+def is_git_dir(d):
+ """ This is taken from the git setup.c:is_git_directory
+ function."""
+ if isdir(d) and \
+ isdir(join(d, 'objects')) and \
+ isdir(join(d, 'refs')):
+ headref = join(d, 'HEAD')
+ return isfile(headref) or \
+ (os.path.islink(headref) and
+ os.readlink(headref).startswith('refs'))
+ return False
-#{ Utility Methods
def stream_copy(source, destination, chunk_size=512*1024):
"""Copy all data from the source stream into the destination stream in chunks
@@ -41,6 +162,87 @@ def stream_copy(source, destination, chunk_size=512*1024):
break
# END reading output stream
return br
+
+def make_sha(source=''):
+ """A python2.4 workaround for the sha/hashlib module fiasco
+ :note: From the dulwich project """
+ try:
+ return hashlib.sha1(source)
+ except NameError:
+ sha1 = sha.sha(source)
+ return sha1
+
+def allocate_memory(size):
+ """:return: a file-protocol accessible memory block of the given size"""
+ if size == 0:
+ return _RandomAccessStringIO('')
+ # END handle empty chunks gracefully
+
+ try:
+ return mmap.mmap(-1, size) # read-write by default
+ except EnvironmentError:
+ # setup real memory instead
+ # this of course may fail if the amount of memory is not available in
+ # one chunk - would only be the case in python 2.4, being more likely on
+ # 32 bit systems.
+ return _RandomAccessStringIO("\0"*size)
+ # END handle memory allocation
+
+
+def file_contents_ro(fd, stream=False, allow_mmap=True):
+ """:return: read-only contents of the file represented by the file descriptor fd
+ :param fd: file descriptor opened for reading
+ :param stream: if False, random access is provided, otherwise the stream interface
+ is provided.
+ :param allow_mmap: if True, its allowed to map the contents into memory, which
+ allows large files to be handled and accessed efficiently. The file-descriptor
+ will change its position if this is False"""
+ try:
+ if allow_mmap:
+ # supports stream and random access
+ try:
+ return mmap.mmap(fd, 0, access=mmap.ACCESS_READ)
+ except EnvironmentError:
+ # python 2.4 issue, 0 wants to be the actual size
+ return mmap.mmap(fd, os.fstat(fd).st_size, access=mmap.ACCESS_READ)
+ # END handle python 2.4
+ except OSError:
+ pass
+ # END exception handling
+
+ # read manully
+ contents = os.read(fd, os.fstat(fd).st_size)
+ if stream:
+ return _RandomAccessStringIO(contents)
+ return contents
+
+def file_contents_ro_filepath(filepath, stream=False, allow_mmap=True, flags=0):
+ """Get the file contents at filepath as fast as possible
+ :return: random access compatible memory of the given filepath
+ :param stream: see ``file_contents_ro``
+ :param allow_mmap: see ``file_contents_ro``
+ :param flags: additional flags to pass to os.open
+ :raise OSError: If the file could not be opened
+ :note: for now we don't try to use O_NOATIME directly as the right value needs to be
+ shared per database in fact. It only makes a real difference for loose object
+ databases anyway, and they use it with the help of the ``flags`` parameter"""
+ fd = os.open(filepath, os.O_RDONLY|getattr(os, 'O_BINARY', 0)|flags)
+ try:
+ return file_contents_ro(fd, stream, allow_mmap)
+ finally:
+ close(fd)
+ # END assure file is closed
+
+def to_hex_sha(sha):
+ """:return: hexified version of sha"""
+ if len(sha) == 40:
+ return sha
+ return bin_to_hex(sha)
+
+def to_bin_sha(sha):
+ if len(sha) == 20:
+ return sha
+ return hex_to_bin(sha)
def join_path(a, *p):
"""Join path tokens together similar to os.path.join, but always use
@@ -61,6 +263,7 @@ def to_native_path_windows(path):
def to_native_path_linux(path):
return path.replace('\\','/')
+
if sys.platform.startswith('win'):
to_native_path = to_native_path_windows
else:
@@ -75,7 +278,7 @@ def join_path_native(a, *p):
needed to play it safe on my dear windows and to assure nice paths that only
use '\'"""
return to_native_path(join_path(a, *p))
-
+
def assure_directory_exists(path, is_file=False):
"""Assure that the directory pointed to by path exists.
