~bzr-pqm/bzr/bzr.dev

# Copyright (C) 2005 Canonical Ltd

# 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; either version 2 of the License, or
# (at your option) any later version.

# 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., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA


import sys
import os
import errno
from warnings import warn
from cStringIO import StringIO


import bzrlib
from bzrlib.inventory import InventoryEntry
import bzrlib.inventory as inventory
from bzrlib.trace import mutter, note
from bzrlib.osutils import (isdir, quotefn, compact_date, rand_bytes, 
                            rename, splitpath, sha_file, appendpath, 
                            file_kind, abspath)
import bzrlib.errors as errors
from bzrlib.errors import (BzrError, InvalidRevisionNumber, InvalidRevisionId,
                           NoSuchRevision, HistoryMissing, NotBranchError,
                           DivergedBranches, LockError, UnlistableStore,
                           UnlistableBranch, NoSuchFile, NotVersionedError)
from bzrlib.textui import show_status
from bzrlib.revision import Revision, is_ancestor, get_intervening_revisions

from bzrlib.delta import compare_trees
from bzrlib.tree import EmptyTree, RevisionTree
from bzrlib.inventory import Inventory
from bzrlib.store import copy_all
from bzrlib.store.compressed_text import CompressedTextStore
from bzrlib.store.text import TextStore
from bzrlib.store.weave import WeaveStore
from bzrlib.testament import Testament
import bzrlib.transactions as transactions
from bzrlib.transport import Transport, get_transport
import bzrlib.xml5
import bzrlib.ui


BZR_BRANCH_FORMAT_4 = "Bazaar-NG branch, format 0.0.4\n"
BZR_BRANCH_FORMAT_5 = "Bazaar-NG branch, format 5\n"
BZR_BRANCH_FORMAT_6 = "Bazaar-NG branch, format 6\n"
## TODO: Maybe include checks for common corruption of newlines, etc?


# TODO: Some operations like log might retrieve the same revisions
# repeatedly to calculate deltas.  We could perhaps have a weakref
# cache in memory to make this faster.  In general anything can be
# cached in memory between lock and unlock operations.

def find_branch(*ignored, **ignored_too):
    # XXX: leave this here for about one release, then remove it
    raise NotImplementedError('find_branch() is not supported anymore, '
                              'please use one of the new branch constructors')


def needs_read_lock(unbound):
    """Decorate unbound to take out and release a read lock."""
    def decorated(self, *args, **kwargs):
        self.lock_read()
        try:
            return unbound(self, *args, **kwargs)
        finally:
            self.unlock()
    return decorated


def needs_write_lock(unbound):
    """Decorate unbound to take out and release a write lock."""
    def decorated(self, *args, **kwargs):
        self.lock_write()
        try:
            return unbound(self, *args, **kwargs)
        finally:
            self.unlock()
    return decorated

######################################################################
# branch objects

class Branch(object):
    """Branch holding a history of revisions.

    base
        Base directory/url of the branch.
    """
    base = None

    def __init__(self, *ignored, **ignored_too):
        raise NotImplementedError('The Branch class is abstract')

    @staticmethod
    def open_downlevel(base):
        """Open a branch which may be of an old format.
        
        Only local branches are supported."""
        return _Branch(get_transport(base), relax_version_check=True)
        
    @staticmethod
    def open(base):
        """Open an existing branch, rooted at 'base' (url)"""
        t = get_transport(base)
        mutter("trying to open %r with transport %r", base, t)
        return _Branch(t)

    @staticmethod
    def open_containing(url):
        """Open an existing branch which contains url.
        
        This probes for a branch at url, and searches upwards from there.

        Basically we keep looking up until we find the control directory or
        run into the root.  If there isn't one, raises NotBranchError.
        If there is one, it is returned, along with the unused portion of url.
        """
        t = get_transport(url)
        while True:
            try:
                return _Branch(t), t.relpath(url)
            except NotBranchError:
                pass
            new_t = t.clone('..')
            if new_t.base == t.base:
                # reached the root, whatever that may be
                raise NotBranchError(path=url)
            t = new_t

    @staticmethod
    def initialize(base):
        """Create a new branch, rooted at 'base' (url)"""
        t = get_transport(base)
        return _Branch(t, init=True)

    def setup_caching(self, cache_root):
        """Subclasses that care about caching should override this, and set
        up cached stores located under cache_root.
        """
        self.cache_root = cache_root


class _Branch(Branch):
    """A branch stored in the actual filesystem.

    Note that it's "local" in the context of the filesystem; it doesn't
    really matter if it's on an nfs/smb/afs/coda/... share, as long as
    it's writable, and can be accessed via the normal filesystem API.

    _lock_mode
        None, or 'r' or 'w'

    _lock_count
        If _lock_mode is true, a positive count of the number of times the
        lock has been taken.

