Show More
@@ -1,318 +1,321 b'' | |||
|
1 | 1 | # convert.py Foreign SCM converter |
|
2 | 2 | # |
|
3 | 3 | # Copyright 2005-2007 Matt Mackall <mpm@selenic.com> |
|
4 | 4 | # |
|
5 | 5 | # This software may be used and distributed according to the terms of the |
|
6 | 6 | # GNU General Public License version 2 or any later version. |
|
7 | 7 | |
|
8 | 8 | '''import revisions from foreign VCS repositories into Mercurial''' |
|
9 | 9 | |
|
10 | 10 | import convcmd |
|
11 | 11 | import cvsps |
|
12 | 12 | import subversion |
|
13 | 13 | from mercurial import commands |
|
14 | 14 | from mercurial.i18n import _ |
|
15 | 15 | |
|
16 | 16 | # Commands definition was moved elsewhere to ease demandload job. |
|
17 | 17 | |
|
18 | 18 | def convert(ui, src, dest=None, revmapfile=None, **opts): |
|
19 | 19 | """convert a foreign SCM repository to a Mercurial one. |
|
20 | 20 | |
|
21 | 21 | Accepted source formats [identifiers]: |
|
22 | 22 | |
|
23 | 23 | - Mercurial [hg] |
|
24 | 24 | - CVS [cvs] |
|
25 | 25 | - Darcs [darcs] |
|
26 | 26 | - git [git] |
|
27 | 27 | - Subversion [svn] |
|
28 | 28 | - Monotone [mtn] |
|
29 | 29 | - GNU Arch [gnuarch] |
|
30 | 30 | - Bazaar [bzr] |
|
31 | 31 | - Perforce [p4] |
|
32 | 32 | |
|
33 | 33 | Accepted destination formats [identifiers]: |
|
34 | 34 | |
|
35 | 35 | - Mercurial [hg] |
|
36 | 36 | - Subversion [svn] (history on branches is not preserved) |
|
37 | 37 | |
|
38 | 38 | If no revision is given, all revisions will be converted. |
|
39 | 39 | Otherwise, convert will only import up to the named revision |
|
40 | 40 | (given in a format understood by the source). |
|
41 | 41 | |
|
42 | 42 | If no destination directory name is specified, it defaults to the |
|
43 | 43 | basename of the source with ``-hg`` appended. If the destination |
|
44 | 44 | repository doesn't exist, it will be created. |
|
45 | 45 | |
|
46 | 46 | By default, all sources except Mercurial will use --branchsort. |
|
47 | 47 | Mercurial uses --sourcesort to preserve original revision numbers |
|
48 | 48 | order. Sort modes have the following effects: |
|
49 | 49 | |
|
50 | 50 | --branchsort convert from parent to child revision when possible, |
|
51 | 51 | which means branches are usually converted one after |
|
52 | 52 | the other. It generates more compact repositories. |
|
53 | 53 | |
|
54 | 54 | --datesort sort revisions by date. Converted repositories have |
|
55 | 55 | good-looking changelogs but are often an order of |
|
56 | 56 | magnitude larger than the same ones generated by |
|
57 | 57 | --branchsort. |
|
58 | 58 | |
|
59 | 59 | --sourcesort try to preserve source revisions order, only |
|
60 | 60 | supported by Mercurial sources. |
|
61 | 61 | |
|
62 | 62 | If <REVMAP> isn't given, it will be put in a default location |
|
63 | 63 | (<dest>/.hg/shamap by default). The <REVMAP> is a simple text file |
|
64 | 64 | that maps each source commit ID to the destination ID for that |
|
65 | 65 | revision, like so:: |
|
66 | 66 | |
|
67 | 67 | <source ID> <destination ID> |
|
68 | 68 | |
|
69 | 69 | If the file doesn't exist, it's automatically created. It's |
|
70 | 70 | updated on each commit copied, so :hg:`convert` can be interrupted |
|
71 | 71 | and can be run repeatedly to copy new commits. |
|
72 | 72 | |
|
73 |
The |
|
|
74 |
|
|
|
75 |
|
|
|
76 |
|
|
|
73 | The authormap is a simple text file that maps each source commit | |
|
74 | author to a destination commit author. It is handy for source SCMs | |
|
75 | that use unix logins to identify authors (eg: CVS). One line per | |
|
76 | author mapping and the line format is:: | |
|
77 | 77 | |
|
78 | 78 | source author = destination author |
|
79 | 79 | |
|
80 | 80 | Empty lines and lines starting with a ``#`` are ignored. |
|
81 | 81 | |
|
82 | 82 | The filemap is a file that allows filtering and remapping of files |
|
83 | 83 | and directories. Each line can contain one of the following |
|
84 | 84 | directives:: |
|
85 | 85 | |
|
86 | 86 | include path/to/file-or-dir |
|
87 | 87 | |
|
88 | 88 | exclude path/to/file-or-dir |
|
89 | 89 | |
|
90 | 90 | rename path/to/source path/to/destination |
|
91 | 91 | |
|
92 | 92 | Comment lines start with ``#``. A specified path matches if it |
|
93 | 93 | equals the full relative name of a file or one of its parent |
|
94 | 94 | directories. The ``include`` or ``exclude`` directive with the |
|
95 | 95 | longest matching path applies, so line order does not matter. |
|
96 | 96 | |
|
97 | 97 | The ``include`` directive causes a file, or all files under a |
|
98 | 98 | directory, to be included in the destination repository, and the |
|
99 | 99 | exclusion of all other files and directories not explicitly |
|
100 | 100 | included. The ``exclude`` directive causes files or directories to |
|
101 | 101 | be omitted. The ``rename`` directive renames a file or directory if |
|
102 | 102 | it is converted. To rename from a subdirectory into the root of |
|
103 | 103 | the repository, use ``.`` as the path to rename to. |
|
104 | 104 | |
|
105 | 105 | The splicemap is a file that allows insertion of synthetic |
|
106 | 106 | history, letting you specify the parents of a revision. This is |
|
107 | 107 | useful if you want to e.g. give a Subversion merge two parents, or |
|
108 | 108 | graft two disconnected series of history together. Each entry |
|
109 | 109 | contains a key, followed by a space, followed by one or two |
|
110 | 110 | comma-separated values:: |
|
111 | 111 | |
|
112 | 112 | key parent1, parent2 |
|
113 | 113 | |
|
114 | 114 | The key is the revision ID in the source |
|
115 | 115 | revision control system whose parents should be modified (same |
|
116 | 116 | format as a key in .hg/shamap). The values are the revision IDs |
|
117 | 117 | (in either the source or destination revision control system) that |
|
118 | 118 | should be used as the new parents for that node. For example, if |
|
119 | 119 | you have merged "release-1.0" into "trunk", then you should |
|
120 | 120 | specify the revision on "trunk" as the first parent and the one on |
|
121 | 121 | the "release-1.0" branch as the second. |
|
122 | 122 | |
|
123 | 123 | The branchmap is a file that allows you to rename a branch when it is |
|
124 | 124 | being brought in from whatever external repository. When used in |
|
125 | 125 | conjunction with a splicemap, it allows for a powerful combination |
|
126 | 126 | to help fix even the most badly mismanaged repositories and turn them |
