Merge branch 'ps/t1509-chroot-test-fixup'
[git/git.git] / Documentation / git-fast-import.txt
1 git-fast-import(1)
2 ==================
3
4 NAME
5 ----
6 git-fast-import - Backend for fast Git data importers
7
8
9 SYNOPSIS
10 --------
11 [verse]
12 frontend | 'git fast-import' [options]
13
14 DESCRIPTION
15 -----------
16 This program is usually not what the end user wants to run directly.
17 Most end users want to use one of the existing frontend programs,
18 which parses a specific type of foreign source and feeds the contents
19 stored there to 'git fast-import'.
20
21 fast-import reads a mixed command/data stream from standard input and
22 writes one or more packfiles directly into the current repository.
23 When EOF is received on standard input, fast import writes out
24 updated branch and tag refs, fully updating the current repository
25 with the newly imported data.
26
27 The fast-import backend itself can import into an empty repository (one that
28 has already been initialized by 'git init') or incrementally
29 update an existing populated repository. Whether or not incremental
30 imports are supported from a particular foreign source depends on
31 the frontend program in use.
32
33
34 OPTIONS
35 -------
36
37 --force::
38 Force updating modified existing branches, even if doing
39 so would cause commits to be lost (as the new commit does
40 not contain the old commit).
41
42 --quiet::
43 Disable all non-fatal output, making fast-import silent when it
44 is successful. This option disables the output shown by
45 --stats.
46
47 --stats::
48 Display some basic statistics about the objects fast-import has
49 created, the packfiles they were stored into, and the
50 memory used by fast-import during this run. Showing this output
51 is currently the default, but can be disabled with --quiet.
52
53 Options for Frontends
54 ~~~~~~~~~~~~~~~~~~~~~
55
56 --cat-blob-fd=<fd>::
57 Write responses to `get-mark`, `cat-blob`, and `ls` queries to the
58 file descriptor <fd> instead of `stdout`. Allows `progress`
59 output intended for the end-user to be separated from other
60 output.
61
62 --date-format=<fmt>::
63 Specify the type of dates the frontend will supply to
64 fast-import within `author`, `committer` and `tagger` commands.
65 See ``Date Formats'' below for details about which formats
66 are supported, and their syntax.
67
68 --done::
69 Terminate with error if there is no `done` command at the end of
70 the stream. This option might be useful for detecting errors
71 that cause the frontend to terminate before it has started to
72 write a stream.
73
74 Locations of Marks Files
75 ~~~~~~~~~~~~~~~~~~~~~~~~
76
77 --export-marks=<file>::
78 Dumps the internal marks table to <file> when complete.
79 Marks are written one per line as `:markid SHA-1`.
80 Frontends can use this file to validate imports after they
81 have been completed, or to save the marks table across
82 incremental runs. As <file> is only opened and truncated
83 at checkpoint (or completion) the same path can also be
84 safely given to --import-marks.
85
86 --import-marks=<file>::
87 Before processing any input, load the marks specified in
88 <file>. The input file must exist, must be readable, and
89 must use the same format as produced by --export-marks.
90 Multiple options may be supplied to import more than one
91 set of marks. If a mark is defined to different values,
92 the last file wins.
93
94 --import-marks-if-exists=<file>::
95 Like --import-marks but instead of erroring out, silently
96 skips the file if it does not exist.
97
98 --[no-]relative-marks::
99 After specifying --relative-marks the paths specified
100 with --import-marks= and --export-marks= are relative
101 to an internal directory in the current repository.
102 In git-fast-import this means that the paths are relative
103 to the .git/info/fast-import directory. However, other
104 importers may use a different location.
105 +
106 Relative and non-relative marks may be combined by interweaving
107 --(no-)-relative-marks with the --(import|export)-marks= options.
108
109 Performance and Compression Tuning
110 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
111
112 --active-branches=<n>::
113 Maximum number of branches to maintain active at once.
114 See ``Memory Utilization'' below for details. Default is 5.
115
116 --big-file-threshold=<n>::
117 Maximum size of a blob that fast-import will attempt to
118 create a delta for, expressed in bytes. The default is 512m
119 (512 MiB). Some importers may wish to lower this on systems
120 with constrained memory.
121
122 --depth=<n>::
123 Maximum delta depth, for blob and tree deltification.
124 Default is 10.
125
126 --export-pack-edges=<file>::
127 After creating a packfile, print a line of data to
128 <file> listing the filename of the packfile and the last
129 commit on each branch that was written to that packfile.
130 This information may be useful after importing projects
131 whose total object set exceeds the 4 GiB packfile limit,
132 as these commits can be used as edge points during calls
133 to 'git pack-objects'.
134
135 --max-pack-size=<n>::
136 Maximum size of each output packfile.
137 The default is unlimited.
138
139
140 Performance
141 -----------
142 The design of fast-import allows it to import large projects in a minimum
143 amount of memory usage and processing time. Assuming the frontend
144 is able to keep up with fast-import and feed it a constant stream of data,
145 import times for projects holding 10+ years of history and containing
146 100,000+ individual commits are generally completed in just 1-2
147 hours on quite modest (~$2,000 USD) hardware.
148
149 Most bottlenecks appear to be in foreign source data access (the
150 source just cannot extract revisions fast enough) or disk IO (fast-import
151 writes as fast as the disk will take the data). Imports will run
152 faster if the source data is stored on a different drive than the
153 destination Git repository (due to less IO contention).
154
155
156 Development Cost
157 ----------------
158 A typical frontend for fast-import tends to weigh in at approximately 200
159 lines of Perl/Python/Ruby code. Most developers have been able to
160 create working importers in just a couple of hours, even though it
161 is their first exposure to fast-import, and sometimes even to Git. This is
162 an ideal situation, given that most conversion tools are throw-away
163 (use once, and never look back).
164
165
166 Parallel Operation
167 ------------------
168 Like 'git push' or 'git fetch', imports handled by fast-import are safe to
169 run alongside parallel `git repack -a -d` or `git gc` invocations,
170 or any other Git operation (including 'git prune', as loose objects
171 are never used by fast-import).
172
173 fast-import does not lock the branch or tag refs it is actively importing.
174 After the import, during its ref update phase, fast-import tests each
175 existing branch ref to verify the update will be a fast-forward
176 update (the commit stored in the ref is contained in the new
177 history of the commit to be written). If the update is not a
178 fast-forward update, fast-import will skip updating that ref and instead
179 prints a warning message. fast-import will always attempt to update all
180 branch refs, and does not stop on the first failure.
181
182 Branch updates can be forced with --force, but it's recommended that
183 this only be used on an otherwise quiet repository. Using --force
184 is not necessary for an initial import into an empty repository.
185
186
187 Technical Discussion
188 --------------------
189 fast-import tracks a set of branches in memory. Any branch can be created
190 or modified at any point during the import process by sending a
191 `commit` command on the input stream. This design allows a frontend
192 program to process an unlimited number of branches simultaneously,
193 generating commits in the order they are available from the source
194 data. It also simplifies the frontend programs considerably.
