6 git-fast-export - Git data exporter
12 'git fast-export' [<options>] | 'git fast-import'
16 This program dumps the given revisions in a form suitable to be piped
17 into 'git fast-import'.
19 You can use it as a human-readable bundle replacement (see
20 linkgit:git-bundle[1]), or as a format that can be edited before being
21 fed to 'git fast-import' in order to do history rewrites (an ability
22 relied on by tools like 'git filter-repo').
27 Insert 'progress' statements every <n> objects, to be shown by
28 'git fast-import' during import.
30 --signed-tags=(verbatim|warn|warn-strip|strip|abort)::
31 Specify how to handle signed tags. Since any transformation
32 after the export can change the tag names (which can also happen
33 when excluding revisions) the signatures will not match.
35 When asking to 'abort' (which is the default), this program will die
36 when encountering a signed tag. With 'strip', the tags will silently
37 be made unsigned, with 'warn-strip' they will be made unsigned but a
38 warning will be displayed, with 'verbatim', they will be silently
39 exported and with 'warn', they will be exported, but you will see a
42 --tag-of-filtered-object=(abort|drop|rewrite)::
43 Specify how to handle tags whose tagged object is filtered out.
44 Since revisions and files to export can be limited by path,
45 tagged objects may be filtered completely.
47 When asking to 'abort' (which is the default), this program will die
48 when encountering such a tag. With 'drop' it will omit such tags from
49 the output. With 'rewrite', if the tagged object is a commit, it will
50 rewrite the tag to tag an ancestor commit (via parent rewriting; see
51 linkgit:git-rev-list[1]).
55 Perform move and/or copy detection, as described in the
56 linkgit:git-diff[1] manual page, and use it to generate
57 rename and copy commands in the output dump.
59 Note that earlier versions of this command did not complain and
60 produced incorrect results if you gave these options.
62 --export-marks=<file>::
63 Dumps the internal marks table to <file> when complete.
64 Marks are written one per line as `:markid SHA-1`. Only marks
65 for revisions are dumped; marks for blobs are ignored.
66 Backends can use this file to validate imports after they
67 have been completed, or to save the marks table across
68 incremental runs. As <file> is only opened and truncated
69 at completion, the same path can also be safely given to
71 The file will not be written if no new object has been
74 --import-marks=<file>::
75 Before processing any input, load the marks specified in
76 <file>. The input file must exist, must be readable, and
77 must use the same format as produced by --export-marks.
80 In addition to labelling blobs and commits with mark ids, also
81 label tags. This is useful in conjunction with
82 `--export-marks` and `--import-marks`, and is also useful (and
83 necessary) for exporting of nested tags. It does not hurt
84 other cases and would be the default, but many fast-import
85 frontends are not prepared to accept tags with mark
88 Any commits (or tags) that have already been marked will not be
89 exported again. If the backend uses a similar --import-marks file,
90 this allows for incremental bidirectional exporting of the repository
91 by keeping the marks the same across runs.
93 --fake-missing-tagger::
94 Some old repositories have tags without a tagger. The
95 fast-import protocol was pretty strict about that, and did not
96 allow that. So fake a tagger to be able to fast-import the
100 Start the stream with a 'feature done' stanza, and terminate
101 it with a 'done' command.
104 Skip output of blob objects and instead refer to blobs via
105 their original SHA-1 hash. This is useful when rewriting the
106 directory structure or history of a repository without
107 touching the contents of individual files. Note that the
108 resulting stream can only be used by a repository which
109 already contains the necessary objects.
112 This option will cause fast-export to issue a "deleteall"
113 directive for each commit followed by a full list of all files
114 in the commit (as opposed to just listing the files which are
115 different from the commit's first parent).
118 Anonymize the contents of the repository while still retaining
119 the shape of the history and stored tree. See the section on
122 --anonymize-map=<from>[:<to>]::
123 Convert token `<from>` to `<to>` in the anonymized output. If
124 `<to>` is omitted, map `<from>` to itself (i.e., do not
125 anonymize it). See the section on `ANONYMIZING` below.
127 --reference-excluded-parents::
128 By default, running a command such as `git fast-export
129 master~5..master` will not include the commit master{tilde}5
130 and will make master{tilde}4 no longer have master{tilde}5 as
131 a parent (though both the old master{tilde}4 and new
132 master{tilde}4 will have all the same files). Use
133 --reference-excluded-parents to instead have the stream
134 refer to commits in the excluded range of history by their
135 sha1sum. Note that the resulting stream can only be used by a
136 repository which already contains the necessary parent
139 --show-original-ids::
140 Add an extra directive to the output for commits and blobs,
141 `original-oid <SHA1SUM>`. While such directives will likely be
142 ignored by importers such as git-fast-import, it may be useful
143 for intermediary filters (e.g. for rewriting commit messages
144 which refer to older commits, or for stripping blobs by id).