|
127 | 127 | into nicely structured Mercurial repositories. The branchmap contains |
|
128 | 128 | lines of the form:: |
|
129 | 129 | |
|
130 | 130 | original_branch_name new_branch_name |
|
131 | 131 | |
|
132 | 132 | where "original_branch_name" is the name of the branch in the |
|
133 | 133 | source repository, and "new_branch_name" is the name of the branch |
|
134 | 134 | is the destination repository. No whitespace is allowed in the |
|
135 | 135 | branch names. This can be used to (for instance) move code in one |
|
136 | 136 | repository from "default" to a named branch. |
|
137 | 137 | |
|
138 | 138 | Mercurial Source |
|
139 | 139 | ---------------- |
|
140 | 140 | |
|
141 | 141 | --config convert.hg.ignoreerrors=False (boolean) |
|
142 | 142 | ignore integrity errors when reading. Use it to fix Mercurial |
|
143 | 143 | repositories with missing revlogs, by converting from and to |
|
144 | 144 | Mercurial. |
|
145 | 145 | --config convert.hg.saverev=False (boolean) |
|
146 | 146 | store original revision ID in changeset (forces target IDs to |
|
147 | 147 | change) |
|
148 | 148 | --config convert.hg.startrev=0 (hg revision identifier) |
|
149 | 149 | convert start revision and its descendants |
|
150 | 150 | |
|
151 | 151 | CVS Source |
|
152 | 152 | ---------- |
|
153 | 153 | |
|
154 | 154 | CVS source will use a sandbox (i.e. a checked-out copy) from CVS |
|
155 | 155 | to indicate the starting point of what will be converted. Direct |
|
156 | 156 | access to the repository files is not needed, unless of course the |
|
157 | 157 | repository is :local:. The conversion uses the top level directory |
|
158 | 158 | in the sandbox to find the CVS repository, and then uses CVS rlog |
|
159 | 159 | commands to find files to convert. This means that unless a |
|
160 | 160 | filemap is given, all files under the starting directory will be |
|
161 | 161 | converted, and that any directory reorganization in the CVS |
|
162 | 162 | sandbox is ignored. |
|
163 | 163 | |
|
164 | 164 | The options shown are the defaults. |
|
165 | 165 | |
|
166 | 166 | --config convert.cvsps.cache=True (boolean) |
|
167 | 167 | Set to False to disable remote log caching, for testing and |
|
168 | 168 | debugging purposes. |
|
169 | 169 | --config convert.cvsps.fuzz=60 (integer) |
|
170 | 170 | Specify the maximum time (in seconds) that is allowed between |
|
171 | 171 | commits with identical user and log message in a single |
|
172 | 172 | changeset. When very large files were checked in as part of a |
|
173 | 173 | changeset then the default may not be long enough. |
|
174 | 174 | --config convert.cvsps.mergeto='{{mergetobranch ([-\\w]+)}}' |
|
175 | 175 | Specify a regular expression to which commit log messages are |
|
176 | 176 | matched. If a match occurs, then the conversion process will |
|
177 | 177 | insert a dummy revision merging the branch on which this log |
|
178 | 178 | message occurs to the branch indicated in the regex. |
|
179 | 179 | --config convert.cvsps.mergefrom='{{mergefrombranch ([-\\w]+)}}' |
|
180 | 180 | Specify a regular expression to which commit log messages are |
|
181 | 181 | matched. If a match occurs, then the conversion process will |
|
182 | 182 | add the most recent revision on the branch indicated in the |
|
183 | 183 | regex as the second parent of the changeset. |
|
184 | 184 | --config hook.cvslog |
|
185 | 185 | Specify a Python function to be called at the end of gathering |
|
186 | 186 | the CVS log. The function is passed a list with the log entries, |
|
187 | 187 | and can modify the entries in-place, or add or delete them. |
|
188 | 188 | --config hook.cvschangesets |
|
189 | 189 | Specify a Python function to be called after the changesets |
|
190 | 190 | are calculated from the the CVS log. The function is passed |
|
191 | 191 | a list with the changeset entries, and can modify the changesets |
|
192 | 192 | in-place, or add or delete them. |
|
193 | 193 | |
|
194 | 194 | An additional "debugcvsps" Mercurial command allows the builtin |
|
195 | 195 | changeset merging code to be run without doing a conversion. Its |
|
196 | 196 | parameters and output are similar to that of cvsps 2.1. Please see |
|
197 | 197 | the command help for more details. |
|
198 | 198 | |
|
199 | 199 | Subversion Source |
|
200 | 200 | ----------------- |
|
201 | 201 | |
|
202 | 202 | Subversion source detects classical trunk/branches/tags layouts. |
|
203 | 203 | By default, the supplied "svn://repo/path/" source URL is |
|
204 | 204 | converted as a single branch. If "svn://repo/path/trunk" exists it |
|
205 | 205 | replaces the default branch. If "svn://repo/path/branches" exists, |
|
206 | 206 | its subdirectories are listed as possible branches. If |
|
207 | 207 | "svn://repo/path/tags" exists, it is looked for tags referencing |
|
208 | 208 | converted branches. Default "trunk", "branches" and "tags" values |
|
209 | 209 | can be overridden with following options. Set them to paths |
|
210 | 210 | relative to the source URL, or leave them blank to disable auto |
|
211 | 211 | detection. |
|
212 | 212 | |
|
213 | 213 | --config convert.svn.branches=branches (directory name) |
|
214 | 214 | specify the directory containing branches |
|
215 | 215 | --config convert.svn.tags=tags (directory name) |
|
216 | 216 | specify the directory containing tags |
|
217 | 217 | --config convert.svn.trunk=trunk (directory name) |
|
218 | 218 | specify the name of the trunk branch |
|
219 | 219 | |
|
220 | 220 | Source history can be retrieved starting at a specific revision, |
|
221 | 221 | instead of being integrally converted. Only single branch |
|
222 | 222 | conversions are supported. |
|
223 | 223 | |
|
224 | 224 | --config convert.svn.startrev=0 (svn revision number) |
|
225 | 225 | specify start Subversion revision. |
|
226 | 226 | |
|
227 | 227 | Perforce Source |
|
228 | 228 | --------------- |
|
229 | 229 | |
|
230 | 230 | The Perforce (P4) importer can be given a p4 depot path or a |
|
231 | 231 | client specification as source. It will convert all files in the |
|
232 | 232 | source to a flat Mercurial repository, ignoring labels, branches |
|
233 | 233 | and integrations. Note that when a depot path is given you then |
|
234 | 234 | usually should specify a target directory, because otherwise the |
|
235 | 235 | target may be named ...-hg. |
|
236 | 236 | |
|
237 | 237 | It is possible to limit the amount of source history to be |