195
196 fast-import does not use or alter the current working directory, or any
197 file within it. (It does however update the current Git repository,
198 as referenced by `GIT_DIR`.) Therefore an import frontend may use
199 the working directory for its own purposes, such as extracting file
200 revisions from the foreign source. This ignorance of the working
201 directory also allows fast-import to run very quickly, as it does not
202 need to perform any costly file update operations when switching
203 between branches.
204
205 Input Format
206 ------------
207 With the exception of raw file data (which Git does not interpret)
208 the fast-import input format is text (ASCII) based. This text based
209 format simplifies development and debugging of frontend programs,
210 especially when a higher level language such as Perl, Python or
211 Ruby is being used.
212
213 fast-import is very strict about its input. Where we say SP below we mean
214 *exactly* one space. Likewise LF means one (and only one) linefeed
215 and HT one (and only one) horizontal tab.
216 Supplying additional whitespace characters will cause unexpected
217 results, such as branch names or file names with leading or trailing
218 spaces in their name, or early termination of fast-import when it encounters
219 unexpected input.
220
221 Stream Comments
222 ~~~~~~~~~~~~~~~
223 To aid in debugging frontends fast-import ignores any line that
224 begins with `#` (ASCII pound/hash) up to and including the line
225 ending `LF`. A comment line may contain any sequence of bytes
226 that does not contain an LF and therefore may be used to include
227 any detailed debugging information that might be specific to the
228 frontend and useful when inspecting a fast-import data stream.
229
230 Date Formats
231 ~~~~~~~~~~~~
232 The following date formats are supported. A frontend should select
233 the format it will use for this import by passing the format name
234 in the --date-format=<fmt> command-line option.
235
236 `raw`::
237 This is the Git native format and is `<time> SP <offutc>`.
238 It is also fast-import's default format, if --date-format was
239 not specified.
240 +
241 The time of the event is specified by `<time>` as the number of
242 seconds since the UNIX epoch (midnight, Jan 1, 1970, UTC) and is
243 written as an ASCII decimal integer.
244 +
245 The local offset is specified by `<offutc>` as a positive or negative
246 offset from UTC. For example EST (which is 5 hours behind UTC)
247 would be expressed in `<tz>` by ``-0500'' while UTC is ``+0000''.
248 The local offset does not affect `<time>`; it is used only as an
249 advisement to help formatting routines display the timestamp.
250 +
251 If the local offset is not available in the source material, use
252 ``+0000'', or the most common local offset. For example many
253 organizations have a CVS repository which has only ever been accessed
254 by users who are located in the same location and time zone. In this
255 case a reasonable offset from UTC could be assumed.
256 +
257 Unlike the `rfc2822` format, this format is very strict. Any
258 variation in formatting will cause fast-import to reject the value.
259
260 `rfc2822`::
261 This is the standard email format as described by RFC 2822.
262 +
263 An example value is ``Tue Feb 6 11:22:18 2007 -0500''. The Git
264 parser is accurate, but a little on the lenient side. It is the
265 same parser used by 'git am' when applying patches
266 received from email.
267 +
268 Some malformed strings may be accepted as valid dates. In some of
269 these cases Git will still be able to obtain the correct date from
270 the malformed string. There are also some types of malformed
271 strings which Git will parse wrong, and yet consider valid.
272 Seriously malformed strings will be rejected.
273 +
274 Unlike the `raw` format above, the time zone/UTC offset information
275 contained in an RFC 2822 date string is used to adjust the date
276 value to UTC prior to storage. Therefore it is important that
277 this information be as accurate as possible.
278 +
279 If the source material uses RFC 2822 style dates,
280 the frontend should let fast-import handle the parsing and conversion
281 (rather than attempting to do it itself) as the Git parser has
282 been well tested in the wild.
283 +
284 Frontends should prefer the `raw` format if the source material
285 already uses UNIX-epoch format, can be coaxed to give dates in that
286 format, or its format is easily convertible to it, as there is no
287 ambiguity in parsing.
288
289 `now`::
290 Always use the current time and time zone. The literal
291 `now` must always be supplied for `<when>`.
292 +
293 This is a toy format. The current time and time zone of this system
294 is always copied into the identity string at the time it is being
295 created by fast-import. There is no way to specify a different time or
296 time zone.
297 +
298 This particular format is supplied as it's short to implement and
299 may be useful to a process that wants to create a new commit
300 right now, without needing to use a working directory or
301 'git update-index'.
302 +
303 If separate `author` and `committer` commands are used in a `commit`
304 the timestamps may not match, as the system clock will be polled
305 twice (once for each command). The only way to ensure that both
306 author and committer identity information has the same timestamp
307 is to omit `author` (thus copying from `committer`) or to use a
308 date format other than `now`.
309
310 Commands
311 ~~~~~~~~
312 fast-import accepts several commands to update the current repository
313 and control the current import process. More detailed discussion
314 (with examples) of each command follows later.
315
316 `commit`::
317 Creates a new branch or updates an existing branch by
318 creating a new commit and updating the branch to point at
319 the newly created commit.
320
321 `tag`::
322 Creates an annotated tag object from an existing commit or
323 branch. Lightweight tags are not supported by this command,
324 as they are not recommended for recording meaningful points
325 in time.
326
327 `reset`::
328 Reset an existing branch (or a new branch) to a specific
329 revision. This command must be used to change a branch to
330 a specific revision without making a commit on it.
331
332 `blob`::
333 Convert raw file data into a blob, for future use in a
334 `commit` command. This command is optional and is not
335 needed to perform an import.
336
337 `checkpoint`::
338 Forces fast-import to close the current packfile, generate its
339 unique SHA-1 checksum and index, and start a new packfile.
340 This command is optional and is not needed to perform
341 an import.
342
343 `progress`::
344 Causes fast-import to echo the entire line to its own
345 standard output. This command is optional and is not needed
346 to perform an import.
347
348 `done`::
349 Marks the end of the stream. This command is optional
350 unless the `done` feature was requested using the
351 `--done` command-line option or `feature done` command.
352
353 `get-mark`::
354 Causes fast-import to print the SHA-1 corresponding to a mark
355 to the file descriptor set with `--cat-blob-fd`, or `stdout` if
356 unspecified.
357
358 `cat-blob`::
359 Causes fast-import to print a blob in 'cat-file --batch'
360 format to the file descriptor set with `--cat-blob-fd` or
361 `stdout` if unspecified.
362
363 `ls`::
364 Causes fast-import to print a line describing a directory
365 entry in 'ls-tree' format to the file descriptor set with
366 `--cat-blob-fd` or `stdout` if unspecified.
367
368 `feature`::
369 Enable the specified feature. This requires that fast-import
370 supports the specified feature, and aborts if it does not.
371
372 `option`::
373 Specify any of the options listed under OPTIONS that do not
374 change stream semantic to suit the frontend's needs. This
375 command is optional and is not needed to perform an import.
