cvs2svn_lib/config.py

   1 # (Be in -*- python -*- mode.)
   2 #
   3 # ====================================================================
   4 # Copyright (c) 2000-2008 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 """This module contains various configuration constants used by cvs2svn."""
  18
  19
  20 SVN_KEYWORDS_VALUE = 'Author Date Id Revision'
  21
  22 # The default names for the trunk/branches/tags directory for each
  23 # project:
  24 DEFAULT_TRUNK_BASE = 'trunk'
  25 DEFAULT_BRANCHES_BASE = 'branches'
  26 DEFAULT_TAGS_BASE = 'tags'
  27
  28 SVNADMIN_EXECUTABLE = 'svnadmin'
  29 CO_EXECUTABLE = 'co'
  30 CVS_EXECUTABLE = 'cvs'
  31 SORT_EXECUTABLE = 'sort'
  32
  33 # A pickled list of the projects defined for this conversion.
  34 PROJECTS = 'projects.pck'
  35
  36 # A file holding the Serializer to be used for
  37 # CVS_REVS_SUMMARY_*_DATAFILE and CVS_SYMBOLS_SYMMARY_*_DATAFILE:
  38 SUMMARY_SERIALIZER = 'summary-serializer.pck'
  39
  40 # The first file contains enough information about each CVSRevision to
  41 # deduce preliminary Changesets.  The second file is a sorted version
  42 # of the first.
  43 CVS_REVS_SUMMARY_DATAFILE = 'revs-summary.txt'
  44 CVS_REVS_SUMMARY_SORTED_DATAFILE = 'revs-summary-s.txt'
  45
  46 # The first file contains enough information about each CVSSymbol to
  47 # deduce preliminary Changesets.  The second file is a sorted version
  48 # of the first.
  49 CVS_SYMBOLS_SUMMARY_DATAFILE = 'symbols-summary.txt'
  50 CVS_SYMBOLS_SUMMARY_SORTED_DATAFILE = 'symbols-summary-s.txt'
  51
  52 # A mapping from CVSItem id to Changeset id.
  53 CVS_ITEM_TO_CHANGESET = 'cvs-item-to-changeset.dat'
  54
  55 # A mapping from CVSItem id to Changeset id, after the
  56 # RevisionChangeset loops have been broken.
  57 CVS_ITEM_TO_CHANGESET_REVBROKEN = 'cvs-item-to-changeset-revbroken.dat'
  58
  59 # A mapping from CVSItem id to Changeset id, after the SymbolChangeset
  60 # loops have been broken.
  61 CVS_ITEM_TO_CHANGESET_SYMBROKEN = 'cvs-item-to-changeset-symbroken.dat'
  62
  63 # A mapping from CVSItem id to Changeset id, after all Changeset
  64 # loops have been broken.
  65 CVS_ITEM_TO_CHANGESET_ALLBROKEN = 'cvs-item-to-changeset-allbroken.dat'
  66
  67 # A mapping from id to Changeset.
  68 CHANGESETS_INDEX = 'changesets-index.dat'
  69 CHANGESETS_STORE = 'changesets.pck'
  70
  71 # A mapping from id to Changeset, after the RevisionChangeset loops
  72 # have been broken.
  73 CHANGESETS_REVBROKEN_INDEX = 'changesets-revbroken-index.dat'
  74 CHANGESETS_REVBROKEN_STORE = 'changesets-revbroken.pck'
  75
  76 # A mapping from id to Changeset, after the RevisionChangesets have
  77 # been sorted and converted into OrderedChangesets.
  78 CHANGESETS_REVSORTED_INDEX = 'changesets-revsorted-index.dat'
  79 CHANGESETS_REVSORTED_STORE = 'changesets-revsorted.pck'
  80
  81 # A mapping from id to Changeset, after the SymbolChangeset loops have
  82 # been broken.
  83 CHANGESETS_SYMBROKEN_INDEX = 'changesets-symbroken-index.dat'
  84 CHANGESETS_SYMBROKEN_STORE = 'changesets-symbroken.pck'
  85
  86 # A mapping from id to Changeset, after all Changeset loops have been
  87 # broken.
  88 CHANGESETS_ALLBROKEN_INDEX = 'changesets-allbroken-index.dat'
  89 CHANGESETS_ALLBROKEN_STORE = 'changesets-allbroken.pck'
  90
  91 # The RevisionChangesets in commit order.  Each line contains the
  92 # changeset id and timestamp of one changeset, in hexadecimal, in the
  93 # order that the changesets should be committed to svn.
  94 CHANGESETS_SORTED_DATAFILE = 'changesets-s.txt'
  95
  96 # A file containing a marshalled copy of all the statistics that have
  97 # been gathered so far is written at the end of each pass as a
  98 # marshalled dictionary.  This is the pattern used to generate the
  99 # filenames.
 100 STATISTICS_FILE = 'statistics-%02d.pck'
 101
 102 # This text file contains records (1 per line) that describe openings
 103 # and closings for copies to tags and branches.  The format is as
 104 # follows:
 105 #
 106 #     SYMBOL_ID SVN_REVNUM TYPE CVS_SYMBOL_ID
 107 #
 108 # where type is either OPENING or CLOSING.  CVS_SYMBOL_ID is the id of
 109 # the CVSSymbol whose opening or closing is being described (in hex).
 110 SYMBOL_OPENINGS_CLOSINGS = 'symbolic-names.txt'
 111 # A sorted version of the above file.  SYMBOL_ID and SVN_REVNUM are
 112 # the primary and secondary sorting criteria.  It is important that
 113 # SYMBOL_IDs be located together to make it quick to read them at
 114 # once.  The order of SVN_REVNUM is only important because it is
 115 # assumed by some internal consistency checks.