|
238 | 238 | converted by specifying an initial Perforce revision. |
|
239 | 239 | |
|
240 | 240 | --config convert.p4.startrev=0 (perforce changelist number) |
|
241 | 241 | specify initial Perforce revision. |
|
242 | 242 | |
|
243 | 243 | Mercurial Destination |
|
244 | 244 | --------------------- |
|
245 | 245 | |
|
246 | 246 | --config convert.hg.clonebranches=False (boolean) |
|
247 | 247 | dispatch source branches in separate clones. |
|
248 | 248 | --config convert.hg.tagsbranch=default (branch name) |
|
249 | 249 | tag revisions branch name |
|
250 | 250 | --config convert.hg.usebranchnames=True (boolean) |
|
251 | 251 | preserve branch names |
|
252 | 252 | |
|
253 | 253 | """ |
|
254 | 254 | return convcmd.convert(ui, src, dest, revmapfile, **opts) |
|
255 | 255 | |
|
256 | 256 | def debugsvnlog(ui, **opts): |
|
257 | 257 | return subversion.debugsvnlog(ui, **opts) |
|
258 | 258 | |
|
259 | 259 | def debugcvsps(ui, *args, **opts): |
|
260 | 260 | '''create changeset information from CVS |
|
261 | 261 | |
|
262 | 262 | This command is intended as a debugging tool for the CVS to |
|
263 | 263 | Mercurial converter, and can be used as a direct replacement for |
|
264 | 264 | cvsps. |
|
265 | 265 | |
|
266 | 266 | Hg debugcvsps reads the CVS rlog for current directory (or any |
|
267 | 267 | named directory) in the CVS repository, and converts the log to a |
|
268 | 268 | series of changesets based on matching commit log entries and |
|
269 | 269 | dates.''' |
|
270 | 270 | return cvsps.debugcvsps(ui, *args, **opts) |
|
271 | 271 | |
|
272 | 272 | commands.norepo += " convert debugsvnlog debugcvsps" |
|
273 | 273 | |
|
274 | 274 | cmdtable = { |
|
275 | 275 | "convert": |
|
276 | 276 | (convert, |
|
277 | 277 | [('A', 'authors', '', |
|
278 |
_('username mapping filename'), |
|
|
278 | _('username mapping filename (DEPRECATED, use --authormap instead)'), | |
|
279 | _('FILE')), | |
|
279 | 280 | ('s', 'source-type', '', |
|
280 | 281 | _('source repository type'), _('TYPE')), |
|
281 | 282 | ('d', 'dest-type', '', |
|
282 | 283 | _('destination repository type'), _('TYPE')), |
|
283 | 284 | ('r', 'rev', '', |
|
284 | 285 | _('import up to target revision REV'), _('REV')), |
|
286 | ('', 'authormap', '', | |
|
287 | _('remap usernames using this file'), _('FILE')), | |
|
285 | 288 | ('', 'filemap', '', |
|
286 | 289 | _('remap file names using contents of file'), _('FILE')), |
|
287 | 290 | ('', 'splicemap', '', |
|
288 | 291 | _('splice synthesized history into place'), _('FILE')), |
|
289 | 292 | ('', 'branchmap', '', |
|
290 | 293 | _('change branch names while converting'), _('FILE')), |
|
291 | 294 | ('', 'branchsort', None, _('try to sort changesets by branches')), |
|
292 | 295 | ('', 'datesort', None, _('try to sort changesets by date')), |
|
293 | 296 | ('', 'sourcesort', None, _('preserve source changesets order'))], |
|
294 | 297 | _('hg convert [OPTION]... SOURCE [DEST [REVMAP]]')), |
|
295 | 298 | "debugsvnlog": |
|
296 | 299 | (debugsvnlog, |
|
297 | 300 | [], |
|
298 | 301 | 'hg debugsvnlog'), |
|
299 | 302 | "debugcvsps": |
|
300 | 303 | (debugcvsps, |
|
301 | 304 | [ |
|
302 | 305 | # Main options shared with cvsps-2.1 |
|
303 | 306 | ('b', 'branches', [], _('only return changes on specified branches')), |
|
304 | 307 | ('p', 'prefix', '', _('prefix to remove from file names')), |
|
305 | 308 | ('r', 'revisions', [], |
|
306 | 309 | _('only return changes after or between specified tags')), |
|
307 | 310 | ('u', 'update-cache', None, _("update cvs log cache")), |
|
308 | 311 | ('x', 'new-cache', None, _("create new cvs log cache")), |
|
309 | 312 | ('z', 'fuzz', 60, _('set commit time fuzz in seconds')), |
|
310 | 313 | ('', 'root', '', _('specify cvsroot')), |
|
311 | 314 | # Options specific to builtin cvsps |
|
312 | 315 | ('', 'parents', '', _('show parent changesets')), |
|
313 | 316 | ('', 'ancestors', '', _('show current changeset in ancestor branches')), |
|
314 | 317 | # Options that are ignored for compatibility with cvsps-2.1 |
|
315 | 318 | ('A', 'cvs-direct', None, _('ignored for compatibility')), |
|
316 | 319 | ], |
|
317 | 320 | _('hg debugcvsps [OPTION]... [PATH]...')), |
|
318 | 321 | } |
@@ -1,430 +1,434 b'' | |||
|
1 | 1 | # convcmd - convert extension commands definition |
|
2 | 2 | # |
|
3 | 3 | # Copyright 2005-2007 Matt Mackall <mpm@selenic.com> |
|
4 | 4 | # |
|
5 | 5 | # This software may be used and distributed according to the terms of the |
|
6 | 6 | # GNU General Public License version 2 or any later version. |
|
7 | 7 | |
|
8 | 8 | from common import NoRepo, MissingTool, SKIPREV, mapfile |
|
9 | 9 | from cvs import convert_cvs |
|
10 | 10 | from darcs import darcs_source |
|
11 | 11 | from git import convert_git |
|
12 | 12 | from hg import mercurial_source, mercurial_sink |
|
13 | 13 | from subversion import svn_source, svn_sink |
|
14 | 14 | from monotone import monotone_source |
|
15 | 15 | from gnuarch import gnuarch_source |
|
16 | 16 | from bzr import bzr_source |
|
17 | 17 | from p4 import p4_source |
|
18 | 18 | import filemap |
|
19 | 19 | |
|
20 | 20 | import os, shutil |
|
21 | 21 | from mercurial import hg, util, encoding |
|
22 | 22 | from mercurial.i18n import _ |
|
23 | 23 | |
|
24 | 24 | orig_encoding = 'ascii' |
|
25 | 25 | |
|
26 | 26 | def recode(s): |
|
27 | 27 | if isinstance(s, unicode): |
|
28 | 28 | return s.encode(orig_encoding, 'replace') |
|
29 | 29 | else: |
|
30 | 30 | return s.decode('utf-8').encode(orig_encoding, 'replace') |
|
31 | 31 | |
|
32 | 32 | source_converters = [ |
|
33 | 33 | ('cvs', convert_cvs, 'branchsort'), |
|
34 | 34 | ('git', convert_git, 'branchsort'), |
|
35 | 35 | ('svn', svn_source, 'branchsort'), |
|
36 | 36 | ('hg', mercurial_source, 'sourcesort'), |
|
37 | 37 | ('darcs', darcs_source, 'branchsort'), |
|
38 | 38 | ('mtn', monotone_source, 'branchsort'), |
|
39 | 39 | ('gnuarch', gnuarch_source, 'branchsort'), |
|
40 | 40 | ('bzr', bzr_source, 'branchsort'), |
|
41 | 41 | ('p4', p4_source, 'branchsort'), |
|
42 | 42 | ] |
|
43 | 43 | |
|
44 | 44 | sink_converters = [ |
|
45 | 45 | ('hg', mercurial_sink), |
|
46 | 46 | ('svn', svn_sink), |
|
47 | 47 | ] |
|
48 | 48 | |
|
49 | 49 | def convertsource(ui, path, type, rev): |
|
50 | 50 | exceptions = [] |
|
51 | 51 | if type and type not in [s[0] for s in source_converters]: |
|
52 | 52 | raise util.Abort(_('%s: invalid source repository type') % type) |
|
53 | 53 | for name, source, sortmode in source_converters: |