@@ -89,138 +292,287 @@ def assure_directory_exists(path, is_file=False):
os.makedirs(path)
return True
return False
-
-def get_user_id():
- """:return: string identifying the currently active system user as name@node
- :note: user can be set with the 'USER' environment variable, usually set on windows"""
- ukn = 'UNKNOWN'
- username = os.environ.get('USER', os.environ.get('USERNAME', ukn))
- if username == ukn and hasattr(os, 'getlogin'):
- username = os.getlogin()
- # END get username from login
- return "%s@%s" % (username, platform.node())
-#} END utilities
-#{ Classes
+#} END routines
+
-class RemoteProgress(object):
+#{ Utilities
+
+class LazyMixin(object):
"""
- Handler providing an interface to parse progress information emitted by git-push
- and git-fetch and to dispatch callbacks allowing subclasses to react to the progress.
+ Base class providing an interface to lazily retrieve attribute values upon
+ first access. If slots are used, memory will only be reserved once the attribute
+ is actually accessed and retrieved the first time. All future accesses will
+ return the cached value as stored in the Instance's dict or slot.
"""
- _num_op_codes = 5
- BEGIN, END, COUNTING, COMPRESSING, WRITING = [1 << x for x in range(_num_op_codes)]
- STAGE_MASK = BEGIN|END
- OP_MASK = ~STAGE_MASK
-
- __slots__ = ("_cur_line", "_seen_ops")
- re_op_absolute = re.compile("(remote: )?([\w\s]+):\s+()(\d+)()(.*)")
- re_op_relative = re.compile("(remote: )?([\w\s]+):\s+(\d+)% \((\d+)/(\d+)\)(.*)")
-
- def __init__(self):
- self._seen_ops = list()
-
- def _parse_progress_line(self, line):
- """Parse progress information from the given line as retrieved by git-push
- or git-fetch
-
- :return: list(line, ...) list of lines that could not be processed"""
- # handle
- # Counting objects: 4, done.
- # Compressing objects: 50% (1/2) \rCompressing objects: 100% (2/2) \rCompressing objects: 100% (2/2), done.
- self._cur_line = line
- sub_lines = line.split('\r')
- failed_lines = list()
- for sline in sub_lines:
- # find esacpe characters and cut them away - regex will not work with
- # them as they are non-ascii. As git might expect a tty, it will send them
- last_valid_index = None
- for i,c in enumerate(reversed(sline)):
- if ord(c) < 32:
- # its a slice index
- last_valid_index = -i-1
- # END character was non-ascii
- # END for each character in sline
- if last_valid_index is not None:
- sline = sline[:last_valid_index]
- # END cut away invalid part
- sline = sline.rstrip()
-
- cur_count, max_count = None, None
- match = self.re_op_relative.match(sline)
- if match is None:
- match = self.re_op_absolute.match(sline)
-
- if not match:
- self.line_dropped(sline)
- failed_lines.append(sline)
- continue
- # END could not get match
-
- op_code = 0
- remote, op_name, percent, cur_count, max_count, message = match.groups()
-
- # get operation id
- if op_name == "Counting objects":
- op_code |= self.COUNTING
- elif op_name == "Compressing objects":
- op_code |= self.COMPRESSING
- elif op_name == "Writing objects":
- op_code |= self.WRITING
- else:
- raise ValueError("Operation name %r unknown" % op_name)
-
- # figure out stage
- if op_code not in self._seen_ops:
- self._seen_ops.append(op_code)
- op_code |= self.BEGIN
- # END begin opcode
-
- if message is None:
- message = ''
- # END message handling
-
- message = message.strip()
- done_token = ', done.'