376
377 `commit`
378 ~~~~~~~~
379 Create or update a branch with a new commit, recording one logical
380 change to the project.
381
382 ....
383 'commit' SP <ref> LF
384 mark?
385 ('author' (SP <name>)? SP LT <email> GT SP <when> LF)?
386 'committer' (SP <name>)? SP LT <email> GT SP <when> LF
387 data
388 ('from' SP <commit-ish> LF)?
389 ('merge' SP <commit-ish> LF)?
390 (filemodify | filedelete | filecopy | filerename | filedeleteall | notemodify)*
391 LF?
392 ....
393
394 where `<ref>` is the name of the branch to make the commit on.
395 Typically branch names are prefixed with `refs/heads/` in
396 Git, so importing the CVS branch symbol `RELENG-1_0` would use
397 `refs/heads/RELENG-1_0` for the value of `<ref>`. The value of
398 `<ref>` must be a valid refname in Git. As `LF` is not valid in
399 a Git refname, no quoting or escaping syntax is supported here.
400
401 A `mark` command may optionally appear, requesting fast-import to save a
402 reference to the newly created commit for future use by the frontend
403 (see below for format). It is very common for frontends to mark
404 every commit they create, thereby allowing future branch creation
405 from any imported commit.
406
407 The `data` command following `committer` must supply the commit
408 message (see below for `data` command syntax). To import an empty
409 commit message use a 0 length data. Commit messages are free-form
410 and are not interpreted by Git. Currently they must be encoded in
411 UTF-8, as fast-import does not permit other encodings to be specified.
412
413 Zero or more `filemodify`, `filedelete`, `filecopy`, `filerename`,
414 `filedeleteall` and `notemodify` commands
415 may be included to update the contents of the branch prior to
416 creating the commit. These commands may be supplied in any order.
417 However it is recommended that a `filedeleteall` command precede
418 all `filemodify`, `filecopy`, `filerename` and `notemodify` commands in
419 the same commit, as `filedeleteall` wipes the branch clean (see below).
420
421 The `LF` after the command is optional (it used to be required).
422
423 `author`
424 ^^^^^^^^
425 An `author` command may optionally appear, if the author information
426 might differ from the committer information. If `author` is omitted
427 then fast-import will automatically use the committer's information for
428 the author portion of the commit. See below for a description of
429 the fields in `author`, as they are identical to `committer`.
430
431 `committer`
432 ^^^^^^^^^^^
433 The `committer` command indicates who made this commit, and when
434 they made it.
435
436 Here `<name>` is the person's display name (for example
437 ``Com M Itter'') and `<email>` is the person's email address
438 (``\cm@example.com''). `LT` and `GT` are the literal less-than (\x3c)
439 and greater-than (\x3e) symbols. These are required to delimit
440 the email address from the other fields in the line. Note that
441 `<name>` and `<email>` are free-form and may contain any sequence
442 of bytes, except `LT`, `GT` and `LF`. `<name>` is typically UTF-8 encoded.
443
444 The time of the change is specified by `<when>` using the date format
445 that was selected by the --date-format=<fmt> command-line option.
446 See ``Date Formats'' above for the set of supported formats, and
447 their syntax.
448
449 `from`
450 ^^^^^^
451 The `from` command is used to specify the commit to initialize
452 this branch from. This revision will be the first ancestor of the
453 new commit. The state of the tree built at this commit will begin
454 with the state at the `from` commit, and be altered by the content
455 modifications in this commit.
456
457 Omitting the `from` command in the first commit of a new branch
458 will cause fast-import to create that commit with no ancestor. This
459 tends to be desired only for the initial commit of a project.
460 If the frontend creates all files from scratch when making a new
461 branch, a `merge` command may be used instead of `from` to start
462 the commit with an empty tree.
463 Omitting the `from` command on existing branches is usually desired,
464 as the current commit on that branch is automatically assumed to
465 be the first ancestor of the new commit.
466
467 As `LF` is not valid in a Git refname or SHA-1 expression, no
468 quoting or escaping syntax is supported within `<commit-ish>`.
469
470 Here `<commit-ish>` is any of the following:
471
472 * The name of an existing branch already in fast-import's internal branch
473 table. If fast-import doesn't know the name, it's treated as a SHA-1
474 expression.
475
476 * A mark reference, `:<idnum>`, where `<idnum>` is the mark number.
477 +
478 The reason fast-import uses `:` to denote a mark reference is this character
479 is not legal in a Git branch name. The leading `:` makes it easy
480 to distinguish between the mark 42 (`:42`) and the branch 42 (`42`
481 or `refs/heads/42`), or an abbreviated SHA-1 which happened to
482 consist only of base-10 digits.
483 +
484 Marks must be declared (via `mark`) before they can be used.
485
486 * A complete 40 byte or abbreviated commit SHA-1 in hex.
487
488 * Any valid Git SHA-1 expression that resolves to a commit. See
489 ``SPECIFYING REVISIONS'' in linkgit:gitrevisions[7] for details.
490
491 * The special null SHA-1 (40 zeros) specifies that the branch is to be
492 removed.
493
494 The special case of restarting an incremental import from the
495 current branch value should be written as:
496 ----
497 from refs/heads/branch^0
498 ----
499 The `^0` suffix is necessary as fast-import does not permit a branch to
500 start from itself, and the branch is created in memory before the
501 `from` command is even read from the input. Adding `^0` will force
502 fast-import to resolve the commit through Git's revision parsing library,
503 rather than its internal branch table, thereby loading in the
504 existing value of the branch.
505
506 `merge`
507 ^^^^^^^
508 Includes one additional ancestor commit. The additional ancestry
509 link does not change the way the tree state is built at this commit.
510 If the `from` command is
511 omitted when creating a new branch, the first `merge` commit will be
512 the first ancestor of the current commit, and the branch will start
513 out with no files. An unlimited number of `merge` commands per
514 commit are permitted by fast-import, thereby establishing an n-way merge.
515
516 Here `<commit-ish>` is any of the commit specification expressions
517 also accepted by `from` (see above).
518
519 `filemodify`
520 ^^^^^^^^^^^^
521 Included in a `commit` command to add a new file or change the
522 content of an existing file. This command has two different means
523 of specifying the content of the file.
524
525 External data format::
526 The data content for the file was already supplied by a prior
527 `blob` command. The frontend just needs to connect it.
528 +
529 ....
530 'M' SP <mode> SP <dataref> SP <path> LF
531 ....
532 +
533 Here usually `<dataref>` must be either a mark reference (`:<idnum>`)
534 set by a prior `blob` command, or a full 40-byte SHA-1 of an
535 existing Git blob object. If `<mode>` is `040000`` then
536 `<dataref>` must be the full 40-byte SHA-1 of an existing
537 Git tree object or a mark reference set with `--import-marks`.
538
539 Inline data format::
540 The data content for the file has not been supplied yet.
541 The frontend wants to supply it as part of this modify
542 command.
543 +
544 ....
545 'M' SP <mode> SP 'inline' SP <path> LF
546 data
547 ....