|
54 | 54 | try: |
|
55 | 55 | if not type or name == type: |
|
56 | 56 | return source(ui, path, rev), sortmode |
|
57 | 57 | except (NoRepo, MissingTool), inst: |
|
58 | 58 | exceptions.append(inst) |
|
59 | 59 | if not ui.quiet: |
|
60 | 60 | for inst in exceptions: |
|
61 | 61 | ui.write("%s\n" % inst) |
|
62 | 62 | raise util.Abort(_('%s: missing or unsupported repository') % path) |
|
63 | 63 | |
|
64 | 64 | def convertsink(ui, path, type): |
|
65 | 65 | if type and type not in [s[0] for s in sink_converters]: |
|
66 | 66 | raise util.Abort(_('%s: invalid destination repository type') % type) |
|
67 | 67 | for name, sink in sink_converters: |
|
68 | 68 | try: |
|
69 | 69 | if not type or name == type: |
|
70 | 70 | return sink(ui, path) |
|
71 | 71 | except NoRepo, inst: |
|
72 | 72 | ui.note(_("convert: %s\n") % inst) |
|
73 | 73 | raise util.Abort(_('%s: unknown repository type') % path) |
|
74 | 74 | |
|
75 | 75 | class progresssource(object): |
|
76 | 76 | def __init__(self, ui, source, filecount): |
|
77 | 77 | self.ui = ui |
|
78 | 78 | self.source = source |
|
79 | 79 | self.filecount = filecount |
|
80 | 80 | self.retrieved = 0 |
|
81 | 81 | |
|
82 | 82 | def getfile(self, file, rev): |
|
83 | 83 | self.retrieved += 1 |
|
84 | 84 | self.ui.progress(_('getting files'), self.retrieved, |
|
85 | 85 | item=file, total=self.filecount) |
|
86 | 86 | return self.source.getfile(file, rev) |
|
87 | 87 | |
|
88 | 88 | def lookuprev(self, rev): |
|
89 | 89 | return self.source.lookuprev(rev) |
|
90 | 90 | |
|
91 | 91 | def close(self): |
|
92 | 92 | self.ui.progress(_('getting files'), None) |
|
93 | 93 | |
|
94 | 94 | class converter(object): |
|
95 | 95 | def __init__(self, ui, source, dest, revmapfile, opts): |
|
96 | 96 | |
|
97 | 97 | self.source = source |
|
98 | 98 | self.dest = dest |
|
99 | 99 | self.ui = ui |
|
100 | 100 | self.opts = opts |
|
101 | 101 | self.commitcache = {} |
|
102 | 102 | self.authors = {} |
|
103 | 103 | self.authorfile = None |
|
104 | 104 | |
|
105 | 105 | # Record converted revisions persistently: maps source revision |
|
106 | 106 | # ID to target revision ID (both strings). (This is how |
|
107 | 107 | # incremental conversions work.) |
|
108 | 108 | self.map = mapfile(ui, revmapfile) |
|
109 | 109 | |
|
110 | 110 | # Read first the dst author map if any |
|
111 | 111 | authorfile = self.dest.authorfile() |
|
112 | 112 | if authorfile and os.path.exists(authorfile): |
|
113 | 113 | self.readauthormap(authorfile) |
|
114 | 114 | # Extend/Override with new author map if necessary |
|
115 |
if opts.get('author |
|
|
116 |
self.readauthormap(opts.get('author |
|
|
115 | if opts.get('authormap'): | |
|
116 | self.readauthormap(opts.get('authormap')) | |
|
117 | 117 | self.authorfile = self.dest.authorfile() |
|
118 | 118 | |
|
119 | 119 | self.splicemap = mapfile(ui, opts.get('splicemap')) |
|
120 | 120 | self.branchmap = mapfile(ui, opts.get('branchmap')) |
|
121 | 121 | |
|
122 | 122 | def walktree(self, heads): |
|
123 | 123 | '''Return a mapping that identifies the uncommitted parents of every |
|
124 | 124 | uncommitted changeset.''' |
|
125 | 125 | visit = heads |
|
126 | 126 | known = set() |
|
127 | 127 | parents = {} |
|
128 | 128 | while visit: |
|
129 | 129 | n = visit.pop(0) |
|
130 | 130 | if n in known or n in self.map: |
|
131 | 131 | continue |
|
132 | 132 | known.add(n) |
|
133 | 133 | self.ui.progress(_('scanning'), len(known), unit=_('revisions')) |
|
134 | 134 | commit = self.cachecommit(n) |
|
135 | 135 | parents[n] = [] |
|
136 | 136 | for p in commit.parents: |
|
137 | 137 | parents[n].append(p) |
|
138 | 138 | visit.append(p) |
|
139 | 139 | self.ui.progress(_('scanning'), None) |
|
140 | 140 | |
|
141 | 141 | return parents |
|
142 | 142 | |
|
143 | 143 | def toposort(self, parents, sortmode): |
|
144 | 144 | '''Return an ordering such that every uncommitted changeset is |
|
145 | 145 | preceeded by all its uncommitted ancestors.''' |
|
146 | 146 | |
|
147 | 147 | def mapchildren(parents): |
|
148 | 148 | """Return a (children, roots) tuple where 'children' maps parent |
|
149 | 149 | revision identifiers to children ones, and 'roots' is the list of |
|
150 | 150 | revisions without parents. 'parents' must be a mapping of revision |
|
151 | 151 | identifier to its parents ones. |
|
152 | 152 | """ |
|
153 | 153 | visit = parents.keys() |
|
154 | 154 | seen = set() |
|
155 | 155 | children = {} |
|
156 | 156 | roots = [] |
|
157 | 157 | |
|
158 | 158 | while visit: |
|
159 | 159 | n = visit.pop(0) |
|
160 | 160 | if n in seen: |
|
161 | 161 | continue |
|
162 | 162 | seen.add(n) |
|
163 | 163 | # Ensure that nodes without parents are present in the |
|
164 | 164 | # 'children' mapping. |
|
165 | 165 | children.setdefault(n, []) |
|
166 | 166 | hasparent = False |
|
167 | 167 | for p in parents[n]: |
|
168 | 168 | if not p in self.map: |
|
169 | 169 | visit.append(p) |
|
170 | 170 | hasparent = True |
|
171 | 171 | children.setdefault(p, []).append(n) |
|
172 | 172 | if not hasparent: |
|
173 | 173 | roots.append(n) |
|
174 | 174 | |
|
175 | 175 | return children, roots |
|
176 | 176 | |
|
177 | 177 | # Sort functions are supposed to take a list of revisions which |
|
178 | 178 | # can be converted immediately and pick one |
|
179 | 179 | |
|
180 | 180 | def makebranchsorter(): |
|
181 | 181 | """If the previously converted revision has a child in the |
|
182 | 182 | eligible revisions list, pick it. Return the list head |
|
183 | 183 | otherwise. Branch sort attempts to minimize branch |
|
184 | 184 | switching, which is harmful for Mercurial backend |
|
185 | 185 | compression. |
|
186 | 186 | """ |
|
187 | 187 | prev = [None] |
|
188 | 188 | def picknext(nodes): |
|
189 | 189 | next = nodes[0] |
|
190 | 190 | for n in nodes: |
|
191 | 191 | if prev[0] in parents[n]: |
|
192 | 192 | next = n |
|
193 | 193 | break |
|
194 | 194 | prev[0] = next |
|
195 | 195 | return next |
|
196 | 196 | return picknext |
|
197 | 197 | |
|
198 | 198 | def makesourcesorter(): |
|
199 | 199 | """Source specific sort.""" |
|
200 | 200 | keyfn = lambda n: self.commitcache[n].sortkey |
|
201 | 201 | def picknext(nodes): |
|
202 | 202 | return sorted(nodes, key=keyfn)[0] |
|
203 | 203 | return picknext |
|
204 | 204 | |
|
205 | 205 | def makedatesorter(): |
|
206 | 206 | """Sort revisions by date.""" |
|
207 | 207 | dates = {} |
|
208 | 208 | def getdate(n): |
|
209 | 209 | if n not in dates: |