146 --reencode=(yes|no|abort)::
147 Specify how to handle `encoding` header in commit objects. When
148 asking to 'abort' (which is the default), this program will die
149 when encountering such a commit object. With 'yes', the commit
150 message will be re-encoded into UTF-8. With 'no', the original
151 encoding will be preserved.
154 Apply the specified refspec to each ref exported. Multiple of them can
157 [<git-rev-list-args>...]::
158 A list of arguments, acceptable to 'git rev-parse' and
159 'git rev-list', that specifies the specific objects and references
160 to export. For example, `master~10..master` causes the
161 current master reference to be exported along with all objects
162 added since its 10th ancestor commit and (unless the
163 --reference-excluded-parents option is specified) all files
164 common to master{tilde}9 and master{tilde}10.
169 -------------------------------------------------------------------
170 $ git fast-export --all | (cd /empty/repository && git fast-import)
171 -------------------------------------------------------------------
173 This will export the whole repository and import it into the existing
174 empty repository. Except for reencoding commits that are not in
175 UTF-8, it would be a one-to-one mirror.
177 -----------------------------------------------------
178 $ git fast-export master~5..master |
179 sed "s|refs/heads/master|refs/heads/other|" |
181 -----------------------------------------------------
183 This makes a new branch called 'other' from 'master~5..master'
184 (i.e. if 'master' has linear history, it will take the last 5 commits).
186 Note that this assumes that none of the blobs and commit messages
187 referenced by that revision range contains the string
194 If the `--anonymize` option is given, git will attempt to remove all
195 identifying information from the repository while still retaining enough
196 of the original tree and history patterns to reproduce some bugs. The
197 goal is that a git bug which is found on a private repository will
198 persist in the anonymized repository, and the latter can be shared with
199 git developers to help solve the bug.
201 With this option, git will replace all refnames, paths, blob contents,
202 commit and tag messages, names, and email addresses in the output with
203 anonymized data. Two instances of the same string will be replaced
204 equivalently (e.g., two commits with the same author will have the same
205 anonymized author in the output, but bear no resemblance to the original
206 author string). The relationship between commits, branches, and tags is
207 retained, as well as the commit timestamps (but the commit messages and
208 refnames bear no resemblance to the originals). The relative makeup of
209 the tree is retained (e.g., if you have a root tree with 10 files and 3
210 trees, so will the output), but their names and the contents of the
211 files will be replaced.
213 If you think you have found a git bug, you can start by exporting an
214 anonymized stream of the whole repository:
216 ---------------------------------------------------
217 $ git fast-export --anonymize --all >anon-stream
218 ---------------------------------------------------
220 Then confirm that the bug persists in a repository created from that
221 stream (many bugs will not, as they really do depend on the exact
222 repository contents):
224 ---------------------------------------------------
227 $ git fast-import <../anon-stream
228 $ ... test your bug ...
229 ---------------------------------------------------
231 If the anonymized repository shows the bug, it may be worth sharing
232 `anon-stream` along with a regular bug report. Note that the anonymized
233 stream compresses very well, so gzipping it is encouraged. If you want
234 to examine the stream to see that it does not contain any private data,
235 you can peruse it directly before sending. You may also want to try:
237 ---------------------------------------------------
238 $ perl -pe 's/\d+/X/g' <anon-stream | sort -u | less
239 ---------------------------------------------------
241 which shows all of the unique lines (with numbers converted to "X", to
242 collapse "User 0", "User 1", etc into "User X"). This produces a much
243 smaller output, and it is usually easy to quickly confirm that there is
244 no private data in the stream.
246 Reproducing some bugs may require referencing particular commits or
247 paths, which becomes challenging after refnames and paths have been
248 anonymized. You can ask for a particular token to be left as-is or
249 mapped to a new value. For example, if you have a bug which reproduces
250 with `git rev-list sensitive -- secret.c`, you can run:
252 ---------------------------------------------------
253 $ git fast-export --anonymize --all \
254 --anonymize-map=sensitive:foo \
255 --anonymize-map=secret.c:bar.c \
257 ---------------------------------------------------
259 After importing the stream, you can then run `git rev-list foo -- bar.c`
260 in the anonymized repository.
262 Note that paths and refnames are split into tokens at slash boundaries.
263 The command above would anonymize `subdir/secret.c` as something like
264 `path123/bar.c`; you could then search for `bar.c` in the anonymized
265 repository to determine the final pathname.
267 To make referencing the final pathname simpler, you can map each path
268 component; so if you also anonymize `subdir` to `publicdir`, then the
269 final pathname would be `publicdir/bar.c`.
274 Since 'git fast-import' cannot tag trees, you will not be
275 able to export the linux.git repository completely, as it contains
276 a tag referencing a tree instead of a commit.
280 linkgit:git-fast-import[1]
284 Part of the linkgit:git[1] suite