548 +
549 See below for a detailed description of the `data` command.
550
551 In both formats `<mode>` is the type of file entry, specified
552 in octal. Git only supports the following modes:
553
554 * `100644` or `644`: A normal (not-executable) file. The majority
555 of files in most projects use this mode. If in doubt, this is
556 what you want.
557 * `100755` or `755`: A normal, but executable, file.
558 * `120000`: A symlink, the content of the file will be the link target.
559 * `160000`: A gitlink, SHA-1 of the object refers to a commit in
560 another repository. Git links can only be specified by SHA or through
561 a commit mark. They are used to implement submodules.
562 * `040000`: A subdirectory. Subdirectories can only be specified by
563 SHA or through a tree mark set with `--import-marks`.
564
565 In both formats `<path>` is the complete path of the file to be added
566 (if not already existing) or modified (if already existing).
567
568 A `<path>` string must use UNIX-style directory separators (forward
569 slash `/`), may contain any byte other than `LF`, and must not
570 start with double quote (`"`).
571
572 A path can use C-style string quoting; this is accepted in all cases
573 and mandatory if the filename starts with double quote or contains
574 `LF`. In C-style quoting, the complete name should be surrounded with
575 double quotes, and any `LF`, backslash, or double quote characters
576 must be escaped by preceding them with a backslash (e.g.,
577 `"path/with\n, \\ and \" in it"`).
578
579 The value of `<path>` must be in canonical form. That is it must not:
580
581 * contain an empty directory component (e.g. `foo//bar` is invalid),
582 * end with a directory separator (e.g. `foo/` is invalid),
583 * start with a directory separator (e.g. `/foo` is invalid),
584 * contain the special component `.` or `..` (e.g. `foo/./bar` and
585 `foo/../bar` are invalid).
586
587 The root of the tree can be represented by an empty string as `<path>`.
588
589 It is recommended that `<path>` always be encoded using UTF-8.
590
591 `filedelete`
592 ^^^^^^^^^^^^
593 Included in a `commit` command to remove a file or recursively
594 delete an entire directory from the branch. If the file or directory
595 removal makes its parent directory empty, the parent directory will
596 be automatically removed too. This cascades up the tree until the
597 first non-empty directory or the root is reached.
598
599 ....
600 'D' SP <path> LF
601 ....
602
603 here `<path>` is the complete path of the file or subdirectory to
604 be removed from the branch.
605 See `filemodify` above for a detailed description of `<path>`.
606
607 `filecopy`
608 ^^^^^^^^^^
609 Recursively copies an existing file or subdirectory to a different
610 location within the branch. The existing file or directory must
611 exist. If the destination exists it will be completely replaced
612 by the content copied from the source.
613
614 ....
615 'C' SP <path> SP <path> LF
616 ....
617
618 here the first `<path>` is the source location and the second
619 `<path>` is the destination. See `filemodify` above for a detailed
620 description of what `<path>` may look like. To use a source path
621 that contains SP the path must be quoted.
622
623 A `filecopy` command takes effect immediately. Once the source
624 location has been copied to the destination any future commands
625 applied to the source location will not impact the destination of
626 the copy.
627
628 `filerename`
629 ^^^^^^^^^^^^
630 Renames an existing file or subdirectory to a different location
631 within the branch. The existing file or directory must exist. If
632 the destination exists it will be replaced by the source directory.
633
634 ....
635 'R' SP <path> SP <path> LF
636 ....
637
638 here the first `<path>` is the source location and the second
639 `<path>` is the destination. See `filemodify` above for a detailed
640 description of what `<path>` may look like. To use a source path
641 that contains SP the path must be quoted.
642
643 A `filerename` command takes effect immediately. Once the source
644 location has been renamed to the destination any future commands
645 applied to the source location will create new files there and not
646 impact the destination of the rename.
647
648 Note that a `filerename` is the same as a `filecopy` followed by a
649 `filedelete` of the source location. There is a slight performance
650 advantage to using `filerename`, but the advantage is so small
651 that it is never worth trying to convert a delete/add pair in
652 source material into a rename for fast-import. This `filerename`
653 command is provided just to simplify frontends that already have
654 rename information and don't want bother with decomposing it into a
655 `filecopy` followed by a `filedelete`.
656
657 `filedeleteall`
658 ^^^^^^^^^^^^^^^
659 Included in a `commit` command to remove all files (and also all
660 directories) from the branch. This command resets the internal
661 branch structure to have no files in it, allowing the frontend
662 to subsequently add all interesting files from scratch.
663
664 ....
665 'deleteall' LF
666 ....
667
668 This command is extremely useful if the frontend does not know
669 (or does not care to know) what files are currently on the branch,
670 and therefore cannot generate the proper `filedelete` commands to
671 update the content.
672
673 Issuing a `filedeleteall` followed by the needed `filemodify`
674 commands to set the correct content will produce the same results
675 as sending only the needed `filemodify` and `filedelete` commands.
676 The `filedeleteall` approach may however require fast-import to use slightly
677 more memory per active branch (less than 1 MiB for even most large
678 projects); so frontends that can easily obtain only the affected
679 paths for a commit are encouraged to do so.
680
681 `notemodify`
682 ^^^^^^^^^^^^
683 Included in a `commit` `<notes_ref>` command to add a new note
684 annotating a `<commit-ish>` or change this annotation contents.
685 Internally it is similar to filemodify 100644 on `<commit-ish>`
686 path (maybe split into subdirectories). It's not advised to
687 use any other commands to write to the `<notes_ref>` tree except
688 `filedeleteall` to delete all existing notes in this tree.
689 This command has two different means of specifying the content
690 of the note.
691
692 External data format::
693 The data content for the note was already supplied by a prior
694 `blob` command. The frontend just needs to connect it to the
695 commit that is to be annotated.
696 +
697 ....
698 'N' SP <dataref> SP <commit-ish> LF
699 ....
700 +
701 Here `<dataref>` can be either a mark reference (`:<idnum>`)
702 set by a prior `blob` command, or a full 40-byte SHA-1 of an
703 existing Git blob object.
704
705 Inline data format::
706 The data content for the note has not been supplied yet.
707 The frontend wants to supply it as part of this modify
708 command.
709 +
710 ....
711 'N' SP 'inline' SP <commit-ish> LF
712 data
713 ....
714 +
715 See below for a detailed description of the `data` command.
716
717 In both formats `<commit-ish>` is any of the commit specification
718 expressions also accepted by `from` (see above).
719
720 `mark`
721 ~~~~~~
722 Arranges for fast-import to save a reference to the current object, allowing
723 the frontend to recall this object at a future point in time, without
724 knowing its SHA-1. Here the current object is the object creation
725 command the `mark` command appears within. This can be `commit`,
726 `tag`, and `blob`, but `commit` is the most common usage.
727
728 ....
729 'mark' SP ':' <idnum> LF
730 ....
731
732 where `<idnum>` is the number assigned by the frontend to this mark.
