/brz/remove-bazaar

#! /usr/bin/python

#

# 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

"""Experiment in converting existing bzr branches to weaves."""

# To make this properly useful

#

# 1. assign text version ids, and put those text versions into

#    the inventory as they're converted.

#

# 2. keep track of the previous version of each file, rather than

#    just using the last one imported

#

# 3. assign entry versions when files are added, renamed or moved.

#

# 4. when merged-in versions are observed, walk down through them

#    to discover everything, then commit bottom-up

#

# 5. track ancestry as things are merged in, and commit that in each

#    revision

#

# Perhaps it's best to first walk the whole graph and make a plan for

# what should be imported in what order?  Need a kind of topological

# sort of all revisions.  (Or do we, can we just before doing a revision

# see that all its parents have either been converted or abandoned?)

if False:

    try:

        import psyco

        psyco.full()

    except ImportError:

        pass

import tempfile

import hotshot, hotshot.stats

import sys

import logging

import time

from bzrlib.branch import Branch, find_branch

from bzrlib.revfile import Revfile

from bzrlib.weave import Weave

from bzrlib.weavefile import read_weave, write_weave

from bzrlib.progress import ProgressBar

from bzrlib.atomicfile import AtomicFile

from bzrlib.xml4 import serializer_v4

from bzrlib.xml5 import serializer_v5

import bzrlib.trace

class Convert(object):

    def __init__(self):

        self.total_revs = 0

        self.converted_revs = 0

        self.text_count = 0

        self.convert()

    def convert(self):

        bzrlib.trace.enable_default_logging()

        self.pb = ProgressBar()

        self.inv_weave = Weave('__inventory')

        self.anc_weave = Weave('__ancestry')

        last_text_sha = {}

        # holds in-memory weaves for all files

        text_weaves = {}

        b = self.branch = Branch('.', relax_version_check=True)

        revno = 1

        rev_history = b.revision_history()

        last_idx = None

        inv_parents = []

        # todo is a stack holding the revisions we still need to process;

        # appending to it adds new highest-priority revisions

        todo = rev_history[:]

        todo.reverse()

        self.total_revs = len(todo)

        while todo:

            self._convert_one_rev(todo.pop())

        self.pb.clear()

        print 'upgraded to weaves:'

        print '  %6d revisions and inventories' % self.converted_revs

        print '  %6d texts' % self.text_count

        self._write_all_weaves()

    def _write_all_weaves(self):

        i = 0

        return ############################################

        # TODO: commit them all atomically at the end, not one by one

        write_atomic_weave(self.inv_weave, 'weaves/inventory.weave')

        write_atomic_weave(self.anc_weave, 'weaves/ancestry.weave')

        for file_id, file_weave in text_weaves.items():

            self.pb.update('writing weave', i, len(text_weaves))

            write_atomic_weave(file_weave, 'weaves/%s.weave' % file_id)

            i += 1

        self.pb.clear()

    def _convert_one_rev(self, rev_id):

        self._bump_progress()

        b = self.branch

        rev_xml = b.revision_store[rev_id].read()

        inv_xml = b.inventory_store[rev_id].read()

        rev = serializer_v4.read_revision_from_string(rev_xml)

        inv = serializer_v4.read_inventory_from_string(inv_xml)

        return ##########################################

        new_idx = self.inv_weave.add(rev_id, inv_parents, inv_xml)

        inv_parents = [new_idx]

        tree = b.revision_tree(rev_id)

        inv = tree.inventory

        # for each file in the inventory, put it into its own revfile

        for file_id in inv:

            ie = inv[file_id]

            if ie.kind != 'file':

                continue

            if last_text_sha.get(file_id) == ie.text_sha1:

                # same as last time

                continue

            last_text_sha[file_id] = ie.text_sha1

            # new text (though possibly already stored); need to store it

            text_lines = tree.get_file(file_id).readlines()

            # if the file's created for the first time in this

            # revision then make a new weave; else find the old one

            if file_id not in text_weaves:

                text_weaves[file_id] = Weave()

            w = text_weaves[file_id]

            # base the new text version off whatever was last

            # (actually it'd be better to track this, to allow for

            # files that are deleted and then reappear)

            last = len(w)

            if last == 0:

                parents = []

            else:

                parents = [last-1]

            w.add(rev_id, parents, text_lines)

            text_count += 1

        revno += 1

    def _bump_progress(self):

        self.converted_revs += 1

        self.pb.update('converting revisions',

                       self.converted_revs,

                       self.total_revs)

def write_atomic_weave(weave, filename):

    inv_wf = AtomicFile(filename)

    try:

        write_weave(weave, inv_wf)

        inv_wf.commit()

    finally:

        inv_wf.close()

def profile_convert(): 

    prof_f = tempfile.NamedTemporaryFile()

    prof = hotshot.Profile(prof_f.name)

    prof.runcall(Convert) 

    prof.close()

    stats = hotshot.stats.load(prof_f.name)

    ##stats.strip_dirs()

    stats.sort_stats('time')

    # XXX: Might like to write to stderr or the trace file instead but

    # print_stats seems hardcoded to stdout

    stats.print_stats(20)

if '-p' in sys.argv[1:]:

    profile_convert()

else:

    Convert()