- if message.endswith(done_token):
- op_code |= self.END
- message = message[:-len(done_token)]
- # END end message handling
-
- self.update(op_code, cur_count, max_count, message)
- # END for each sub line
- return failed_lines
- def line_dropped(self, line):
- """Called whenever a line could not be understood and was therefore dropped."""
+ __slots__ = tuple()
+
+ def __getattr__(self, attr):
+ """
+ Whenever an attribute is requested that we do not know, we allow it
+ to be created and set. Next time the same attribute is reqeusted, it is simply
+ returned from our dict/slots. """
+ self._set_cache_(attr)
+ # will raise in case the cache was not created
+ return object.__getattribute__(self, attr)
+
+ def _set_cache_(self, attr):
+ """
+ This method should be overridden in the derived class.
+ It should check whether the attribute named by attr can be created
+ and cached. Do nothing if you do not know the attribute or call your subclass
+
+ The derived class may create as many additional attributes as it deems
+ necessary in case a git command returns more information than represented
+ in the single attribute."""
pass
+
+
+class LockedFD(object):
+ """
+ This class facilitates a safe read and write operation to a file on disk.
+ If we write to 'file', we obtain a lock file at 'file.lock' and write to
+ that instead. If we succeed, the lock file will be renamed to overwrite
+ the original file.
+
+ When reading, we obtain a lock file, but to prevent other writers from
+ succeeding while we are reading the file.
+
+ This type handles error correctly in that it will assure a consistent state
+ on destruction.
- def update(self, op_code, cur_count, max_count=None, message=''):
- """Called whenever the progress changes
+ :note: with this setup, parallel reading is not possible"""
+ __slots__ = ("_filepath", '_fd', '_write')
+
+ def __init__(self, filepath):
+ """Initialize an instance with the givne filepath"""
+ self._filepath = filepath
+ self._fd = None
+ self._write = None # if True, we write a file
+
+ def __del__(self):
+ # will do nothing if the file descriptor is already closed
+ if self._fd is not None:
+ self.rollback()
+
+ def _lockfilepath(self):
+ return "%s.lock" % self._filepath
+
+ def open(self, write=False, stream=False):
+ """
+ Open the file descriptor for reading or writing, both in binary mode.
+
+ :param write: if True, the file descriptor will be opened for writing. Other
+ wise it will be opened read-only.
+ :param stream: if True, the file descriptor will be wrapped into a simple stream
+ object which supports only reading or writing
+ :return: fd to read from or write to. It is still maintained by this instance
+ and must not be closed directly
+ :raise IOError: if the lock could not be retrieved
+ :raise OSError: If the actual file could not be opened for reading
+ :note: must only be called once"""
+ if self._write is not None:
+ raise AssertionError("Called %s multiple times" % self.open)
+
+ self._write = write
+
+ # try to open the lock file
+ binary = getattr(os, 'O_BINARY', 0)
+ lockmode = os.O_WRONLY | os.O_CREAT | os.O_EXCL | binary
+ try:
+ fd = os.open(self._lockfilepath(), lockmode, 0600)
+ if not write:
+ os.close(fd)
+ else:
+ self._fd = fd
+ # END handle file descriptor
+ except OSError:
+ raise IOError("Lock at %r could not be obtained" % self._lockfilepath())
+ # END handle lock retrieval
+
+ # open actual file if required
+ if self._fd is None:
+ # we could specify exlusive here, as we obtained the lock anyway
+ try:
+ self._fd = os.open(self._filepath, os.O_RDONLY | binary)
+ except:
+ # assure we release our lockfile
+ os.remove(self._lockfilepath())
+ raise
+ # END handle lockfile
+ # END open descriptor for reading
+
+ if stream:
+ # need delayed import
+ from stream import FDStream
+ return FDStream(self._fd)
+ else:
+ return self._fd
+ # END handle stream
+
+ def commit(self):
+ """When done writing, call this function to commit your changes into the
+ actual file.
+ The file descriptor will be closed, and the lockfile handled.
+ :note: can be called multiple times"""
+ self._end_writing(successful=True)
+
+ def rollback(self):
+ """Abort your operation without any changes. The file descriptor will be
+ closed, and the lock released.