733 The value of `<idnum>` is expressed as an ASCII decimal integer.
734 The value 0 is reserved and cannot be used as
735 a mark. Only values greater than or equal to 1 may be used as marks.
736
737 New marks are created automatically. Existing marks can be moved
738 to another object simply by reusing the same `<idnum>` in another
739 `mark` command.
740
741 `tag`
742 ~~~~~
743 Creates an annotated tag referring to a specific commit. To create
744 lightweight (non-annotated) tags see the `reset` command below.
745
746 ....
747 'tag' SP <name> LF
748 'from' SP <commit-ish> LF
749 'tagger' (SP <name>)? SP LT <email> GT SP <when> LF
750 data
751 ....
752
753 where `<name>` is the name of the tag to create.
754
755 Tag names are automatically prefixed with `refs/tags/` when stored
756 in Git, so importing the CVS branch symbol `RELENG-1_0-FINAL` would
757 use just `RELENG-1_0-FINAL` for `<name>`, and fast-import will write the
758 corresponding ref as `refs/tags/RELENG-1_0-FINAL`.
759
760 The value of `<name>` must be a valid refname in Git and therefore
761 may contain forward slashes. As `LF` is not valid in a Git refname,
762 no quoting or escaping syntax is supported here.
763
764 The `from` command is the same as in the `commit` command; see
765 above for details.
766
767 The `tagger` command uses the same format as `committer` within
768 `commit`; again see above for details.
769
770 The `data` command following `tagger` must supply the annotated tag
771 message (see below for `data` command syntax). To import an empty
772 tag message use a 0 length data. Tag messages are free-form and are
773 not interpreted by Git. Currently they must be encoded in UTF-8,
774 as fast-import does not permit other encodings to be specified.
775
776 Signing annotated tags during import from within fast-import is not
777 supported. Trying to include your own PGP/GPG signature is not
778 recommended, as the frontend does not (easily) have access to the
779 complete set of bytes which normally goes into such a signature.
780 If signing is required, create lightweight tags from within fast-import with
781 `reset`, then create the annotated versions of those tags offline
782 with the standard 'git tag' process.
783
784 `reset`
785 ~~~~~~~
786 Creates (or recreates) the named branch, optionally starting from
787 a specific revision. The reset command allows a frontend to issue
788 a new `from` command for an existing branch, or to create a new
789 branch from an existing commit without creating a new commit.
790
791 ....
792 'reset' SP <ref> LF
793 ('from' SP <commit-ish> LF)?
794 LF?
795 ....
796
797 For a detailed description of `<ref>` and `<commit-ish>` see above
798 under `commit` and `from`.
799
800 The `LF` after the command is optional (it used to be required).
801
802 The `reset` command can also be used to create lightweight
803 (non-annotated) tags. For example:
804
805 ====
806 reset refs/tags/938
807 from :938
808 ====
809
810 would create the lightweight tag `refs/tags/938` referring to
811 whatever commit mark `:938` references.
812
813 `blob`
814 ~~~~~~
815 Requests writing one file revision to the packfile. The revision
816 is not connected to any commit; this connection must be formed in
817 a subsequent `commit` command by referencing the blob through an
818 assigned mark.
819
820 ....
821 'blob' LF
822 mark?
823 data
824 ....
825
826 The mark command is optional here as some frontends have chosen
827 to generate the Git SHA-1 for the blob on their own, and feed that
828 directly to `commit`. This is typically more work than it's worth
829 however, as marks are inexpensive to store and easy to use.
830
831 `data`
832 ~~~~~~
833 Supplies raw data (for use as blob/file content, commit messages, or
834 annotated tag messages) to fast-import. Data can be supplied using an exact
835 byte count or delimited with a terminating line. Real frontends
836 intended for production-quality conversions should always use the
837 exact byte count format, as it is more robust and performs better.
838 The delimited format is intended primarily for testing fast-import.
839
840 Comment lines appearing within the `<raw>` part of `data` commands
841 are always taken to be part of the body of the data and are therefore
842 never ignored by fast-import. This makes it safe to import any
843 file/message content whose lines might start with `#`.
844
845 Exact byte count format::
846 The frontend must specify the number of bytes of data.
847 +
848 ....
849 'data' SP <count> LF
850 <raw> LF?
851 ....
852 +
853 where `<count>` is the exact number of bytes appearing within
854 `<raw>`. The value of `<count>` is expressed as an ASCII decimal
855 integer. The `LF` on either side of `<raw>` is not
856 included in `<count>` and will not be included in the imported data.
857 +
858 The `LF` after `<raw>` is optional (it used to be required) but
859 recommended. Always including it makes debugging a fast-import
860 stream easier as the next command always starts in column 0
861 of the next line, even if `<raw>` did not end with an `LF`.
862
863 Delimited format::
864 A delimiter string is used to mark the end of the data.
865 fast-import will compute the length by searching for the delimiter.
866 This format is primarily useful for testing and is not
867 recommended for real data.
868 +
869 ....
870 'data' SP '<<' <delim> LF
871 <raw> LF
872 <delim> LF
873 LF?
874 ....
875 +
876 where `<delim>` is the chosen delimiter string. The string `<delim>`
877 must not appear on a line by itself within `<raw>`, as otherwise
878 fast-import will think the data ends earlier than it really does. The `LF`
879 immediately trailing `<raw>` is part of `<raw>`. This is one of
880 the limitations of the delimited format, it is impossible to supply
881 a data chunk which does not have an LF as its last byte.
882 +
883 The `LF` after `<delim> LF` is optional (it used to be required).
884
885 `checkpoint`
886 ~~~~~~~~~~~~
887 Forces fast-import to close the current packfile, start a new one, and to
888 save out all current branch refs, tags and marks.
889
890 ....
891 'checkpoint' LF
892 LF?
893 ....
894
895 Note that fast-import automatically switches packfiles when the current
896 packfile reaches --max-pack-size, or 4 GiB, whichever limit is
897 smaller. During an automatic packfile switch fast-import does not update
898 the branch refs, tags or marks.
899
900 As a `checkpoint` can require a significant amount of CPU time and
901 disk IO (to compute the overall pack SHA-1 checksum, generate the
902 corresponding index file, and update the refs) it can easily take
903 several minutes for a single `checkpoint` command to complete.
904
905 Frontends may choose to issue checkpoints during extremely large
906 and long running imports, or when they need to allow another Git
907 process access to a branch. However given that a 30 GiB Subversion
908 repository can be loaded into Git through fast-import in about 3 hours,
909 explicit checkpointing may not be necessary.
910
911 The `LF` after the command is optional (it used to be required).
912
913 `progress`
914 ~~~~~~~~~~
915 Causes fast-import to print the entire `progress` line unmodified to
916 its standard output channel (file descriptor 1) when the command is
917 processed from the input stream. The command otherwise has no impact
918 on the current import, or on any of fast-import's internal state.
919
920 ....
921 'progress' SP <any> LF
922 LF?
923 ....
924
925 The `<any>` part of the command may contain any sequence of bytes
926 that does not contain `LF`. The `LF` after the command is optional.