    _lock
        Lock object from bzrlib.lock.
    """
    # We actually expect this class to be somewhat short-lived; part of its
    # purpose is to try to isolate what bits of the branch logic are tied to
    # filesystem access, so that in a later step, we can extricate them to
    # a separarte ("storage") class.
    _lock_mode = None
    _lock_count = None
    _lock = None
    _inventory_weave = None
    
    # Map some sort of prefix into a namespace
    # stuff like "revno:10", "revid:", etc.
    # This should match a prefix with a function which accepts
    REVISION_NAMESPACES = {}

    def push_stores(self, branch_to):
        """Copy the content of this branches store to branch_to."""
        if (self._branch_format != branch_to._branch_format
            or self._branch_format != 4):
            from bzrlib.fetch import greedy_fetch
            mutter("falling back to fetch logic to push between %s(%s) and %s(%s)",
                   self, self._branch_format, branch_to, branch_to._branch_format)
            greedy_fetch(to_branch=branch_to, from_branch=self,
                         revision=self.last_revision())
            return

        store_pairs = ((self.text_store,      branch_to.text_store),
                       (self.inventory_store, branch_to.inventory_store),
                       (self.revision_store,  branch_to.revision_store))
        try:
            for from_store, to_store in store_pairs: 
                copy_all(from_store, to_store)
        except UnlistableStore:
            raise UnlistableBranch(from_store)

    def __init__(self, transport, init=False,
                 relax_version_check=False):
        """Create new branch object at a particular location.

        transport -- A Transport object, defining how to access files.
                (If a string, transport.transport() will be used to
                create a Transport object)
        
        init -- If True, create new control files in a previously
             unversioned directory.  If False, the branch must already
             be versioned.

        relax_version_check -- If true, the usual check for the branch
            version is not applied.  This is intended only for
            upgrade/recovery type use; it's not guaranteed that
            all operations will work on old format branches.

        In the test suite, creation of new trees is tested using the
        `ScratchBranch` class.
        """
        assert isinstance(transport, Transport), \
            "%r is not a Transport" % transport
        self._transport = transport
        if init:
            self._make_control()
        self._check_format(relax_version_check)

        def get_store(name, compressed=True, prefixed=False):
            # FIXME: This approach of assuming stores are all entirely compressed
            # or entirely uncompressed is tidy, but breaks upgrade from 
            # some existing branches where there's a mixture; we probably 
            # still want the option to look for both.
            relpath = self._rel_controlfilename(name)
            if compressed:
                store = CompressedTextStore(self._transport.clone(relpath),
                                            prefixed=prefixed)
            else:
                store = TextStore(self._transport.clone(relpath),
                                  prefixed=prefixed)
            #if self._transport.should_cache():
            #    cache_path = os.path.join(self.cache_root, name)
            #    os.mkdir(cache_path)
            #    store = bzrlib.store.CachedStore(store, cache_path)
            return store
        def get_weave(name, prefixed=False):
            relpath = self._rel_controlfilename(name)
            ws = WeaveStore(self._transport.clone(relpath), prefixed=prefixed)
            if self._transport.should_cache():
                ws.enable_cache = True
            return ws

        if self._branch_format == 4:
            self.inventory_store = get_store('inventory-store')
            self.text_store = get_store('text-store')
            self.revision_store = get_store('revision-store')
        elif self._branch_format == 5:
            self.control_weaves = get_weave('')
            self.weave_store = get_weave('weaves')
            self.revision_store = get_store('revision-store', compressed=False)
        elif self._branch_format == 6:
            self.control_weaves = get_weave('')
            self.weave_store = get_weave('weaves', prefixed=True)
            self.revision_store = get_store('revision-store', compressed=False,
                                            prefixed=True)
        self.revision_store.register_suffix('sig')
        self._transaction = None

    def __str__(self):
        return '%s(%r)' % (self.__class__.__name__, self._transport.base)


    __repr__ = __str__


    def __del__(self):
        if self._lock_mode or self._lock:
            # XXX: This should show something every time, and be suitable for
            # headless operation and embedding
            warn("branch %r was not explicitly unlocked" % self)
            self._lock.unlock()

        # TODO: It might be best to do this somewhere else,
        # but it is nice for a Branch object to automatically
        # cache it's information.
        # Alternatively, we could have the Transport objects cache requests
        # See the earlier discussion about how major objects (like Branch)
        # should never expect their __del__ function to run.
        if hasattr(self, 'cache_root') and self.cache_root is not None:
            try:
                import shutil
                shutil.rmtree(self.cache_root)
            except:
                pass
            self.cache_root = None

    def _get_base(self):
        if self._transport:
            return self._transport.base
        return None

    base = property(_get_base, doc="The URL for the root of this branch.")

    def _finish_transaction(self):
        """Exit the current transaction."""
        if self._transaction is None:
            raise errors.LockError('Branch %s is not in a transaction' %
                                   self)
        transaction = self._transaction
        self._transaction = None
        transaction.finish()

    def get_transaction(self):
        """Return the current active transaction.