+ :note: can be called multiple times"""
+ self._end_writing(successful=False)
+
+ def _end_writing(self, successful=True):
+ """Handle the lock according to the write mode """
+ if self._write is None:
+ raise AssertionError("Cannot end operation if it wasn't started yet")
+
+ if self._fd is None:
+ return
+
+ os.close(self._fd)
+ self._fd = None
- :param op_code:
- Integer allowing to be compared against Operation IDs and stage IDs.
+ lockfile = self._lockfilepath()
+ if self._write and successful:
+ # on windows, rename does not silently overwrite the existing one
+ if sys.platform == "win32":
+ if isfile(self._filepath):
+ os.remove(self._filepath)
+ # END remove if exists
+ # END win32 special handling
+ os.rename(lockfile, self._filepath)
- Stage IDs are BEGIN and END. BEGIN will only be set once for each Operation
- ID as well as END. It may be that BEGIN and END are set at once in case only
- one progress message was emitted due to the speed of the operation.
- Between BEGIN and END, none of these flags will be set
+ # assure others can at least read the file - the tmpfile left it at rw--
+ # We may also write that file, on windows that boils down to a remove-
+ # protection as well
+ chmod(self._filepath, 0644)
+ else:
+ # just delete the file so far, we failed
+ os.remove(lockfile)
+ # END successful handling
+
+
+class LockFile(object):
+ """Provides methods to obtain, check for, and release a file based lock which
+ should be used to handle concurrent access to the same file.
+
+ As we are a utility class to be derived from, we only use protected methods.
+
+ Locks will automatically be released on destruction"""
+ __slots__ = ("_file_path", "_owns_lock")
+
+ def __init__(self, file_path):
+ self._file_path = file_path
+ self._owns_lock = False
+
+ def __del__(self):
+ self._release_lock()
+
+ def _lock_file_path(self):
+ """:return: Path to lockfile"""
+ return "%s.lock" % (self._file_path)
+
+ def _has_lock(self):
+ """:return: True if we have a lock and if the lockfile still exists
+ :raise AssertionError: if our lock-file does not exist"""
+ if not self._owns_lock:
+ return False
+
+ return True
+
+ def _obtain_lock_or_raise(self):
+ """Create a lock file as flag for other instances, mark our instance as lock-holder
+
+ :raise IOError: if a lock was already present or a lock file could not be written"""
+ if self._has_lock():
+ return
+ lock_file = self._lock_file_path()
+ if os.path.isfile(lock_file):
+ raise IOError("Lock for file %r did already exist, delete %r in case the lock is illegal" % (self._file_path, lock_file))
- Operation IDs are all held within the OP_MASK. Only one Operation ID will
- be active per call.
- :param cur_count: Current absolute count of items
+ try:
+ fd = os.open(lock_file, os.O_WRONLY | os.O_CREAT | os.O_EXCL, 0)
+ os.close(fd)
+ except OSError,e:
+ raise IOError(str(e))
+
+ self._owns_lock = True
+
+ def _obtain_lock(self):
+ """The default implementation will raise if a lock cannot be obtained.
+ Subclasses may override this method to provide a different implementation"""
+ return self._obtain_lock_or_raise()
+
+ def _release_lock(self):
+ """Release our lock if we have one"""
+ if not self._has_lock():
+ return
- :param max_count:
- The maximum count of items we expect. It may be None in case there is
- no maximum number of items or if it is (yet) unknown.
+ # if someone removed our file beforhand, lets just flag this issue
+ # instead of failing, to make it more usable.
+ lfp = self._lock_file_path()
+ try:
+ # on bloody windows, the file needs write permissions to be removable.
+ # Why ...
+ if os.name == 'nt':
+ os.chmod(lfp, 0777)
+ # END handle win32
+ os.remove(lfp)
+ except OSError:
+ pass
+ self._owns_lock = False
+
+
+class BlockingLockFile(LockFile):
+ """The lock file will block until a lock could be obtained, or fail after
+ a specified timeout.