927 Callers may wish to process the output through a tool such as sed to
928 remove the leading part of the line, for example:
929
930 ====
931 frontend | git fast-import | sed 's/^progress //'
932 ====
933
934 Placing a `progress` command immediately after a `checkpoint` will
935 inform the reader when the `checkpoint` has been completed and it
936 can safely access the refs that fast-import updated.
937
938 `get-mark`
939 ~~~~~~~~~~
940 Causes fast-import to print the SHA-1 corresponding to a mark to
941 stdout or to the file descriptor previously arranged with the
942 `--cat-blob-fd` argument. The command otherwise has no impact on the
943 current import; its purpose is to retrieve SHA-1s that later commits
944 might want to refer to in their commit messages.
945
946 ....
947 'get-mark' SP ':' <idnum> LF
948 ....
949
950 This command can be used anywhere in the stream that comments are
951 accepted. In particular, the `get-mark` command can be used in the
952 middle of a commit but not in the middle of a `data` command.
953
954 See ``Responses To Commands'' below for details about how to read
955 this output safely.
956
957 `cat-blob`
958 ~~~~~~~~~~
959 Causes fast-import to print a blob to a file descriptor previously
960 arranged with the `--cat-blob-fd` argument. The command otherwise
961 has no impact on the current import; its main purpose is to
962 retrieve blobs that may be in fast-import's memory but not
963 accessible from the target repository.
964
965 ....
966 'cat-blob' SP <dataref> LF
967 ....
968
969 The `<dataref>` can be either a mark reference (`:<idnum>`)
970 set previously or a full 40-byte SHA-1 of a Git blob, preexisting or
971 ready to be written.
972
973 Output uses the same format as `git cat-file --batch`:
974
975 ====
976 <sha1> SP 'blob' SP <size> LF
977 <contents> LF
978 ====
979
980 This command can be used anywhere in the stream that comments are
981 accepted. In particular, the `cat-blob` command can be used in the
982 middle of a commit but not in the middle of a `data` command.
983
984 See ``Responses To Commands'' below for details about how to read
985 this output safely.
986
987 `ls`
988 ~~~~
989 Prints information about the object at a path to a file descriptor
990 previously arranged with the `--cat-blob-fd` argument. This allows
991 printing a blob from the active commit (with `cat-blob`) or copying a
992 blob or tree from a previous commit for use in the current one (with
993 `filemodify`).
994
995 The `ls` command can be used anywhere in the stream that comments are
996 accepted, including the middle of a commit.
997
998 Reading from the active commit::
999 This form can only be used in the middle of a `commit`.
1000 The path names a directory entry within fast-import's
1001 active commit. The path must be quoted in this case.
1002 +
1003 ....
1004 'ls' SP <path> LF
1005 ....
1006
1007 Reading from a named tree::
1008 The `<dataref>` can be a mark reference (`:<idnum>`) or the
1009 full 40-byte SHA-1 of a Git tag, commit, or tree object,
1010 preexisting or waiting to be written.
1011 The path is relative to the top level of the tree
1012 named by `<dataref>`.
1013 +
1014 ....
1015 'ls' SP <dataref> SP <path> LF
1016 ....
1017
1018 See `filemodify` above for a detailed description of `<path>`.
1019
1020 Output uses the same format as `git ls-tree <tree> -- <path>`:
1021
1022 ====
1023 <mode> SP ('blob' | 'tree' | 'commit') SP <dataref> HT <path> LF
1024 ====
1025
1026 The <dataref> represents the blob, tree, or commit object at <path>
1027 and can be used in later 'get-mark', 'cat-blob', 'filemodify', or
1028 'ls' commands.
1029
1030 If there is no file or subtree at that path, 'git fast-import' will
1031 instead report
1032
1033 ====
1034 missing SP <path> LF
1035 ====
1036
1037 See ``Responses To Commands'' below for details about how to read
1038 this output safely.
1039
1040 `feature`
1041 ~~~~~~~~~
1042 Require that fast-import supports the specified feature, or abort if
1043 it does not.
1044
1045 ....
1046 'feature' SP <feature> ('=' <argument>)? LF
1047 ....
1048
1049 The <feature> part of the command may be any one of the following:
1050
1051 date-format::
1052 export-marks::
1053 relative-marks::
1054 no-relative-marks::
1055 force::
1056 Act as though the corresponding command-line option with
1057 a leading '--' was passed on the command line
1058 (see OPTIONS, above).
1059
1060 import-marks::
1061 import-marks-if-exists::
1062 Like --import-marks except in two respects: first, only one
1063 "feature import-marks" or "feature import-marks-if-exists"
1064 command is allowed per stream; second, an --import-marks=
1065 or --import-marks-if-exists command-line option overrides
1066 any of these "feature" commands in the stream; third,
1067 "feature import-marks-if-exists" like a corresponding
1068 command-line option silently skips a nonexistent file.
1069
1070 get-mark::
1071 cat-blob::
1072 ls::
1073 Require that the backend support the 'get-mark', 'cat-blob',
1074 or 'ls' command respectively.
1075 Versions of fast-import not supporting the specified command
1076 will exit with a message indicating so.
1077 This lets the import error out early with a clear message,
1078 rather than wasting time on the early part of an import
1079 before the unsupported command is detected.
1080
1081 notes::
1082 Require that the backend support the 'notemodify' (N)
1083 subcommand to the 'commit' command.
1084 Versions of fast-import not supporting notes will exit
1085 with a message indicating so.
1086
1087 done::
1088 Error out if the stream ends without a 'done' command.
1089 Without this feature, errors causing the frontend to end
1090 abruptly at a convenient point in the stream can go
1091 undetected. This may occur, for example, if an import
1092 front end dies in mid-operation without emitting SIGTERM
1093 or SIGKILL at its subordinate git fast-import instance.
1094
1095 `option`
1096 ~~~~~~~~
1097 Processes the specified option so that git fast-import behaves in a
1098 way that suits the frontend's needs.
1099 Note that options specified by the frontend are overridden by any
1100 options the user may specify to git fast-import itself.
1101
1102 ....
1103 'option' SP <option> LF
1104 ....
1105
1106 The `<option>` part of the command may contain any of the options
1107 listed in the OPTIONS section that do not change import semantics,
1108 without the leading '--' and is treated in the same way.
1109
1110 Option commands must be the first commands on the input (not counting
1111 feature commands), to give an option command after any non-option
1112 command is an error.
1113
1114 The following command-line options change import semantics and may therefore
1115 not be passed as option:
1116
1117 * date-format
1118 * import-marks
1119 * export-marks
1120 * cat-blob-fd
1121 * force
1122
1123 `done`
1124 ~~~~~~
1125 If the `done` feature is not in use, treated as if EOF was read.
1126 This can be used to tell fast-import to finish early.
1127
1128 If the `--done` command-line option or `feature done` command is
1129 in use, the `done` command is mandatory and marks the end of the
1130 stream.
