cvs2svn_lib/external_blob_generator.py

   1 # (Be in -*- python -*- mode.)
   2 #
   3 # ====================================================================
   4 # Copyright (c) 2009-2010 CollabNet.  All rights reserved.
   5 #
   6 # This software is licensed as described in the file COPYING, which
   7 # you should have received as part of this distribution.  The terms
   8 # are also available at http://subversion.tigris.org/license-1.html.
   9 # If newer versions of this license are posted there, you may use a
  10 # newer version instead, at your option.
  11 #
  12 # This software consists of voluntary contributions made by many
  13 # individuals.  For exact contribution history, see the revision
  14 # history and logs, available at http://cvs2svn.tigris.org/.
  15 # ====================================================================
  16
  17 """Use the generate_blobs.py script to generate git blobs.
  18
  19 Use a separate program, generate_blobs.py, to generate a git blob file
  20 directly from the RCS files, setting blob marks that we choose.  This
  21 method is very much faster then generating the blobs from within the
  22 main program for several reasons:
  23
  24 * The revision fulltexts are generated using internal code (rather
  25   than spawning rcs or cvs once per revision).  This gain is analogous
  26   to the benefit of using --use-internal-co rather than --use-cvs or
  27   --use-rcs for cvs2svn.
  28
  29 * Intermediate revisions' fulltext can usually be held in RAM rather
  30   than being written to temporary storage, and output as they are
  31   generated (git-fast-import doesn't care about their order).
  32
  33 * The generate_blobs.py script runs in parallel to the main cvs2git
  34   script, allowing benefits to be had from multiple CPUs.
  35
  36 """
  37
  38 import sys
  39 import os
  40 import subprocess
  41 import cPickle as pickle
  42 import itertools
  43
  44 from cvs2svn_lib.common import FatalError
  45 from cvs2svn_lib.log import Log
  46 from cvs2svn_lib.symbol import Trunk
  47 from cvs2svn_lib.cvs_item import CVSRevisionDelete
  48 from cvs2svn_lib.cvs_item import CVSSymbol
  49 from cvs2svn_lib.revision_manager import RevisionCollector
  50 from cvs2svn_lib.key_generator import KeyGenerator
  51
  52
  53 class ExternalBlobGenerator(RevisionCollector):
  54   """Have generate_blobs.py output file revisions to a blob file."""
  55
  56   def __init__(self, blob_filename):
  57     self.blob_filename = blob_filename
  58
  59   def start(self):
  60     self._mark_generator = KeyGenerator()
  61     Log().normal('Starting generate_blobs.py...')
  62     self._popen = subprocess.Popen(
  63         [
  64             sys.executable,
  65             os.path.join(os.path.dirname(__file__), 'generate_blobs.py'),
  66             self.blob_filename,
  67             ],
  68         stdin=subprocess.PIPE,
  69         )
  70
  71   def _process_symbol(self, cvs_symbol, cvs_file_items):
  72     """Record the original source of CVS_SYMBOL.
  73
  74     Determine the original revision source of CVS_SYMBOL, and store it
  75     as the symbol's revision_reader_token."""
  76
  77     cvs_source = cvs_symbol.get_cvs_revision_source(cvs_file_items)
  78     cvs_symbol.revision_reader_token = cvs_source.revision_reader_token
  79
  80   def process_file(self, cvs_file_items):
  81     marks = {}
  82     for lod_items in cvs_file_items.iter_lods():
  83       for cvs_rev in lod_items.cvs_revisions:
  84         if not isinstance(cvs_rev, CVSRevisionDelete):
  85           mark = self._mark_generator.gen_id()
  86           cvs_rev.revision_reader_token = mark
  87           marks[cvs_rev.rev] = mark
  88
  89     # A separate pickler is used for each dump(), so that its memo
  90     # doesn't grow very large.  The default ASCII protocol is used so
  91     # that this works without changes on systems that distinguish
  92     # between text and binary files.
  93     pickle.dump((cvs_file_items.cvs_file.filename, marks), self._popen.stdin)
  94     self._popen.stdin.flush()
  95
  96     # Now that all CVSRevisions' revision_reader_tokens are set,
  97     # iterate through symbols and set their tokens to those of their
  98     # original source revisions:
  99     for lod_items in cvs_file_items.iter_lods():
 100       if lod_items.cvs_branch is not None:
 101         self._process_symbol(lod_items.cvs_branch, cvs_file_items)
 102       for cvs_tag in lod_items.cvs_tags:
 103         self._process_symbol(cvs_tag, cvs_file_items)
 104
 105   def finish(self):
 106     self._popen.stdin.close()
 107     Log().normal('Waiting for generate_blobs.py to finish...')
 108     returncode = self._popen.wait()
 109     if returncode:
 110       raise FatalError(
 111           'generate_blobs.py failed with return code %s.' % (returncode,)
 112           )
 113     else:
 114       Log().normal('generate_blobs.py is done.')
 115
 116