        If no transaction is active, this returns a passthrough object
        for which all data is immedaitely flushed and no caching happens.
        """
        if self._transaction is None:
            return transactions.PassThroughTransaction()
        else:
            return self._transaction

    def _set_transaction(self, new_transaction):
        """Set a new active transaction."""
        if self._transaction is not None:
            raise errors.LockError('Branch %s is in a transaction already.' %
                                   self)
        self._transaction = new_transaction

    def lock_write(self):
        mutter("lock write: %s (%s)", self, self._lock_count)
        # TODO: Upgrade locking to support using a Transport,
        # and potentially a remote locking protocol
        if self._lock_mode:
            if self._lock_mode != 'w':
                raise LockError("can't upgrade to a write lock from %r" %
                                self._lock_mode)
            self._lock_count += 1
        else:
            self._lock = self._transport.lock_write(
                    self._rel_controlfilename('branch-lock'))
            self._lock_mode = 'w'
            self._lock_count = 1
            self._set_transaction(transactions.PassThroughTransaction())

    def lock_read(self):
        mutter("lock read: %s (%s)", self, self._lock_count)
        if self._lock_mode:
            assert self._lock_mode in ('r', 'w'), \
                   "invalid lock mode %r" % self._lock_mode
            self._lock_count += 1
        else:
            self._lock = self._transport.lock_read(
                    self._rel_controlfilename('branch-lock'))
            self._lock_mode = 'r'
            self._lock_count = 1
            self._set_transaction(transactions.ReadOnlyTransaction())
            # 5K may be excessive, but hey, its a knob.
            self.get_transaction().set_cache_size(5000)
                        
    def unlock(self):
        mutter("unlock: %s (%s)", self, self._lock_count)
        if not self._lock_mode:
            raise LockError('branch %r is not locked' % (self))

        if self._lock_count > 1:
            self._lock_count -= 1
        else:
            self._finish_transaction()
            self._lock.unlock()
            self._lock = None
            self._lock_mode = self._lock_count = None

    def abspath(self, name):
        """Return absolute filename for something in the branch
        
        XXX: Robert Collins 20051017 what is this used for? why is it a branch
        method and not a tree method.
        """
        return self._transport.abspath(name)

    def _rel_controlfilename(self, file_or_path):
        if not isinstance(file_or_path, basestring):
            file_or_path = '/'.join(file_or_path)
        if file_or_path == '':
            return bzrlib.BZRDIR
        return bzrlib.transport.urlescape(bzrlib.BZRDIR + '/' + file_or_path)

    def controlfilename(self, file_or_path):
        """Return location relative to branch."""
        return self._transport.abspath(self._rel_controlfilename(file_or_path))


    def controlfile(self, file_or_path, mode='r'):
        """Open a control file for this branch.

        There are two classes of file in the control directory: text
        and binary.  binary files are untranslated byte streams.  Text
        control files are stored with Unix newlines and in UTF-8, even
        if the platform or locale defaults are different.

        Controlfiles should almost never be opened in write mode but
        rather should be atomically copied and replaced using atomicfile.
        """
        import codecs

        relpath = self._rel_controlfilename(file_or_path)
        #TODO: codecs.open() buffers linewise, so it was overloaded with
        # a much larger buffer, do we need to do the same for getreader/getwriter?
        if mode == 'rb': 
            return self._transport.get(relpath)
        elif mode == 'wb':
            raise BzrError("Branch.controlfile(mode='wb') is not supported, use put_controlfiles")
        elif mode == 'r':
            return codecs.getreader('utf-8')(self._transport.get(relpath), errors='replace')
        elif mode == 'w':
            raise BzrError("Branch.controlfile(mode='w') is not supported, use put_controlfiles")
        else:
            raise BzrError("invalid controlfile mode %r" % mode)

    def put_controlfile(self, path, f, encode=True):
        """Write an entry as a controlfile.

        :param path: The path to put the file, relative to the .bzr control
                     directory
        :param f: A file-like or string object whose contents should be copied.
        :param encode:  If true, encode the contents as utf-8
        """
        self.put_controlfiles([(path, f)], encode=encode)

    def put_controlfiles(self, files, encode=True):
        """Write several entries as controlfiles.

        :param files: A list of [(path, file)] pairs, where the path is the directory
                      underneath the bzr control directory
        :param encode:  If true, encode the contents as utf-8
        """
        import codecs
        ctrl_files = []
        for path, f in files:
            if encode:
                if isinstance(f, basestring):
                    f = f.encode('utf-8', 'replace')
                else:
                    f = codecs.getwriter('utf-8')(f, errors='replace')
            path = self._rel_controlfilename(path)
            ctrl_files.append((path, f))
        self._transport.put_multi(ctrl_files)

    def _make_control(self):
        from bzrlib.inventory import Inventory
        from bzrlib.weavefile import write_weave_v5
        from bzrlib.weave import Weave
        
        # Create an empty inventory
        sio = StringIO()
        # if we want per-tree root ids then this is the place to set
        # them; they're not needed for now and so ommitted for
        # simplicity.
        bzrlib.xml5.serializer_v5.write_inventory(Inventory(), sio)
        empty_inv = sio.getvalue()
        sio = StringIO()
        bzrlib.weavefile.write_weave_v5(Weave(), sio)
        empty_weave = sio.getvalue()

        dirs = [[], 'revision-store', 'weaves']
        files = [('README', 
            "This is a Bazaar-NG control directory.\n"
            "Do not change any files in this directory.\n"),
            ('branch-format', BZR_BRANCH_FORMAT_6),
            ('revision-history', ''),
            ('branch-name', ''),
            ('branch-lock', ''),
            ('pending-merges', ''),
            ('inventory', empty_inv),
            ('inventory.weave', empty_weave),
            ('ancestry.weave', empty_weave)
        ]
        cfn = self._rel_controlfilename
        self._transport.mkdir_multi([cfn(d) for d in dirs])
        self.put_controlfiles(files)
        mutter('created control directory in ' + self._transport.base)

    def _check_format(self, relax_version_check):
        """Check this branch format is supported.