+
+ :note: If the directory containing the lock was removed, an exception will
+ be raised during the blocking period, preventing hangs as the lock
+ can never be obtained."""
+ __slots__ = ("_check_interval", "_max_block_time")
+ def __init__(self, file_path, check_interval_s=0.3, max_block_time_s=sys.maxint):
+ """Configure the instance
- :param message:
- In case of the 'WRITING' operation, it contains the amount of bytes
- transferred. It may possibly be used for other purposes as well.
+ :parm check_interval_s:
+ Period of time to sleep until the lock is checked the next time.
+ By default, it waits a nearly unlimited time
- You may read the contents of the current line in self._cur_line"""
- pass
+ :parm max_block_time_s: Maximum amount of seconds we may lock"""
+ super(BlockingLockFile, self).__init__(file_path)
+ self._check_interval = check_interval_s
+ self._max_block_time = max_block_time_s
+
+ def _obtain_lock(self):
+ """This method blocks until it obtained the lock, or raises IOError if
+ it ran out of time or if the parent directory was not available anymore.
+ If this method returns, you are guranteed to own the lock"""
+ starttime = time.time()
+ maxtime = starttime + float(self._max_block_time)
+ while True:
+ try:
+ super(BlockingLockFile, self)._obtain_lock()
+ except IOError:
+ # synity check: if the directory leading to the lockfile is not
+ # readable anymore, raise an execption
+ curtime = time.time()
+ if not os.path.isdir(os.path.dirname(self._lock_file_path())):
+ msg = "Directory containing the lockfile %r was not readable anymore after waiting %g seconds" % (self._lock_file_path(), curtime - starttime)
+ raise IOError(msg)
+ # END handle missing directory
+
+ if curtime >= maxtime:
+ msg = "Waited %g seconds for lock at %r" % ( maxtime - starttime, self._lock_file_path())
+ raise IOError(msg)
+ # END abort if we wait too long
+ time.sleep(self._check_interval)
+ else:
+ break
+ # END endless loop
class Actor(object):
@@ -274,20 +626,20 @@ class Actor(object):
m = cls.name_email_regex.search(string)
if m:
name, email = m.groups()
- return Actor(name, email)
+ return cls(name, email)
else:
m = cls.name_only_regex.search(string)
if m:
- return Actor(m.group(1), None)
+ return cls(m.group(1), None)
else:
# assume best and use the whole string as name
- return Actor(string, None)
+ return cls(string, None)
# END special case name
# END handle name/email matching
@classmethod
def _main_actor(cls, env_name, env_email, config_reader=None):
- actor = Actor('', '')
+ actor = cls('', '')
default_email = get_user_id()
default_name = default_email.split('@')[0]
@@ -324,6 +676,95 @@ class Actor(object):
return cls._main_actor(cls.env_author_name, cls.env_author_email, config_reader)
+class Iterable(object):
+ """Defines an interface for iterable items which is to assure a uniform
+ way to retrieve and iterate items within the git repository"""
+ __slots__ = tuple()
+ _id_attribute_ = "attribute that most suitably identifies your instance"
+
+ @classmethod
+ def list_items(cls, repo, *args, **kwargs):
+ """
+ Find all items of this type - subclasses can specify args and kwargs differently.
+ If no args are given, subclasses are obliged to return all items if no additional
+ arguments arg given.
+
+ :note: Favor the iter_items method as it will
+
+ :return:list(Item,...) list of item instances"""
+ out_list = IterableList( cls._id_attribute_ )
+ out_list.extend(cls.iter_items(repo, *args, **kwargs))
+ return out_list
+
+
+ @classmethod
+ def iter_items(cls, repo, *args, **kwargs):
+ """For more information about the arguments, see list_items
+ :return: iterator yielding Items"""
+ raise NotImplementedError("To be implemented by Subclass")
+
+
+class IterableList(list):
+ """
+ List of iterable objects allowing to query an object by id or by named index::
+
+ heads = repo.heads
+ heads.master
+ heads['master']
+ heads[0]
+
+ It requires an id_attribute name to be set which will be queried from its
+ contained items to have a means for comparison.