|
210 | 210 | dates[n] = util.parsedate(self.commitcache[n].date) |
|
211 | 211 | return dates[n] |
|
212 | 212 | |
|
213 | 213 | def picknext(nodes): |
|
214 | 214 | return min([(getdate(n), n) for n in nodes])[1] |
|
215 | 215 | |
|
216 | 216 | return picknext |
|
217 | 217 | |
|
218 | 218 | if sortmode == 'branchsort': |
|
219 | 219 | picknext = makebranchsorter() |
|
220 | 220 | elif sortmode == 'datesort': |
|
221 | 221 | picknext = makedatesorter() |
|
222 | 222 | elif sortmode == 'sourcesort': |
|
223 | 223 | picknext = makesourcesorter() |
|
224 | 224 | else: |
|
225 | 225 | raise util.Abort(_('unknown sort mode: %s') % sortmode) |
|
226 | 226 | |
|
227 | 227 | children, actives = mapchildren(parents) |
|
228 | 228 | |
|
229 | 229 | s = [] |
|
230 | 230 | pendings = {} |
|
231 | 231 | while actives: |
|
232 | 232 | n = picknext(actives) |
|
233 | 233 | actives.remove(n) |
|
234 | 234 | s.append(n) |
|
235 | 235 | |
|
236 | 236 | # Update dependents list |
|
237 | 237 | for c in children.get(n, []): |
|
238 | 238 | if c not in pendings: |
|
239 | 239 | pendings[c] = [p for p in parents[c] if p not in self.map] |
|
240 | 240 | try: |
|
241 | 241 | pendings[c].remove(n) |
|
242 | 242 | except ValueError: |
|
243 | 243 | raise util.Abort(_('cycle detected between %s and %s') |
|
244 | 244 | % (recode(c), recode(n))) |
|
245 | 245 | if not pendings[c]: |
|
246 | 246 | # Parents are converted, node is eligible |
|
247 | 247 | actives.insert(0, c) |
|
248 | 248 | pendings[c] = None |
|
249 | 249 | |
|
250 | 250 | if len(s) != len(parents): |
|
251 | 251 | raise util.Abort(_("not all revisions were sorted")) |
|
252 | 252 | |
|
253 | 253 | return s |
|
254 | 254 | |
|
255 | 255 | def writeauthormap(self): |
|
256 | 256 | authorfile = self.authorfile |
|
257 | 257 | if authorfile: |
|
258 | 258 | self.ui.status(_('Writing author map file %s\n') % authorfile) |
|
259 | 259 | ofile = open(authorfile, 'w+') |
|
260 | 260 | for author in self.authors: |
|
261 | 261 | ofile.write("%s=%s\n" % (author, self.authors[author])) |
|
262 | 262 | ofile.close() |
|
263 | 263 | |
|
264 | 264 | def readauthormap(self, authorfile): |
|
265 | 265 | afile = open(authorfile, 'r') |
|
266 | 266 | for line in afile: |
|
267 | 267 | |
|
268 | 268 | line = line.strip() |
|
269 | 269 | if not line or line.startswith('#'): |
|
270 | 270 | continue |
|
271 | 271 | |
|
272 | 272 | try: |
|
273 | 273 | srcauthor, dstauthor = line.split('=', 1) |
|
274 | 274 | except ValueError: |
|
275 | 275 | msg = _('Ignoring bad line in author map file %s: %s\n') |
|
276 | 276 | self.ui.warn(msg % (authorfile, line.rstrip())) |
|
277 | 277 | continue |
|
278 | 278 | |
|
279 | 279 | srcauthor = srcauthor.strip() |
|
280 | 280 | dstauthor = dstauthor.strip() |
|
281 | 281 | if self.authors.get(srcauthor) in (None, dstauthor): |
|
282 | 282 | msg = _('mapping author %s to %s\n') |
|
283 | 283 | self.ui.debug(msg % (srcauthor, dstauthor)) |
|
284 | 284 | self.authors[srcauthor] = dstauthor |
|
285 | 285 | continue |
|
286 | 286 | |
|
287 | 287 | m = _('overriding mapping for author %s, was %s, will be %s\n') |
|
288 | 288 | self.ui.status(m % (srcauthor, self.authors[srcauthor], dstauthor)) |
|
289 | 289 | |
|
290 | 290 | afile.close() |
|
291 | 291 | |
|
292 | 292 | def cachecommit(self, rev): |
|
293 | 293 | commit = self.source.getcommit(rev) |
|
294 | 294 | commit.author = self.authors.get(commit.author, commit.author) |
|
295 | 295 | commit.branch = self.branchmap.get(commit.branch, commit.branch) |
|
296 | 296 | self.commitcache[rev] = commit |
|
297 | 297 | return commit |
|
298 | 298 | |
|
299 | 299 | def copy(self, rev): |
|
300 | 300 | commit = self.commitcache[rev] |
|
301 | 301 | |
|
302 | 302 | changes = self.source.getchanges(rev) |
|
303 | 303 | if isinstance(changes, basestring): |
|
304 | 304 | if changes == SKIPREV: |
|
305 | 305 | dest = SKIPREV |
|
306 | 306 | else: |
|
307 | 307 | dest = self.map[changes] |
|
308 | 308 | self.map[rev] = dest |
|
309 | 309 | return |
|
310 | 310 | files, copies = changes |
|
311 | 311 | pbranches = [] |
|
312 | 312 | if commit.parents: |
|
313 | 313 | for prev in commit.parents: |
|
314 | 314 | if prev not in self.commitcache: |
|
315 | 315 | self.cachecommit(prev) |
|
316 | 316 | pbranches.append((self.map[prev], |
|
317 | 317 | self.commitcache[prev].branch)) |
|
318 | 318 | self.dest.setbranch(commit.branch, pbranches) |
|
319 | 319 | try: |
|
320 | 320 | parents = self.splicemap[rev].replace(',', ' ').split() |
|
321 | 321 | self.ui.status(_('spliced in %s as parents of %s\n') % |
|
322 | 322 | (parents, rev)) |
|
323 | 323 | parents = [self.map.get(p, p) for p in parents] |
|
324 | 324 | except KeyError: |
|
325 | 325 | parents = [b[0] for b in pbranches] |
|
326 | 326 | source = progresssource(self.ui, self.source, len(files)) |
|
327 | 327 | newnode = self.dest.putcommit(files, copies, parents, commit, |
|
328 | 328 | source, self.map) |
|
329 | 329 | source.close() |
|
330 | 330 | self.source.converted(rev, newnode) |
|
331 | 331 | self.map[rev] = newnode |
|
332 | 332 | |
|
333 | 333 | def convert(self, sortmode): |
|
334 | 334 | try: |
|
335 | 335 | self.source.before() |
|
336 | 336 | self.dest.before() |
|
337 | 337 | self.source.setrevmap(self.map) |
|
338 | 338 | self.ui.status(_("scanning source...\n")) |
|
339 | 339 | heads = self.source.getheads() |
|
340 | 340 | parents = self.walktree(heads) |
|
341 | 341 | self.ui.status(_("sorting...\n")) |
|
342 | 342 | t = self.toposort(parents, sortmode) |
|
343 | 343 | num = len(t) |
|
344 | 344 | c = None |
|
345 | 345 | |
|
346 | 346 | self.ui.status(_("converting...\n")) |
|
347 | 347 | for i, c in enumerate(t): |
|
348 | 348 | num -= 1 |
|
349 | 349 | desc = self.commitcache[c].desc |
|
350 | 350 | if "\n" in desc: |
|
351 | 351 | desc = desc.splitlines()[0] |
|
352 | 352 | # convert log message to local encoding without using |
|
353 | 353 | # tolocal() because encoding.encoding conver() use it as |
|
354 | 354 | # 'utf-8' |
|
355 | 355 | self.ui.status("%d %s\n" % (num, recode(desc))) |
|
356 | 356 | self.ui.note(_("source: %s\n") % recode(c)) |
|
357 | 357 | self.ui.progress(_('converting'), i, unit=_('revisions'), |
|
358 | 358 | total=len(t)) |
|
359 | 359 | self.copy(c) |
|
360 | 360 | self.ui.progress(_('converting'), None) |
|
361 | 361 | |
|
362 | 362 | tags = self.source.gettags() |
|
363 | 363 | ctags = {} |
|
364 | 364 | for k in tags: |
|
365 | 365 | v = tags[k] |