        The format level is stored, as an integer, in
        self._branch_format for code that needs to check it later.

        In the future, we might need different in-memory Branch
        classes to support downlevel branches.  But not yet.
        """
        try:
            fmt = self.controlfile('branch-format', 'r').read()
        except NoSuchFile:
            raise NotBranchError(path=self.base)
        mutter("got branch format %r", fmt)
        if fmt == BZR_BRANCH_FORMAT_6:
            self._branch_format = 6
        elif fmt == BZR_BRANCH_FORMAT_5:
            self._branch_format = 5
        elif fmt == BZR_BRANCH_FORMAT_4:
            self._branch_format = 4

        if (not relax_version_check
            and self._branch_format not in (5, 6)):
            raise errors.UnsupportedFormatError(
                           'sorry, branch format %r not supported' % fmt,
                           ['use a different bzr version',
                            'or remove the .bzr directory'
                            ' and "bzr init" again'])

    def get_root_id(self):
        """Return the id of this branches root"""
        inv = self.read_working_inventory()
        return inv.root.file_id

    def set_root_id(self, file_id):
        inv = self.read_working_inventory()
        orig_root_id = inv.root.file_id
        del inv._byid[inv.root.file_id]
        inv.root.file_id = file_id
        inv._byid[inv.root.file_id] = inv.root
        for fid in inv:
            entry = inv[fid]
            if entry.parent_id in (None, orig_root_id):
                entry.parent_id = inv.root.file_id
        self._write_inventory(inv)

    @needs_read_lock
    def read_working_inventory(self):
        """Read the working inventory."""
        # ElementTree does its own conversion from UTF-8, so open in
        # binary.
        f = self.controlfile('inventory', 'rb')
        return bzrlib.xml5.serializer_v5.read_inventory(f)

    @needs_write_lock
    def _write_inventory(self, inv):
        """Update the working inventory.

        That is to say, the inventory describing changes underway, that
        will be committed to the next revision.
        """
        from cStringIO import StringIO
        sio = StringIO()
        bzrlib.xml5.serializer_v5.write_inventory(inv, sio)
        sio.seek(0)
        # Transport handles atomicity
        self.put_controlfile('inventory', sio)
        
        mutter('wrote working inventory')
            
    inventory = property(read_working_inventory, _write_inventory, None,
                         """Inventory for the working copy.""")

    @needs_write_lock
    def add(self, files, ids=None):
        """Make files versioned.

        Note that the command line normally calls smart_add instead,
        which can automatically recurse.

        This puts the files in the Added state, so that they will be
        recorded by the next commit.

        files
            List of paths to add, relative to the base of the tree.

        ids
            If set, use these instead of automatically generated ids.
            Must be the same length as the list of files, but may
            contain None for ids that are to be autogenerated.

        TODO: Perhaps have an option to add the ids even if the files do
              not (yet) exist.

        TODO: Perhaps yield the ids and paths as they're added.
        """
        # TODO: Re-adding a file that is removed in the working copy
        # should probably put it back with the previous ID.
        if isinstance(files, basestring):
            assert(ids is None or isinstance(ids, basestring))
            files = [files]
            if ids is not None:
                ids = [ids]

        if ids is None:
            ids = [None] * len(files)
        else:
            assert(len(ids) == len(files))

        inv = self.read_working_inventory()
        for f,file_id in zip(files, ids):
            if is_control_file(f):
                raise BzrError("cannot add control file %s" % quotefn(f))

            fp = splitpath(f)

            if len(fp) == 0:
                raise BzrError("cannot add top-level %r" % f)

            fullpath = os.path.normpath(self.abspath(f))

            try:
                kind = file_kind(fullpath)
            except OSError:
                # maybe something better?
                raise BzrError('cannot add: not a regular file, symlink or directory: %s' % quotefn(f))

            if not InventoryEntry.versionable_kind(kind):
                raise BzrError('cannot add: not a versionable file ('
                               'i.e. regular file, symlink or directory): %s' % quotefn(f))

            if file_id is None:
                file_id = gen_file_id(f)
            inv.add_path(f, kind=kind, file_id=file_id)

            mutter("add file %s file_id:{%s} kind=%r" % (f, file_id, kind))

        self._write_inventory(inv)

    @needs_read_lock
    def print_file(self, file, revno):
        """Print `file` to stdout."""
        tree = self.revision_tree(self.get_rev_id(revno))
        # use inventory as it was in that revision
        file_id = tree.inventory.path2id(file)
        if not file_id:
            raise BzrError("%r is not present in revision %s" % (file, revno))
        tree.print_file(file_id)

    # FIXME: this doesn't need to be a branch method
    def set_inventory(self, new_inventory_list):
        from bzrlib.inventory import Inventory, InventoryEntry
        inv = Inventory(self.get_root_id())
        for path, file_id, parent, kind in new_inventory_list:
            name = os.path.basename(path)
            if name == "":
                continue
            # fixme, there should be a factory function inv,add_?? 
            if kind == 'directory':
                inv.add(inventory.InventoryDirectory(file_id, name, parent))
            elif kind == 'file':
                inv.add(inventory.InventoryFile(file_id, name, parent))
            elif kind == 'symlink':
                inv.add(inventory.InventoryLink(file_id, name, parent))
            else:
                raise BzrError("unknown kind %r" % kind)
        self._write_inventory(inv)

    def unknowns(self):
        """Return all unknown files.