+
+ A prefix can be specified which is to be used in case the id returned by the
+ items always contains a prefix that does not matter to the user, so it
+ can be left out."""
+ __slots__ = ('_id_attr', '_prefix')
+
+ def __new__(cls, id_attr, prefix=''):
+ return super(IterableList,cls).__new__(cls)
+
+ def __init__(self, id_attr, prefix=''):
+ self._id_attr = id_attr
+ self._prefix = prefix
+ if not isinstance(id_attr, basestring):
+ raise ValueError("First parameter must be a string identifying the name-property. Extend the list after initialization")
+ # END help debugging !
+
+ def __getattr__(self, attr):
+ attr = self._prefix + attr
+ for item in self:
+ if getattr(item, self._id_attr) == attr:
+ return item
+ # END for each item
+ return list.__getattribute__(self, attr)
+
+ def __getitem__(self, index):
+ if isinstance(index, int):
+ return list.__getitem__(self,index)
+
+ try:
+ return getattr(self, index)
+ except AttributeError:
+ raise IndexError( "No item found with id %r" % (self._prefix + index) )
+
+
+
+#} END utilities
+
+#{ Classes
+
+class RepoAliasMixin(object):
+ """Simple utility providing a repo-property which resolves to the 'odb' attribute
+ of the actual type. This is for api compatability only, as the types previously
+ held repository instances, now they hold odb instances instead"""
+ __slots__ = tuple()
+
+ @property
+ def repo(self):
+ return self.odb
+
+
class Stats(object):
"""
Represents stat information as presented by git at the end of a merge. It is
@@ -407,195 +848,4 @@ class IndexFileSHA1Writer(object):
return self.f.tell()
-class LockFile(object):
- """Provides methods to obtain, check for, and release a file based lock which
- should be used to handle concurrent access to the same file.
-
- As we are a utility class to be derived from, we only use protected methods.
-
- Locks will automatically be released on destruction"""
- __slots__ = ("_file_path", "_owns_lock")
-
- def __init__(self, file_path):
- self._file_path = file_path
- self._owns_lock = False
-
- def __del__(self):
- self._release_lock()
-
- def _lock_file_path(self):
- """:return: Path to lockfile"""
- return "%s.lock" % (self._file_path)
-
- def _has_lock(self):
- """:return: True if we have a lock and if the lockfile still exists
- :raise AssertionError: if our lock-file does not exist"""
- if not self._owns_lock:
- return False
-
- return True
-
- def _obtain_lock_or_raise(self):
- """Create a lock file as flag for other instances, mark our instance as lock-holder
-
- :raise IOError: if a lock was already present or a lock file could not be written"""
- if self._has_lock():
- return
- lock_file = self._lock_file_path()
- if os.path.isfile(lock_file):
- raise IOError("Lock for file %r did already exist, delete %r in case the lock is illegal" % (self._file_path, lock_file))
-
- try:
- fd = os.open(lock_file, os.O_WRONLY | os.O_CREAT | os.O_EXCL, 0)
- os.close(fd)
- except OSError,e:
- raise IOError(str(e))
-
- self._owns_lock = True
-
- def _obtain_lock(self):
- """The default implementation will raise if a lock cannot be obtained.
- Subclasses may override this method to provide a different implementation"""
- return self._obtain_lock_or_raise()
-
- def _release_lock(self):
- """Release our lock if we have one"""
- if not self._has_lock():
- return
-
- # if someone removed our file beforhand, lets just flag this issue
- # instead of failing, to make it more usable.
- lfp = self._lock_file_path()
- try:
- # on bloody windows, the file needs write permissions to be removable.
- # Why ...
- if os.name == 'nt':
- os.chmod(lfp, 0777)
- # END handle win32
- os.remove(lfp)
- except OSError:
- pass
- self._owns_lock = False
-
-
-class BlockingLockFile(LockFile):
- """The lock file will block until a lock could be obtained, or fail after
- a specified timeout.