 116 SYMBOL_OPENINGS_CLOSINGS_SORTED = 'symbolic-names-s.txt'
 117
 118 # Skeleton version of the repository filesystem.  See class
 119 # RepositoryMirror for how these work.
 120 MIRROR_NODES_INDEX_TABLE = 'mirror-nodes-index.dat'
 121 MIRROR_NODES_STORE = 'mirror-nodes.pck'
 122
 123 # Offsets pointing to the beginning of each symbol's records in
 124 # SYMBOL_OPENINGS_CLOSINGS_SORTED.  This file contains a pickled map
 125 # from symbol_id to file offset.
 126 SYMBOL_OFFSETS_DB = 'symbol-offsets.pck'
 127
 128 # Pickled map of CVSFile.id to instance.
 129 CVS_FILES_DB = 'cvs-files.pck'
 130
 131 # A series of records.  The first is a pickled serializer.  Each
 132 # subsequent record is a serialized list of all CVSItems applying to a
 133 # CVSFile.
 134 CVS_ITEMS_STORE = 'cvs-items.pck'
 135
 136 # The same as above, but with the CVSItems ordered in groups based on
 137 # their initial changesets.  CVSItems will usually be accessed one
 138 # changeset at a time, so this ordering helps disk locality (even
 139 # though some of the changesets will later be broken up).
 140 CVS_ITEMS_SORTED_INDEX_TABLE = 'cvs-items-sorted-index.dat'
 141 CVS_ITEMS_SORTED_STORE = 'cvs-items-sorted.pck'
 142
 143 # A record of all symbolic names that will be processed in the
 144 # conversion.  This file contains a pickled list of TypedSymbol
 145 # objects.
 146 SYMBOL_DB = 'symbols.pck'
 147
 148 # A pickled list of the statistics for all symbols.  Each entry in the
 149 # list is an instance of cvs2svn_lib.symbol_statistics._Stats.
 150 SYMBOL_STATISTICS = 'symbol-statistics.pck'
 151
 152 # These two databases provide a bidirectional mapping between
 153 # CVSRevision.ids (in hex) and Subversion revision numbers.
 154 #
 155 # The first maps CVSRevision.id to the SVN revision number of which it
 156 # is a part (more than one CVSRevision can map to the same SVN
 157 # revision number).
 158 #
 159 # The second maps Subversion revision numbers (as hex strings) to
 160 # pickled SVNCommit instances.
 161 CVS_REVS_TO_SVN_REVNUMS = 'cvs-revs-to-svn-revnums.dat'
 162
 163 # This database maps Subversion revision numbers to pickled SVNCommit
 164 # instances.
 165 SVN_COMMITS_INDEX_TABLE = 'svn-commits-index.dat'
 166 SVN_COMMITS_STORE = 'svn-commits.pck'
 167
 168 # How many bytes to read at a time from a pipe.  128 kiB should be
 169 # large enough to be efficient without wasting too much memory.
 170 PIPE_READ_SIZE = 128 * 1024
 171
 172 # Records the author and log message for each changeset.  The database
 173 # contains a map metadata_id -> (author, logmessage).  Each
 174 # CVSRevision that is eligible to be combined into the same SVN commit
 175 # is assigned the same id.  Note that the (author, logmessage) pairs
 176 # are not necessarily all distinct; other data are taken into account
 177 # when constructing ids.
 178 METADATA_INDEX_TABLE = 'metadata-index.dat'
 179 METADATA_STORE = 'metadata.pck'
 180
 181 # The same, after it has been cleaned up for the chosen output option:
 182 METADATA_CLEAN_INDEX_TABLE = 'metadata-clean-index.dat'
 183 METADATA_CLEAN_STORE = 'metadata-clean.pck'
 184
 185 # The following four databases are used in conjunction with --use-internal-co.
 186
 187 # Records the RCS deltas for all CVS revisions.  The deltas are to be
 188 # applied forward, i.e. those from trunk are reversed wrt RCS.
 189 RCS_DELTAS_INDEX_TABLE = 'rcs-deltas-index.dat'
 190 RCS_DELTAS_STORE = 'rcs-deltas.pck'
 191
 192 # Records the revision tree of each RCS file.  The format is a list of
 193 # list of integers.  The outer list holds lines of development, the inner list
 194 # revisions within the LODs, revisions are CVSItem ids.  Branches "closer
 195 # to the trunk" appear later.  Revisions are sorted by reverse chronological
 196 # order.  The last revision of each branch is the revision it sprouts from.
 197 # Revisions that represent deletions at the end of a branch are omitted.
 198 RCS_TREES_INDEX_TABLE = 'rcs-trees-index.dat'
 199 RCS_TREES_STORE = 'rcs-trees.pck'
 200
 201 # Records the revision tree of each RCS file after removing revisions
 202 # belonging to excluded branches.  Note that the branch ordering is arbitrary
 203 # in this file.
 204 RCS_TREES_FILTERED_INDEX_TABLE = 'rcs-trees-filtered-index.dat'
 205 RCS_TREES_FILTERED_STORE = 'rcs-trees-filtered.pck'
 206
 207 # At any given time during OutputPass, holds the full text of each CVS
 208 # revision that was checked out already and still has descendants that will
 209 # be checked out.
 210 CVS_CHECKOUT_DB = 'cvs-checkout.db'
 211
 212 # End of DBs related to --use-internal-co.
 213
 214 # If this run will output directly to a Subversion repository, then
 215 # this is the name of the file that each revision will temporarily be
 216 # written to prior to writing it into the repository.
 217 DUMPFILE = 'svn.dump'
 218
 219 # flush a commit if a 5 minute gap occurs.
 220 COMMIT_THRESHOLD = 5 * 60
 221