        These are files in the working directory that are not versioned or
        control files or ignored.
        
        >>> from bzrlib.workingtree import WorkingTree
        >>> b = ScratchBranch(files=['foo', 'foo~'])
        >>> map(str, b.unknowns())
        ['foo']
        >>> b.add('foo')
        >>> list(b.unknowns())
        []
        >>> WorkingTree(b.base, b).remove('foo')
        >>> list(b.unknowns())
        [u'foo']
        """
        return self.working_tree().unknowns()

    @needs_write_lock
    def append_revision(self, *revision_ids):
        for revision_id in revision_ids:
            mutter("add {%s} to revision-history" % revision_id)
        rev_history = self.revision_history()
        rev_history.extend(revision_ids)
        self.set_revision_history(rev_history)

    @needs_write_lock
    def set_revision_history(self, rev_history):
        self.put_controlfile('revision-history', '\n'.join(rev_history))

    def has_revision(self, revision_id):
        """True if this branch has a copy of the revision.

        This does not necessarily imply the revision is merge
        or on the mainline."""
        return (revision_id is None
                or self.revision_store.has_id(revision_id))

    @needs_read_lock
    def get_revision_xml_file(self, revision_id):
        """Return XML file object for revision object."""
        if not revision_id or not isinstance(revision_id, basestring):
            raise InvalidRevisionId(revision_id)
        try:
            return self.revision_store.get(revision_id)
        except (IndexError, KeyError):
            raise bzrlib.errors.NoSuchRevision(self, revision_id)

    #deprecated
    get_revision_xml = get_revision_xml_file

    def get_revision_xml(self, revision_id):
        return self.get_revision_xml_file(revision_id).read()


    def get_revision(self, revision_id):
        """Return the Revision object for a named revision"""
        xml_file = self.get_revision_xml_file(revision_id)

        try:
            r = bzrlib.xml5.serializer_v5.read_revision(xml_file)
        except SyntaxError, e:
            raise bzrlib.errors.BzrError('failed to unpack revision_xml',
                                         [revision_id,
                                          str(e)])
            
        assert r.revision_id == revision_id
        return r

    def get_revision_delta(self, revno):
        """Return the delta for one revision.

        The delta is relative to its mainline predecessor, or the
        empty tree for revision 1.
        """
        assert isinstance(revno, int)
        rh = self.revision_history()
        if not (1 <= revno <= len(rh)):
            raise InvalidRevisionNumber(revno)

        # revno is 1-based; list is 0-based

        new_tree = self.revision_tree(rh[revno-1])
        if revno == 1:
            old_tree = EmptyTree()
        else:
            old_tree = self.revision_tree(rh[revno-2])

        return compare_trees(old_tree, new_tree)

    def get_revision_sha1(self, revision_id):
        """Hash the stored value of a revision, and return it."""
        # In the future, revision entries will be signed. At that
        # point, it is probably best *not* to include the signature
        # in the revision hash. Because that lets you re-sign
        # the revision, (add signatures/remove signatures) and still
        # have all hash pointers stay consistent.
        # But for now, just hash the contents.
        return bzrlib.osutils.sha_file(self.get_revision_xml_file(revision_id))

    def get_ancestry(self, revision_id):
        """Return a list of revision-ids integrated by a revision.
        
        This currently returns a list, but the ordering is not guaranteed:
        treat it as a set.
        """
        if revision_id is None:
            return [None]
        w = self.get_inventory_weave()
        return [None] + map(w.idx_to_name,
                            w.inclusions([w.lookup(revision_id)]))

    def get_inventory_weave(self):
        return self.control_weaves.get_weave('inventory',
                                             self.get_transaction())

    def get_inventory(self, revision_id):
        """Get Inventory object by hash."""
        xml = self.get_inventory_xml(revision_id)
        return bzrlib.xml5.serializer_v5.read_inventory_from_string(xml)

    def get_inventory_xml(self, revision_id):
        """Get inventory XML as a file object."""
        try:
            assert isinstance(revision_id, basestring), type(revision_id)
            iw = self.get_inventory_weave()
            return iw.get_text(iw.lookup(revision_id))
        except IndexError:
            raise bzrlib.errors.HistoryMissing(self, 'inventory', revision_id)

    def get_inventory_sha1(self, revision_id):
        """Return the sha1 hash of the inventory entry
        """
        return self.get_revision(revision_id).inventory_sha1

    def get_revision_inventory(self, revision_id):
        """Return inventory of a past revision."""
        # TODO: Unify this with get_inventory()
        # bzr 0.0.6 and later imposes the constraint that the inventory_id
        # must be the same as its revision, so this is trivial.
        if revision_id == None:
            return Inventory(self.get_root_id())
        else:
            return self.get_inventory(revision_id)

    @needs_read_lock
    def revision_history(self):
        """Return sequence of revision hashes on to this branch."""
        transaction = self.get_transaction()
        history = transaction.map.find_revision_history()
        if history is not None:
            mutter("cache hit for revision-history in %s", self)
            return list(history)
        history = [l.rstrip('\r\n') for l in
                self.controlfile('revision-history', 'r').readlines()]
        transaction.map.add_revision_history(history)
        # this call is disabled because revision_history is 
        # not really an object yet, and the transaction is for objects.
        # transaction.register_clean(history, precious=True)
        return list(history)

    def revno(self):
        """Return current revision number for this branch.