|
366 | 366 | if self.map.get(v, SKIPREV) != SKIPREV: |
|
367 | 367 | ctags[k] = self.map[v] |
|
368 | 368 | |
|
369 | 369 | if c and ctags: |
|
370 | 370 | nrev, tagsparent = self.dest.puttags(ctags) |
|
371 | 371 | if nrev and tagsparent: |
|
372 | 372 | # write another hash correspondence to override the previous |
|
373 | 373 | # one so we don't end up with extra tag heads |
|
374 | 374 | tagsparents = [e for e in self.map.iteritems() |
|
375 | 375 | if e[1] == tagsparent] |
|
376 | 376 | if tagsparents: |
|
377 | 377 | self.map[tagsparents[0][0]] = nrev |
|
378 | 378 | |
|
379 | 379 | self.writeauthormap() |
|
380 | 380 | finally: |
|
381 | 381 | self.cleanup() |
|
382 | 382 | |
|
383 | 383 | def cleanup(self): |
|
384 | 384 | try: |
|
385 | 385 | self.dest.after() |
|
386 | 386 | finally: |
|
387 | 387 | self.source.after() |
|
388 | 388 | self.map.close() |
|
389 | 389 | |
|
390 | 390 | def convert(ui, src, dest=None, revmapfile=None, **opts): |
|
391 | 391 | global orig_encoding |
|
392 | 392 | orig_encoding = encoding.encoding |
|
393 | 393 | encoding.encoding = 'UTF-8' |
|
394 | 394 | |
|
395 | # support --authors as an alias for --authormap | |
|
396 | if not opts.get('authormap'): | |
|
397 | opts['authormap'] = opts.get('authors') | |
|
398 | ||
|
395 | 399 | if not dest: |
|
396 | 400 | dest = hg.defaultdest(src) + "-hg" |
|
397 | 401 | ui.status(_("assuming destination %s\n") % dest) |
|
398 | 402 | |
|
399 | 403 | destc = convertsink(ui, dest, opts.get('dest_type')) |
|
400 | 404 | |
|
401 | 405 | try: |
|
402 | 406 | srcc, defaultsort = convertsource(ui, src, opts.get('source_type'), |
|
403 | 407 | opts.get('rev')) |
|
404 | 408 | except Exception: |
|
405 | 409 | for path in destc.created: |
|
406 | 410 | shutil.rmtree(path, True) |
|
407 | 411 | raise |
|
408 | 412 | |
|
409 | 413 | sortmodes = ('branchsort', 'datesort', 'sourcesort') |
|
410 | 414 | sortmode = [m for m in sortmodes if opts.get(m)] |
|
411 | 415 | if len(sortmode) > 1: |
|
412 | 416 | raise util.Abort(_('more than one sort mode specified')) |
|
413 | 417 | sortmode = sortmode and sortmode[0] or defaultsort |
|
414 | 418 | if sortmode == 'sourcesort' and not srcc.hasnativeorder(): |
|
415 | 419 | raise util.Abort(_('--sourcesort is not supported by this data source')) |
|
416 | 420 | |
|
417 | 421 | fmap = opts.get('filemap') |
|
418 | 422 | if fmap: |
|
419 | 423 | srcc = filemap.filemap_source(ui, srcc, fmap) |
|
420 | 424 | destc.setfilemapmode(True) |
|
421 | 425 | |
|
422 | 426 | if not revmapfile: |
|
423 | 427 | try: |
|
424 | 428 | revmapfile = destc.revmapfile() |
|
425 | 429 | except: |
|
426 | 430 | revmapfile = os.path.join(destc, "map") |
|
427 | 431 | |
|
428 | 432 | c = converter(ui, srcc, destc, revmapfile, opts) |
|
429 | 433 | c.convert(sortmode) |
|
430 | 434 |
@@ -1,318 +1,318 b'' | |||
|
1 | 1 | hg convert [OPTION]... SOURCE [DEST [REVMAP]] |
|
2 | 2 | |
|
3 | 3 | convert a foreign SCM repository to a Mercurial one. |
|
4 | 4 | |
|
5 | 5 | Accepted source formats [identifiers]: |
|
6 | 6 | |
|
7 | 7 | - Mercurial [hg] |
|
8 | 8 | - CVS [cvs] |
|
9 | 9 | - Darcs [darcs] |
|
10 | 10 | - git [git] |
|
11 | 11 | - Subversion [svn] |
|
12 | 12 | - Monotone [mtn] |
|
13 | 13 | - GNU Arch [gnuarch] |
|
14 | 14 | - Bazaar [bzr] |
|
15 | 15 | - Perforce [p4] |
|
16 | 16 | |
|
17 | 17 | Accepted destination formats [identifiers]: |
|
18 | 18 | |
|
19 | 19 | - Mercurial [hg] |
|
20 | 20 | - Subversion [svn] (history on branches is not preserved) |
|
21 | 21 | |
|
22 | 22 | If no revision is given, all revisions will be converted. Otherwise, |
|
23 | 23 | convert will only import up to the named revision (given in a format |
|
24 | 24 | understood by the source). |
|
25 | 25 | |
|
26 | 26 | If no destination directory name is specified, it defaults to the basename |
|
27 | 27 | of the source with "-hg" appended. If the destination repository doesn't |
|
28 | 28 | exist, it will be created. |
|
29 | 29 | |
|
30 | 30 | By default, all sources except Mercurial will use --branchsort. Mercurial |
|
31 | 31 | uses --sourcesort to preserve original revision numbers order. Sort modes |
|
32 | 32 | have the following effects: |
|
33 | 33 | |
|
34 | 34 | --branchsort convert from parent to child revision when possible, which |
|
35 | 35 | means branches are usually converted one after the other. It |
|
36 | 36 | generates more compact repositories. |
|
37 | 37 | --datesort sort revisions by date. Converted repositories have good- |
|
38 | 38 | looking changelogs but are often an order of magnitude |
|
39 | 39 | larger than the same ones generated by --branchsort. |
|
40 | 40 | --sourcesort try to preserve source revisions order, only supported by |
|
41 | 41 | Mercurial sources. |
|
42 | 42 | |
|
43 | 43 | If <REVMAP> isn't given, it will be put in a default location |
|
44 | 44 | (<dest>/.hg/shamap by default). The <REVMAP> is a simple text file that |
|
45 | 45 | maps each source commit ID to the destination ID for that revision, like |
|
46 | 46 | so: |
|
47 | 47 | |
|
48 | 48 | <source ID> <destination ID> |
|
49 | 49 | |
|
50 | 50 | If the file doesn't exist, it's automatically created. It's updated on |
|
51 | 51 | each commit copied, so "hg convert" can be interrupted and can be run |
|
52 | 52 | repeatedly to copy new commits. |
|
53 | 53 | |
|
54 |
The |
|
|
55 |
|
|
|
56 |
|
|
|
57 |
|
|
|
54 | The authormap is a simple text file that maps each source commit author to | |
|
55 | a destination commit author. It is handy for source SCMs that use unix | |
|
56 | logins to identify authors (eg: CVS). One line per author mapping and the | |
|
57 | line format is: | |
|
58 | 58 | |
|
59 | 59 | source author = destination author |
|
60 | 60 | |
|
61 | 61 | Empty lines and lines starting with a "#" are ignored. |
|
62 | 62 | |
|
63 | 63 | The filemap is a file that allows filtering and remapping of files and |
|
64 | 64 | directories. Each line can contain one of the following directives: |
|
65 | 65 | |
|
66 | 66 | include path/to/file-or-dir |
|
67 | 67 | |
|
68 | 68 | exclude path/to/file-or-dir |
|
69 | 69 | |
|
70 | 70 | rename path/to/source path/to/destination |
|
71 | 71 | |
|
72 | 72 | Comment lines start with "#". A specified path matches if it equals the |
|
73 | 73 | full relative name of a file or one of its parent directories. The |
|
74 | 74 | "include" or "exclude" directive with the longest matching path applies, |
|