-
- :note: If the directory containing the lock was removed, an exception will
- be raised during the blocking period, preventing hangs as the lock
- can never be obtained."""
- __slots__ = ("_check_interval", "_max_block_time")
- def __init__(self, file_path, check_interval_s=0.3, max_block_time_s=sys.maxint):
- """Configure the instance
-
- :parm check_interval_s:
- Period of time to sleep until the lock is checked the next time.
- By default, it waits a nearly unlimited time
-
- :parm max_block_time_s: Maximum amount of seconds we may lock"""
- super(BlockingLockFile, self).__init__(file_path)
- self._check_interval = check_interval_s
- self._max_block_time = max_block_time_s
-
- def _obtain_lock(self):
- """This method blocks until it obtained the lock, or raises IOError if
- it ran out of time or if the parent directory was not available anymore.
- If this method returns, you are guranteed to own the lock"""
- starttime = time.time()
- maxtime = starttime + float(self._max_block_time)
- while True:
- try:
- super(BlockingLockFile, self)._obtain_lock()
- except IOError:
- # synity check: if the directory leading to the lockfile is not
- # readable anymore, raise an execption
- curtime = time.time()
- if not os.path.isdir(os.path.dirname(self._lock_file_path())):
- msg = "Directory containing the lockfile %r was not readable anymore after waiting %g seconds" % (self._lock_file_path(), curtime - starttime)
- raise IOError(msg)
- # END handle missing directory
-
- if curtime >= maxtime:
- msg = "Waited %g seconds for lock at %r" % ( maxtime - starttime, self._lock_file_path())
- raise IOError(msg)
- # END abort if we wait too long
- time.sleep(self._check_interval)
- else:
- break
- # END endless loop
-
-
-class IterableList(list):
- """
- List of iterable objects allowing to query an object by id or by named index::
-
- heads = repo.heads
- heads.master
- heads['master']
- heads[0]
-
- It requires an id_attribute name to be set which will be queried from its
- contained items to have a means for comparison.
-
- A prefix can be specified which is to be used in case the id returned by the
- items always contains a prefix that does not matter to the user, so it
- can be left out."""
- __slots__ = ('_id_attr', '_prefix')
-
- def __new__(cls, id_attr, prefix=''):
- return super(IterableList,cls).__new__(cls)
-
- def __init__(self, id_attr, prefix=''):
- self._id_attr = id_attr
- self._prefix = prefix
- if not isinstance(id_attr, basestring):
- raise ValueError("First parameter must be a string identifying the name-property. Extend the list after initialization")
- # END help debugging !
-
- def __getattr__(self, attr):
- attr = self._prefix + attr
- for item in self:
- if getattr(item, self._id_attr) == attr:
- return item
- # END for each item
- return list.__getattribute__(self, attr)
-
- def __getitem__(self, index):
- if isinstance(index, int):
- return list.__getitem__(self,index)
-
- try:
- return getattr(self, index)
- except AttributeError:
- raise IndexError( "No item found with id %r" % (self._prefix + index) )
-
-
-class Iterable(object):
- """Defines an interface for iterable items which is to assure a uniform
- way to retrieve and iterate items within the git repository"""
- __slots__ = tuple()
- _id_attribute_ = "attribute that most suitably identifies your instance"
-
- @classmethod
- def list_items(cls, repo, *args, **kwargs):
- """
- Find all items of this type - subclasses can specify args and kwargs differently.
- If no args are given, subclasses are obliged to return all items if no additional
- arguments arg given.
-
- :note: Favor the iter_items method as it will
-
- :return:list(Item,...) list of item instances"""
- out_list = IterableList( cls._id_attribute_ )
- out_list.extend(cls.iter_items(repo, *args, **kwargs))
- return out_list
-
-
- @classmethod
- def iter_items(cls, repo, *args, **kwargs):
- """For more information about the arguments, see list_items
- :return: iterator yielding Items"""
- raise NotImplementedError("To be implemented by Subclass")
-
#} END classes