        That is equivalent to the number of revisions committed to
        this branch.
        """
        return len(self.revision_history())

    def last_revision(self):
        """Return last patch hash, or None if no history.
        """
        ph = self.revision_history()
        if ph:
            return ph[-1]
        else:
            return None

    def missing_revisions(self, other, stop_revision=None, diverged_ok=False):
        """Return a list of new revisions that would perfectly fit.
        
        If self and other have not diverged, return a list of the revisions
        present in other, but missing from self.

        >>> from bzrlib.commit import commit
        >>> bzrlib.trace.silent = True
        >>> br1 = ScratchBranch()
        >>> br2 = ScratchBranch()
        >>> br1.missing_revisions(br2)
        []
        >>> commit(br2, "lala!", rev_id="REVISION-ID-1")
        >>> br1.missing_revisions(br2)
        [u'REVISION-ID-1']
        >>> br2.missing_revisions(br1)
        []
        >>> commit(br1, "lala!", rev_id="REVISION-ID-1")
        >>> br1.missing_revisions(br2)
        []
        >>> commit(br2, "lala!", rev_id="REVISION-ID-2A")
        >>> br1.missing_revisions(br2)
        [u'REVISION-ID-2A']
        >>> commit(br1, "lala!", rev_id="REVISION-ID-2B")
        >>> br1.missing_revisions(br2)
        Traceback (most recent call last):
        DivergedBranches: These branches have diverged.
        """
        self_history = self.revision_history()
        self_len = len(self_history)
        other_history = other.revision_history()
        other_len = len(other_history)
        common_index = min(self_len, other_len) -1
        if common_index >= 0 and \
            self_history[common_index] != other_history[common_index]:
            raise DivergedBranches(self, other)

        if stop_revision is None:
            stop_revision = other_len
        else:
            assert isinstance(stop_revision, int)
            if stop_revision > other_len:
                raise bzrlib.errors.NoSuchRevision(self, stop_revision)
        return other_history[self_len:stop_revision]

    def update_revisions(self, other, stop_revision=None):
        """Pull in new perfect-fit revisions."""
        # FIXME: If the branches have diverged, but the latest
        # revision in this branch is completely merged into the other,
        # then we should still be able to pull.
        from bzrlib.fetch import greedy_fetch
        if stop_revision is None:
            stop_revision = other.last_revision()
        ### Should this be checking is_ancestor instead of revision_history?
        if (stop_revision is not None and 
            stop_revision in self.revision_history()):
            return
        greedy_fetch(to_branch=self, from_branch=other,
                     revision=stop_revision)
        pullable_revs = self.pullable_revisions(other, stop_revision)
        if len(pullable_revs) > 0:
            self.append_revision(*pullable_revs)

    def pullable_revisions(self, other, stop_revision):
        other_revno = other.revision_id_to_revno(stop_revision)
        try:
            return self.missing_revisions(other, other_revno)
        except DivergedBranches, e:
            try:
                pullable_revs = get_intervening_revisions(self.last_revision(),
                                                          stop_revision, self)
                assert self.last_revision() not in pullable_revs
                return pullable_revs
            except bzrlib.errors.NotAncestor:
                if is_ancestor(self.last_revision(), stop_revision, self):
                    return []
                else:
                    raise e
        
    def commit(self, *args, **kw):
        from bzrlib.commit import Commit
        Commit().commit(self, *args, **kw)
    
    def revision_id_to_revno(self, revision_id):
        """Given a revision id, return its revno"""
        if revision_id is None:
            return 0
        history = self.revision_history()
        try:
            return history.index(revision_id) + 1
        except ValueError:
            raise bzrlib.errors.NoSuchRevision(self, revision_id)

    def get_rev_id(self, revno, history=None):
        """Find the revision id of the specified revno."""
        if revno == 0:
            return None
        if history is None:
            history = self.revision_history()
        elif revno <= 0 or revno > len(history):
            raise bzrlib.errors.NoSuchRevision(self, revno)
        return history[revno - 1]

    def revision_tree(self, revision_id):
        """Return Tree for a revision on this branch.

        `revision_id` may be None for the null revision, in which case
        an `EmptyTree` is returned."""
        # TODO: refactor this to use an existing revision object
        # so we don't need to read it in twice.
        if revision_id == None:
            return EmptyTree()
        else:
            inv = self.get_revision_inventory(revision_id)
            return RevisionTree(self.weave_store, inv, revision_id)

    def working_tree(self):
        """Return a `Tree` for the working copy."""
        from bzrlib.workingtree import WorkingTree
        # TODO: In the future, perhaps WorkingTree should utilize Transport
        # RobertCollins 20051003 - I don't think it should - working trees are
        # much more complex to keep consistent than our careful .bzr subset.
        # instead, we should say that working trees are local only, and optimise
        # for that.
        return WorkingTree(self.base, branch=self)


    def basis_tree(self):
        """Return `Tree` object for last revision.