75 | 75 | so line order does not matter. |
|
76 | 76 | |
|
77 | 77 | The "include" directive causes a file, or all files under a directory, to |
|
78 | 78 | be included in the destination repository, and the exclusion of all other |
|
79 | 79 | files and directories not explicitly included. The "exclude" directive |
|
80 | 80 | causes files or directories to be omitted. The "rename" directive renames |
|
81 | 81 | a file or directory if it is converted. To rename from a subdirectory into |
|
82 | 82 | the root of the repository, use "." as the path to rename to. |
|
83 | 83 | |
|
84 | 84 | The splicemap is a file that allows insertion of synthetic history, |
|
85 | 85 | letting you specify the parents of a revision. This is useful if you want |
|
86 | 86 | to e.g. give a Subversion merge two parents, or graft two disconnected |
|
87 | 87 | series of history together. Each entry contains a key, followed by a |
|
88 | 88 | space, followed by one or two comma-separated values: |
|
89 | 89 | |
|
90 | 90 | key parent1, parent2 |
|
91 | 91 | |
|
92 | 92 | The key is the revision ID in the source revision control system whose |
|
93 | 93 | parents should be modified (same format as a key in .hg/shamap). The |
|
94 | 94 | values are the revision IDs (in either the source or destination revision |
|
95 | 95 | control system) that should be used as the new parents for that node. For |
|
96 | 96 | example, if you have merged "release-1.0" into "trunk", then you should |
|
97 | 97 | specify the revision on "trunk" as the first parent and the one on the |
|
98 | 98 | "release-1.0" branch as the second. |
|
99 | 99 | |
|
100 | 100 | The branchmap is a file that allows you to rename a branch when it is |
|
101 | 101 | being brought in from whatever external repository. When used in |
|
102 | 102 | conjunction with a splicemap, it allows for a powerful combination to help |
|
103 | 103 | fix even the most badly mismanaged repositories and turn them into nicely |
|
104 | 104 | structured Mercurial repositories. The branchmap contains lines of the |
|
105 | 105 | form: |
|
106 | 106 | |
|
107 | 107 | original_branch_name new_branch_name |
|
108 | 108 | |
|
109 | 109 | where "original_branch_name" is the name of the branch in the source |
|
110 | 110 | repository, and "new_branch_name" is the name of the branch is the |
|
111 | 111 | destination repository. No whitespace is allowed in the branch names. This |
|
112 | 112 | can be used to (for instance) move code in one repository from "default" |
|
113 | 113 | to a named branch. |
|
114 | 114 | |
|
115 | 115 | Mercurial Source |
|
116 | 116 | ---------------- |
|
117 | 117 | |
|
118 | 118 | --config convert.hg.ignoreerrors=False (boolean) |
|
119 | 119 | ignore integrity errors when reading. Use it to fix Mercurial |
|
120 | 120 | repositories with missing revlogs, by converting from and to |
|
121 | 121 | Mercurial. |
|
122 | 122 | |
|
123 | 123 | --config convert.hg.saverev=False (boolean) |
|
124 | 124 | store original revision ID in changeset (forces target IDs to change) |
|
125 | 125 | |
|
126 | 126 | --config convert.hg.startrev=0 (hg revision identifier) |
|
127 | 127 | convert start revision and its descendants |
|
128 | 128 | |
|
129 | 129 | CVS Source |
|
130 | 130 | ---------- |
|
131 | 131 | |
|
132 | 132 | CVS source will use a sandbox (i.e. a checked-out copy) from CVS to |
|
133 | 133 | indicate the starting point of what will be converted. Direct access to |
|
134 | 134 | the repository files is not needed, unless of course the repository is |
|
135 | 135 | :local:. The conversion uses the top level directory in the sandbox to |
|
136 | 136 | find the CVS repository, and then uses CVS rlog commands to find files to |
|
137 | 137 | convert. This means that unless a filemap is given, all files under the |
|
138 | 138 | starting directory will be converted, and that any directory |
|
139 | 139 | reorganization in the CVS sandbox is ignored. |
|
140 | 140 | |
|
141 | 141 | The options shown are the defaults. |
|
142 | 142 | |
|
143 | 143 | --config convert.cvsps.cache=True (boolean) |
|
144 | 144 | Set to False to disable remote log caching, for testing and debugging |
|
145 | 145 | purposes. |
|
146 | 146 | |
|
147 | 147 | --config convert.cvsps.fuzz=60 (integer) |
|
148 | 148 | Specify the maximum time (in seconds) that is allowed between commits |
|
149 | 149 | with identical user and log message in a single changeset. When very |
|
150 | 150 | large files were checked in as part of a changeset then the default |
|
151 | 151 | may not be long enough. |
|
152 | 152 | |
|
153 | 153 | --config convert.cvsps.mergeto='{{mergetobranch ([-\w]+)}}' |
|
154 | 154 | Specify a regular expression to which commit log messages are matched. |
|
155 | 155 | If a match occurs, then the conversion process will insert a dummy |
|
156 | 156 | revision merging the branch on which this log message occurs to the |
|
157 | 157 | branch indicated in the regex. |
|
158 | 158 | |
|
159 | 159 | --config convert.cvsps.mergefrom='{{mergefrombranch ([-\w]+)}}' |
|
160 | 160 | Specify a regular expression to which commit log messages are matched. |
|
161 | 161 | If a match occurs, then the conversion process will add the most |
|
162 | 162 | recent revision on the branch indicated in the regex as the second |
|
163 | 163 | parent of the changeset. |
|
164 | 164 | |
|
165 | 165 | --config hook.cvslog |
|
166 | 166 | Specify a Python function to be called at the end of gathering the CVS |
|
167 | 167 | log. The function is passed a list with the log entries, and can |
|
168 | 168 | modify the entries in-place, or add or delete them. |
|
169 | 169 | |
|
170 | 170 | --config hook.cvschangesets |
|
171 | 171 | Specify a Python function to be called after the changesets are |
|
172 | 172 | calculated from the the CVS log. The function is passed a list with |
|
173 | 173 | the changeset entries, and can modify the changesets in-place, or add |
|
174 | 174 | or delete them. |
|
175 | 175 | |
|
176 | 176 | An additional "debugcvsps" Mercurial command allows the builtin changeset |
|
177 | 177 | merging code to be run without doing a conversion. Its parameters and |
|
178 | 178 | output are similar to that of cvsps 2.1. Please see the command help for |
|
179 | 179 | more details. |
|
180 | 180 | |
|
181 | 181 | Subversion Source |
|
182 | 182 | ----------------- |
|
183 | 183 | |
|
184 | 184 | Subversion source detects classical trunk/branches/tags layouts. By |