1131
1132 Responses To Commands
1133 ---------------------
1134 New objects written by fast-import are not available immediately.
1135 Most fast-import commands have no visible effect until the next
1136 checkpoint (or completion). The frontend can send commands to
1137 fill fast-import's input pipe without worrying about how quickly
1138 they will take effect, which improves performance by simplifying
1139 scheduling.
1140
1141 For some frontends, though, it is useful to be able to read back
1142 data from the current repository as it is being updated (for
1143 example when the source material describes objects in terms of
1144 patches to be applied to previously imported objects). This can
1145 be accomplished by connecting the frontend and fast-import via
1146 bidirectional pipes:
1147
1148 ====
1149 mkfifo fast-import-output
1150 frontend <fast-import-output |
1151 git fast-import >fast-import-output
1152 ====
1153
1154 A frontend set up this way can use `progress`, `get-mark`, `ls`, and
1155 `cat-blob` commands to read information from the import in progress.
1156
1157 To avoid deadlock, such frontends must completely consume any
1158 pending output from `progress`, `ls`, `get-mark`, and `cat-blob` before
1159 performing writes to fast-import that might block.
1160
1161 Crash Reports
1162 -------------
1163 If fast-import is supplied invalid input it will terminate with a
1164 non-zero exit status and create a crash report in the top level of
1165 the Git repository it was importing into. Crash reports contain
1166 a snapshot of the internal fast-import state as well as the most
1167 recent commands that lead up to the crash.
1168
1169 All recent commands (including stream comments, file changes and
1170 progress commands) are shown in the command history within the crash
1171 report, but raw file data and commit messages are excluded from the
1172 crash report. This exclusion saves space within the report file
1173 and reduces the amount of buffering that fast-import must perform
1174 during execution.
1175
1176 After writing a crash report fast-import will close the current
1177 packfile and export the marks table. This allows the frontend
1178 developer to inspect the repository state and resume the import from
1179 the point where it crashed. The modified branches and tags are not
1180 updated during a crash, as the import did not complete successfully.
1181 Branch and tag information can be found in the crash report and
1182 must be applied manually if the update is needed.
1183
1184 An example crash:
1185
1186 ====
1187 $ cat >in <<END_OF_INPUT
1188 # my very first test commit
1189 commit refs/heads/master
1190 committer Shawn O. Pearce <spearce> 19283 -0400
1191 # who is that guy anyway?
1192 data <<EOF
1193 this is my commit
1194 EOF
1195 M 644 inline .gitignore
1196 data <<EOF
1197 .gitignore
1198 EOF
1199 M 777 inline bob
1200 END_OF_INPUT
1201
1202 $ git fast-import <in
1203 fatal: Corrupt mode: M 777 inline bob
1204 fast-import: dumping crash report to .git/fast_import_crash_8434
1205
1206 $ cat .git/fast_import_crash_8434
1207 fast-import crash report:
1208 fast-import process: 8434
1209 parent process : 1391
1210 at Sat Sep 1 00:58:12 2007
1211
1212 fatal: Corrupt mode: M 777 inline bob
1213
1214 Most Recent Commands Before Crash
1215 ---------------------------------
1216 # my very first test commit
1217 commit refs/heads/master
1218 committer Shawn O. Pearce <spearce> 19283 -0400
1219 # who is that guy anyway?
1220 data <<EOF
1221 M 644 inline .gitignore
1222 data <<EOF
1223 * M 777 inline bob
1224
1225 Active Branch LRU
1226 -----------------
1227 active_branches = 1 cur, 5 max
1228
1229 pos clock name
1230 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1231 1) 0 refs/heads/master
1232
1233 Inactive Branches
1234 -----------------
1235 refs/heads/master:
1236 status : active loaded dirty
1237 tip commit : 0000000000000000000000000000000000000000
1238 old tree : 0000000000000000000000000000000000000000
1239 cur tree : 0000000000000000000000000000000000000000
1240 commit clock: 0
1241 last pack :
1242
1243
1244 -------------------
1245 END OF CRASH REPORT
1246 ====
1247
1248 Tips and Tricks
1249 ---------------
1250 The following tips and tricks have been collected from various
1251 users of fast-import, and are offered here as suggestions.
1252
1253 Use One Mark Per Commit
1254 ~~~~~~~~~~~~~~~~~~~~~~~
1255 When doing a repository conversion, use a unique mark per commit
1256 (`mark :<n>`) and supply the --export-marks option on the command
1257 line. fast-import will dump a file which lists every mark and the Git
1258 object SHA-1 that corresponds to it. If the frontend can tie
1259 the marks back to the source repository, it is easy to verify the
1260 accuracy and completeness of the import by comparing each Git
1261 commit to the corresponding source revision.
1262
1263 Coming from a system such as Perforce or Subversion this should be
1264 quite simple, as the fast-import mark can also be the Perforce changeset
1265 number or the Subversion revision number.
1266
1267 Freely Skip Around Branches
1268 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1269 Don't bother trying to optimize the frontend to stick to one branch
1270 at a time during an import. Although doing so might be slightly
1271 faster for fast-import, it tends to increase the complexity of the frontend
1272 code considerably.
1273
1274 The branch LRU builtin to fast-import tends to behave very well, and the
1275 cost of activating an inactive branch is so low that bouncing around
1276 between branches has virtually no impact on import performance.
1277
1278 Handling Renames
1279 ~~~~~~~~~~~~~~~~
1280 When importing a renamed file or directory, simply delete the old
1281 name(s) and modify the new name(s) during the corresponding commit.
1282 Git performs rename detection after-the-fact, rather than explicitly
1283 during a commit.
1284
1285 Use Tag Fixup Branches
1286 ~~~~~~~~~~~~~~~~~~~~~~
1287 Some other SCM systems let the user create a tag from multiple
1288 files which are not from the same commit/changeset. Or to create
1289 tags which are a subset of the files available in the repository.
1290
1291 Importing these tags as-is in Git is impossible without making at
1292 least one commit which ``fixes up'' the files to match the content
1293 of the tag. Use fast-import's `reset` command to reset a dummy branch
1294 outside of your normal branch space to the base commit for the tag,
1295 then commit one or more file fixup commits, and finally tag the
1296 dummy branch.
1297
1298 For example since all normal branches are stored under `refs/heads/`
1299 name the tag fixup branch `TAG_FIXUP`. This way it is impossible for
1300 the fixup branch used by the importer to have namespace conflicts
1301 with real branches imported from the source (the name `TAG_FIXUP`
1302 is not `refs/heads/TAG_FIXUP`).
1303
1304 When committing fixups, consider using `merge` to connect the
1305 commit(s) which are supplying file revisions to the fixup branch.
1306 Doing so will allow tools such as 'git blame' to track
1307 through the real commit history and properly annotate the source
1308 files.
1309
1310 After fast-import terminates the frontend will need to do `rm .git/TAG_FIXUP`
1311 to remove the dummy branch.
1312
1313 Import Now, Repack Later
1314 ~~~~~~~~~~~~~~~~~~~~~~~~
1315 As soon as fast-import completes the Git repository is completely valid
1316 and ready for use. Typically this takes only a very short time,
1317 even for considerably large projects (100,000+ commits).