        If there are no revisions yet, return an `EmptyTree`.
        """
        return self.revision_tree(self.last_revision())

    @needs_write_lock
    def rename_one(self, from_rel, to_rel):
        """Rename one file.

        This can change the directory or the filename or both.
        """
        tree = self.working_tree()
        inv = tree.inventory
        if not tree.has_filename(from_rel):
            raise BzrError("can't rename: old working file %r does not exist" % from_rel)
        if tree.has_filename(to_rel):
            raise BzrError("can't rename: new working file %r already exists" % to_rel)

        file_id = inv.path2id(from_rel)
        if file_id == None:
            raise BzrError("can't rename: old name %r is not versioned" % from_rel)

        if inv.path2id(to_rel):
            raise BzrError("can't rename: new name %r is already versioned" % to_rel)

        to_dir, to_tail = os.path.split(to_rel)
        to_dir_id = inv.path2id(to_dir)
        if to_dir_id == None and to_dir != '':
            raise BzrError("can't determine destination directory id for %r" % to_dir)

        mutter("rename_one:")
        mutter("  file_id    {%s}" % file_id)
        mutter("  from_rel   %r" % from_rel)
        mutter("  to_rel     %r" % to_rel)
        mutter("  to_dir     %r" % to_dir)
        mutter("  to_dir_id  {%s}" % to_dir_id)

        inv.rename(file_id, to_dir_id, to_tail)

        from_abs = self.abspath(from_rel)
        to_abs = self.abspath(to_rel)
        try:
            rename(from_abs, to_abs)
        except OSError, e:
            raise BzrError("failed to rename %r to %r: %s"
                    % (from_abs, to_abs, e[1]),
                    ["rename rolled back"])

        self._write_inventory(inv)

    @needs_write_lock
    def move(self, from_paths, to_name):
        """Rename files.

        to_name must exist as a versioned directory.

        If to_name exists and is a directory, the files are moved into
        it, keeping their old names.  If it is a directory, 

        Note that to_name is only the last component of the new name;
        this doesn't change the directory.

        This returns a list of (from_path, to_path) pairs for each
        entry that is moved.
        """
        result = []
        ## TODO: Option to move IDs only
        assert not isinstance(from_paths, basestring)
        tree = self.working_tree()
        inv = tree.inventory
        to_abs = self.abspath(to_name)
        if not isdir(to_abs):
            raise BzrError("destination %r is not a directory" % to_abs)
        if not tree.has_filename(to_name):
            raise BzrError("destination %r not in working directory" % to_abs)
        to_dir_id = inv.path2id(to_name)
        if to_dir_id == None and to_name != '':
            raise BzrError("destination %r is not a versioned directory" % to_name)
        to_dir_ie = inv[to_dir_id]
        if to_dir_ie.kind not in ('directory', 'root_directory'):
            raise BzrError("destination %r is not a directory" % to_abs)

        to_idpath = inv.get_idpath(to_dir_id)

        for f in from_paths:
            if not tree.has_filename(f):
                raise BzrError("%r does not exist in working tree" % f)
            f_id = inv.path2id(f)
            if f_id == None:
                raise BzrError("%r is not versioned" % f)
            name_tail = splitpath(f)[-1]
            dest_path = appendpath(to_name, name_tail)
            if tree.has_filename(dest_path):
                raise BzrError("destination %r already exists" % dest_path)
            if f_id in to_idpath:
                raise BzrError("can't move %r to a subdirectory of itself" % f)

        # OK, so there's a race here, it's possible that someone will
        # create a file in this interval and then the rename might be
        # left half-done.  But we should have caught most problems.

        for f in from_paths:
            name_tail = splitpath(f)[-1]
            dest_path = appendpath(to_name, name_tail)
            result.append((f, dest_path))
            inv.rename(inv.path2id(f), to_dir_id, name_tail)
            try:
                rename(self.abspath(f), self.abspath(dest_path))
            except OSError, e:
                raise BzrError("failed to rename %r to %r: %s" % (f, dest_path, e[1]),
                        ["rename rolled back"])

        self._write_inventory(inv)
        return result


    def revert(self, filenames, old_tree=None, backups=True):
        """Restore selected files to the versions from a previous tree.

        backups
            If true (default) backups are made of files before
            they're renamed.
        """
        from bzrlib.atomicfile import AtomicFile
        from bzrlib.osutils import backup_file
        
        inv = self.read_working_inventory()
        if old_tree is None:
            old_tree = self.basis_tree()
        old_inv = old_tree.inventory

        nids = []
        for fn in filenames:
            file_id = inv.path2id(fn)
            if not file_id:
                raise NotVersionedError(path=fn)
            if not old_inv.has_id(file_id):
                raise BzrError("file not present in old tree", fn, file_id)
            nids.append((fn, file_id))
            
        # TODO: Rename back if it was previously at a different location

        # TODO: If given a directory, restore the entire contents from
        # the previous version.

        # TODO: Make a backup to a temporary file.

        # TODO: If the file previously didn't exist, delete it?
        for fn, file_id in nids:
            backup_file(fn)
            
            f = AtomicFile(fn, 'wb')
            try:
                f.write(old_tree.get_file(file_id).read())
                f.commit()
            finally:
                f.close()


    def pending_merges(self):
        """Return a list of pending merges.