|
185 | 185 | default, the supplied "svn://repo/path/" source URL is converted as a |
|
186 | 186 | single branch. If "svn://repo/path/trunk" exists it replaces the default |
|
187 | 187 | branch. If "svn://repo/path/branches" exists, its subdirectories are |
|
188 | 188 | listed as possible branches. If "svn://repo/path/tags" exists, it is |
|
189 | 189 | looked for tags referencing converted branches. Default "trunk", |
|
190 | 190 | "branches" and "tags" values can be overridden with following options. Set |
|
191 | 191 | them to paths relative to the source URL, or leave them blank to disable |
|
192 | 192 | auto detection. |
|
193 | 193 | |
|
194 | 194 | --config convert.svn.branches=branches (directory name) |
|
195 | 195 | specify the directory containing branches |
|
196 | 196 | |
|
197 | 197 | --config convert.svn.tags=tags (directory name) |
|
198 | 198 | specify the directory containing tags |
|
199 | 199 | |
|
200 | 200 | --config convert.svn.trunk=trunk (directory name) |
|
201 | 201 | specify the name of the trunk branch |
|
202 | 202 | |
|
203 | 203 | Source history can be retrieved starting at a specific revision, instead |
|
204 | 204 | of being integrally converted. Only single branch conversions are |
|
205 | 205 | supported. |
|
206 | 206 | |
|
207 | 207 | --config convert.svn.startrev=0 (svn revision number) |
|
208 | 208 | specify start Subversion revision. |
|
209 | 209 | |
|
210 | 210 | Perforce Source |
|
211 | 211 | --------------- |
|
212 | 212 | |
|
213 | 213 | The Perforce (P4) importer can be given a p4 depot path or a client |
|
214 | 214 | specification as source. It will convert all files in the source to a flat |
|
215 | 215 | Mercurial repository, ignoring labels, branches and integrations. Note |
|
216 | 216 | that when a depot path is given you then usually should specify a target |
|
217 | 217 | directory, because otherwise the target may be named ...-hg. |
|
218 | 218 | |
|
219 | 219 | It is possible to limit the amount of source history to be converted by |
|
220 | 220 | specifying an initial Perforce revision. |
|
221 | 221 | |
|
222 | 222 | --config convert.p4.startrev=0 (perforce changelist number) |
|
223 | 223 | specify initial Perforce revision. |
|
224 | 224 | |
|
225 | 225 | Mercurial Destination |
|
226 | 226 | --------------------- |
|
227 | 227 | |
|
228 | 228 | --config convert.hg.clonebranches=False (boolean) |
|
229 | 229 | dispatch source branches in separate clones. |
|
230 | 230 | |
|
231 | 231 | --config convert.hg.tagsbranch=default (branch name) |
|
232 | 232 | tag revisions branch name |
|
233 | 233 | |
|
234 | 234 | --config convert.hg.usebranchnames=True (boolean) |
|
235 | 235 | preserve branch names |
|
236 | 236 | |
|
237 | 237 | options: |
|
238 | 238 | |
|
239 | -A --authors FILE username mapping filename | |
|
240 | 239 | -s --source-type TYPE source repository type |
|
241 | 240 | -d --dest-type TYPE destination repository type |
|
242 | 241 | -r --rev REV import up to target revision REV |
|
242 | --authormap FILE remap usernames using this file | |
|
243 | 243 | --filemap FILE remap file names using contents of file |
|
244 | 244 | --splicemap FILE splice synthesized history into place |
|
245 | 245 | --branchmap FILE change branch names while converting |
|
246 | 246 | --branchsort try to sort changesets by branches |
|
247 | 247 | --datesort try to sort changesets by date |
|
248 | 248 | --sourcesort preserve source changesets order |
|
249 | 249 | |
|
250 | 250 | use "hg -v help convert" to show global options |
|
251 | 251 | adding a |
|
252 | 252 | assuming destination a-hg |
|
253 | 253 | initializing destination a-hg repository |
|
254 | 254 | scanning source... |
|
255 | 255 | sorting... |
|
256 | 256 | converting... |
|
257 | 257 | 4 a |
|
258 | 258 | 3 b |
|
259 | 259 | 2 c |
|
260 | 260 | 1 d |
|
261 | 261 | 0 e |
|
262 | 262 | pulling from ../a |
|
263 | 263 | searching for changes |
|
264 | 264 | no changes found |
|
265 | 265 | % should fail |
|
266 | 266 | initializing destination bogusfile repository |
|
267 | 267 | abort: cannot create new bundle repository |
|
268 | 268 | % should fail |
|
269 | 269 | abort: Permission denied: bogusdir |
|
270 | 270 | % should succeed |
|
271 | 271 | initializing destination bogusdir repository |
|
272 | 272 | scanning source... |
|
273 | 273 | sorting... |
|
274 | 274 | converting... |
|
275 | 275 | 4 a |
|
276 | 276 | 3 b |
|
277 | 277 | 2 c |
|
278 | 278 | 1 d |
|
279 | 279 | 0 e |
|
280 | 280 | % test pre and post conversion actions |
|
281 | 281 | run hg source pre-conversion action |
|
282 | 282 | run hg sink pre-conversion action |
|
283 | 283 | run hg sink post-conversion action |
|
284 | 284 | run hg source post-conversion action |
|
285 | 285 | % converting empty dir should fail nicely |
|
286 | 286 | assuming destination emptydir-hg |
|
287 | 287 | initializing destination emptydir-hg repository |
|
288 | 288 | emptydir does not look like a CVS checkout |
|
289 | 289 | emptydir does not look like a Git repository |
|
290 | 290 | emptydir does not look like a Subversion repository |
|
291 | 291 | emptydir is not a local Mercurial repository |
|
292 | 292 | emptydir does not look like a darcs repository |
|
293 | 293 | emptydir does not look like a monotone repository |
|
294 | 294 | emptydir does not look like a GNU Arch repository |
|
295 | 295 | emptydir does not look like a Bazaar repository |
|
296 | 296 | cannot find required "p4" tool |
|
297 | 297 | abort: emptydir: missing or unsupported repository |
|
298 | 298 | % convert with imaginary source type |
|
299 | 299 | initializing destination a-foo repository |
|
300 | 300 | abort: foo: invalid source repository type |
|
301 | 301 | % convert with imaginary sink type |
|
302 | 302 | abort: foo: invalid destination repository type |
|
303 | 303 | |
|
304 | 304 | % testing: convert must not produce duplicate entries in fncache |
|
305 | 305 | initializing destination b repository |
|
306 | 306 | scanning source... |
|
307 | 307 | sorting... |
|
308 | 308 | converting... |
|
309 | 309 | 4 a |
|
310 | 310 | 3 b |
|
311 | 311 | 2 c |
|
312 | 312 | 1 d |
|
313 | 313 | 0 e |
|
314 | 314 | % contents of fncache file: |
|
315 | 315 | data/a.i |
|
316 | 316 | data/b.i |
|
317 | 317 | % test bogus URL |
|
318 | 318 | abort: bzr+ssh://foobar@selenic.com/baz: missing or unsupported repository |
General Comments 0
You need to be logged in to leave comments.
Login now