1318
1319 However repacking the repository is necessary to improve data
1320 locality and access performance. It can also take hours on extremely
1321 large projects (especially if -f and a large --window parameter is
1322 used). Since repacking is safe to run alongside readers and writers,
1323 run the repack in the background and let it finish when it finishes.
1324 There is no reason to wait to explore your new Git project!
1325
1326 If you choose to wait for the repack, don't try to run benchmarks
1327 or performance tests until repacking is completed. fast-import outputs
1328 suboptimal packfiles that are simply never seen in real use
1329 situations.
1330
1331 Repacking Historical Data
1332 ~~~~~~~~~~~~~~~~~~~~~~~~~
1333 If you are repacking very old imported data (e.g. older than the
1334 last year), consider expending some extra CPU time and supplying
1335 --window=50 (or higher) when you run 'git repack'.
1336 This will take longer, but will also produce a smaller packfile.
1337 You only need to expend the effort once, and everyone using your
1338 project will benefit from the smaller repository.
1339
1340 Include Some Progress Messages
1341 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1342 Every once in a while have your frontend emit a `progress` message
1343 to fast-import. The contents of the messages are entirely free-form,
1344 so one suggestion would be to output the current month and year
1345 each time the current commit date moves into the next month.
1346 Your users will feel better knowing how much of the data stream
1347 has been processed.
1348
1349
1350 Packfile Optimization
1351 ---------------------
1352 When packing a blob fast-import always attempts to deltify against the last
1353 blob written. Unless specifically arranged for by the frontend,
1354 this will probably not be a prior version of the same file, so the
1355 generated delta will not be the smallest possible. The resulting
1356 packfile will be compressed, but will not be optimal.
1357
1358 Frontends which have efficient access to all revisions of a
1359 single file (for example reading an RCS/CVS ,v file) can choose
1360 to supply all revisions of that file as a sequence of consecutive
1361 `blob` commands. This allows fast-import to deltify the different file
1362 revisions against each other, saving space in the final packfile.
1363 Marks can be used to later identify individual file revisions during
1364 a sequence of `commit` commands.
1365
1366 The packfile(s) created by fast-import do not encourage good disk access
1367 patterns. This is caused by fast-import writing the data in the order
1368 it is received on standard input, while Git typically organizes
1369 data within packfiles to make the most recent (current tip) data
1370 appear before historical data. Git also clusters commits together,
1371 speeding up revision traversal through better cache locality.
1372
1373 For this reason it is strongly recommended that users repack the
1374 repository with `git repack -a -d` after fast-import completes, allowing
1375 Git to reorganize the packfiles for faster data access. If blob
1376 deltas are suboptimal (see above) then also adding the `-f` option
1377 to force recomputation of all deltas can significantly reduce the
1378 final packfile size (30-50% smaller can be quite typical).
1379
1380
1381 Memory Utilization
1382 ------------------
1383 There are a number of factors which affect how much memory fast-import
1384 requires to perform an import. Like critical sections of core
1385 Git, fast-import uses its own memory allocators to amortize any overheads
1386 associated with malloc. In practice fast-import tends to amortize any
1387 malloc overheads to 0, due to its use of large block allocations.
1388
1389 per object
1390 ~~~~~~~~~~
1391 fast-import maintains an in-memory structure for every object written in
1392 this execution. On a 32 bit system the structure is 32 bytes,
1393 on a 64 bit system the structure is 40 bytes (due to the larger
1394 pointer sizes). Objects in the table are not deallocated until
1395 fast-import terminates. Importing 2 million objects on a 32 bit system
1396 will require approximately 64 MiB of memory.
1397
1398 The object table is actually a hashtable keyed on the object name
1399 (the unique SHA-1). This storage configuration allows fast-import to reuse
1400 an existing or already written object and avoid writing duplicates
1401 to the output packfile. Duplicate blobs are surprisingly common
1402 in an import, typically due to branch merges in the source.
1403
1404 per mark
1405 ~~~~~~~~
1406 Marks are stored in a sparse array, using 1 pointer (4 bytes or 8
1407 bytes, depending on pointer size) per mark. Although the array
1408 is sparse, frontends are still strongly encouraged to use marks
1409 between 1 and n, where n is the total number of marks required for
1410 this import.
1411
1412 per branch
1413 ~~~~~~~~~~
1414 Branches are classified as active and inactive. The memory usage
1415 of the two classes is significantly different.
1416
1417 Inactive branches are stored in a structure which uses 96 or 120
1418 bytes (32 bit or 64 bit systems, respectively), plus the length of
1419 the branch name (typically under 200 bytes), per branch. fast-import will
1420 easily handle as many as 10,000 inactive branches in under 2 MiB
1421 of memory.
1422
1423 Active branches have the same overhead as inactive branches, but
1424 also contain copies of every tree that has been recently modified on
1425 that branch. If subtree `include` has not been modified since the
1426 branch became active, its contents will not be loaded into memory,
1427 but if subtree `src` has been modified by a commit since the branch
1428 became active, then its contents will be loaded in memory.
1429
1430 As active branches store metadata about the files contained on that
1431 branch, their in-memory storage size can grow to a considerable size
1432 (see below).
1433
1434 fast-import automatically moves active branches to inactive status based on
1435 a simple least-recently-used algorithm. The LRU chain is updated on
1436 each `commit` command. The maximum number of active branches can be
1437 increased or decreased on the command line with --active-branches=.
1438
1439 per active tree
1440 ~~~~~~~~~~~~~~~
1441 Trees (aka directories) use just 12 bytes of memory on top of the
1442 memory required for their entries (see ``per active file'' below).
1443 The cost of a tree is virtually 0, as its overhead amortizes out
1444 over the individual file entries.
1445
1446 per active file entry
1447 ~~~~~~~~~~~~~~~~~~~~~
1448 Files (and pointers to subtrees) within active trees require 52 or 64
1449 bytes (32/64 bit platforms) per entry. To conserve space, file and
1450 tree names are pooled in a common string table, allowing the filename
1451 ``Makefile'' to use just 16 bytes (after including the string header
1452 overhead) no matter how many times it occurs within the project.
1453
1454 The active branch LRU, when coupled with the filename string pool
1455 and lazy loading of subtrees, allows fast-import to efficiently import
1456 projects with 2,000+ branches and 45,114+ files in a very limited
1457 memory footprint (less than 2.7 MiB per active branch).
1458
1459 Signals
1460 -------
1461 Sending *SIGUSR1* to the 'git fast-import' process ends the current
1462 packfile early, simulating a `checkpoint` command. The impatient
1463 operator can use this facility to peek at the objects and refs from an
1464 import in progress, at the cost of some added running time and worse
1465 compression.
1466
1467 SEE ALSO
1468 --------
1469 linkgit:git-fast-export[1]
1470
1471 GIT
1472 ---
1473 Part of the linkgit:git[1] suite