        These are revisions that have been merged into the working
        directory but not yet committed.
        """
        cfn = self._rel_controlfilename('pending-merges')
        if not self._transport.has(cfn):
            return []
        p = []
        for l in self.controlfile('pending-merges', 'r').readlines():
            p.append(l.rstrip('\n'))
        return p


    def add_pending_merge(self, *revision_ids):
        # TODO: Perhaps should check at this point that the
        # history of the revision is actually present?
        p = self.pending_merges()
        updated = False
        for rev_id in revision_ids:
            if rev_id in p:
                continue
            p.append(rev_id)
            updated = True
        if updated:
            self.set_pending_merges(p)

    @needs_write_lock
    def set_pending_merges(self, rev_list):
        self.put_controlfile('pending-merges', '\n'.join(rev_list))

    def get_parent(self):
        """Return the parent location of the branch.

        This is the default location for push/pull/missing.  The usual
        pattern is that the user can override it by specifying a
        location.
        """
        import errno
        _locs = ['parent', 'pull', 'x-pull']
        for l in _locs:
            try:
                return self.controlfile(l, 'r').read().strip('\n')
            except IOError, e:
                if e.errno != errno.ENOENT:
                    raise
        return None

    @needs_write_lock
    def set_parent(self, url):
        # TODO: Maybe delete old location files?
        from bzrlib.atomicfile import AtomicFile
        f = AtomicFile(self.controlfilename('parent'))
        try:
            f.write(url + '\n')
            f.commit()
        finally:
            f.close()

    def check_revno(self, revno):
        """\
        Check whether a revno corresponds to any revision.
        Zero (the NULL revision) is considered valid.
        """
        if revno != 0:
            self.check_real_revno(revno)
            
    def check_real_revno(self, revno):
        """\
        Check whether a revno corresponds to a real revision.
        Zero (the NULL revision) is considered invalid
        """
        if revno < 1 or revno > self.revno():
            raise InvalidRevisionNumber(revno)
        
    def sign_revision(self, revision_id, gpg_strategy):
        plaintext = Testament.from_revision(self, revision_id).as_short_text()
        self.store_revision_signature(gpg_strategy, plaintext, revision_id)

    @needs_write_lock
    def store_revision_signature(self, gpg_strategy, plaintext, revision_id):
        self.revision_store.add(StringIO(gpg_strategy.sign(plaintext)), 
                                revision_id, "sig")


class ScratchBranch(_Branch):
    """Special test class: a branch that cleans up after itself.

    >>> b = ScratchBranch()
    >>> isdir(b.base)
    True
    >>> bd = b.base
    >>> b._transport.__del__()
    >>> isdir(bd)
    False
    """

    def __init__(self, files=[], dirs=[], transport=None):
        """Make a test branch.

        This creates a temporary directory and runs init-tree in it.

        If any files are listed, they are created in the working copy.
        """
        if transport is None:
            transport = bzrlib.transport.local.ScratchTransport()
            super(ScratchBranch, self).__init__(transport, init=True)
        else:
            super(ScratchBranch, self).__init__(transport)

        for d in dirs:
            self._transport.mkdir(d)
            
        for f in files:
            self._transport.put(f, 'content of %s' % f)


    def clone(self):
        """
        >>> orig = ScratchBranch(files=["file1", "file2"])
        >>> clone = orig.clone()
        >>> if os.name != 'nt':
        ...   os.path.samefile(orig.base, clone.base)
        ... else:
        ...   orig.base == clone.base
        ...
        False
        >>> os.path.isfile(os.path.join(clone.base, "file1"))
        True
        """
        from shutil import copytree
        from tempfile import mkdtemp
        base = mkdtemp()
        os.rmdir(base)
        copytree(self.base, base, symlinks=True)
        return ScratchBranch(
            transport=bzrlib.transport.local.ScratchTransport(base))
    

######################################################################
# predicates


def is_control_file(filename):
    ## FIXME: better check
    filename = os.path.normpath(filename)
    while filename != '':
        head, tail = os.path.split(filename)
        ## mutter('check %r for control file' % ((head, tail), ))
        if tail == bzrlib.BZRDIR:
            return True
        if filename == head:
            break
        filename = head
    return False



def gen_file_id(name):
    """Return new file id.

    This should probably generate proper UUIDs, but for the moment we
    cope with just randomness because running uuidgen every time is
    slow."""
    import re
    from binascii import hexlify
    from time import time

    # get last component
    idx = name.rfind('/')
    if idx != -1:
        name = name[idx+1 : ]
    idx = name.rfind('\\')
    if idx != -1:
        name = name[idx+1 : ]

    # make it not a hidden file
    name = name.lstrip('.')

    # remove any wierd characters; we don't escape them but rather
    # just pull them out
    name = re.sub(r'[^\w.]', '', name)

    s = hexlify(rand_bytes(8))
    return '-'.join((name, compact_date(time()), s))


def gen_root_id():
    """Return a new tree-root file id."""
    return gen_file_id('TREE_ROOT')