summaryrefslogtreecommitdiff
path: root/README
blob: 6b38a7aa7a711c327d82e2da5e474544a7befc34 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
////////////////////////////////////////////////////////////////
 
	GIT - the stupid content tracker
 
////////////////////////////////////////////////////////////////
"git" can mean anything, depending on your mood.
 
 - random three-letter combination that is pronounceable, and not
   actually used by any common UNIX command.  The fact that it is a
   mispronunciation of "get" may or may not be relevant.
 - stupid. contemptible and despicable. simple. Take your pick from the
   dictionary of slang.
 - "global information tracker": you're in a good mood, and it actually
   works for you. Angels sing, and a light suddenly fills the room. 
 - "goddamn idiotic truckload of sh*t": when it breaks
 
This is a stupid (but extremely fast) directory content manager.  It
doesn't do a whole lot, but what it 'does' do is track directory
contents efficiently. 
 
There are two object abstractions: the "object database", and the
"current directory cache" aka "index".
 
The Object Database
~~~~~~~~~~~~~~~~~~~
The object database is literally just a content-addressable collection
of objects.  All objects are named by their content, which is
approximated by the SHA1 hash of the object itself.  Objects may refer
to other objects (by referencing their SHA1 hash), and so you can
build up a hierarchy of objects.
 
All objects have a statically determined "type" aka "tag", which is
determined at object creation time, and which identifies the format of
the object (i.e. how it is used, and how it can refer to other
objects).  There are currently four different object types: "blob",
"tree", "commit" and "tag".
 
A "blob" object cannot refer to any other object, and is, like the tag
implies, a pure storage object containing some user data.  It is used to
actually store the file data, i.e. a blob object is associated with some
particular version of some file. 
 
A "tree" object is an object that ties one or more "blob" objects into a
directory structure. In addition, a tree object can refer to other tree
objects, thus creating a directory hierarchy. 
 
A "commit" object ties such directory hierarchies together into
a DAG of revisions - each "commit" is associated with exactly one tree
(the directory hierarchy at the time of the commit). In addition, a
"commit" refers to one or more "parent" commit objects that describe the
history of how we arrived at that directory hierarchy.
 
As a special case, a commit object with no parents is called the "root"
object, and is the point of an initial project commit.  Each project
must have at least one root, and while you can tie several different
root objects together into one project by creating a commit object which
has two or more separate roots as its ultimate parents, that's probably
just going to confuse people.  So aim for the notion of "one root object
per project", even if git itself does not enforce that. 
 
A "tag" object symbolically identifies and can be used to sign other
objects. It contains the identifier and type of another object, a
symbolic name (of course!) and, optionally, a signature.
 
Regardless of object type, all objects share the following
characteristics: they are all deflated with zlib, and have a header
that not only specifies their tag, but also provides size information
about the data in the object.  It's worth noting that the SHA1 hash
that is used to name the object is the hash of the original data
plus this header, so `sha1sum` 'file' does not match the object name
for 'file'.
(Historical note: in the dawn of the age of git the hash
was the sha1 of the 'compressed' object.)
 
As a result, the general consistency of an object can always be tested
independently of the contents or the type of the object: all objects can
be validated by verifying that (a) their hashes match the content of the
file and (b) the object successfully inflates to a stream of bytes that
forms a sequence of <ascii tag without space> + <space> + <ascii decimal
size> + <byte\0> + <binary object data>. 
 
The structured objects can further have their structure and
connectivity to other objects verified. This is generally done with
the `git-fsck-objects` program, which generates a full dependency graph
of all objects, and verifies their internal consistency (in addition
to just verifying their superficial consistency through the hash).
 
The object types in some more detail:
 
Blob Object
~~~~~~~~~~~
A "blob" object is nothing but a binary blob of data, and doesn't
refer to anything else.  There is no signature or any other
verification of the data, so while the object is consistent (it 'is'
indexed by its sha1 hash, so the data itself is certainly correct), it
has absolutely no other attributes.  No name associations, no
permissions.  It is purely a blob of data (i.e. normally "file
contents").
 
In particular, since the blob is entirely defined by its data, if two
files in a directory tree (or in multiple different versions of the
repository) have the same contents, they will share the same blob
object. The object is totally independent of its location in the
directory tree, and renaming a file does not change the object that
file is associated with in any way.
 
A blob is typically created when link:git-update-index.html[git-update-index]
is run, and its data can be accessed by link:git-cat-file.html[git-cat-file].
 
Tree Object
~~~~~~~~~~~
The next hierarchical object type is the "tree" object.  A tree object
is a list of mode/name/blob data, sorted by name.  Alternatively, the
mode data may specify a directory mode, in which case instead of
naming a blob, that name is associated with another TREE object.
 
Like the "blob" object, a tree object is uniquely determined by the
set contents, and so two separate but identical trees will always
share the exact same object. This is true at all levels, i.e. it's
true for a "leaf" tree (which does not refer to any other trees, only
blobs) as well as for a whole subdirectory.
 
For that reason a "tree" object is just a pure data abstraction: it
has no history, no signatures, no verification of validity, except
that since the contents are again protected by the hash itself, we can
trust that the tree is immutable and its contents never change.
 
So you can trust the contents of a tree to be valid, the same way you
can trust the contents of a blob, but you don't know where those
contents 'came' from.
 
Side note on trees: since a "tree" object is a sorted list of
"filename+content", you can create a diff between two trees without
actually having to unpack two trees.  Just ignore all common parts,
and your diff will look right.  In other words, you can effectively
(and efficiently) tell the difference between any two random trees by
O(n) where "n" is the size of the difference, rather than the size of
the tree.
 
Side note 2 on trees: since the name of a "blob" depends entirely and
exclusively on its contents (i.e. there are no names or permissions
involved), you can see trivial renames or permission changes by
noticing that the blob stayed the same.  However, renames with data
changes need a smarter "diff" implementation.
 
A tree is created with link:git-write-tree.html[git-write-tree] and
its data can be accessed by link:git-ls-tree.html[git-ls-tree].
Two trees can be compared with link:git-diff-tree.html[git-diff-tree].
 
Commit Object
~~~~~~~~~~~~~
The "commit" object is an object that introduces the notion of
history into the picture.  In contrast to the other objects, it
doesn't just describe the physical state of a tree, it describes how
we got there, and why.
 
A "commit" is defined by the tree-object that it results in, the
parent commits (zero, one or more) that led up to that point, and a
comment on what happened.  Again, a commit is not trusted per se:
the contents are well-defined and "safe" due to the cryptographically
strong signatures at all levels, but there is no reason to believe
that the tree is "good" or that the merge information makes sense.
The parents do not have to actually have any relationship with the
result, for example.
 
Note on commits: unlike real SCM's, commits do not contain
rename information or file mode change information.  All of that is
implicit in the trees involved (the result tree, and the result trees
of the parents), and describing that makes no sense in this idiotic
file manager.
 
A commit is created with link:git-commit-tree.html[git-commit-tree] and
its data can be accessed by link:git-cat-file.html[git-cat-file].
 
Trust
~~~~~
An aside on the notion of "trust". Trust is really outside the scope
of "git", but it's worth noting a few things.  First off, since
everything is hashed with SHA1, you 'can' trust that an object is
intact and has not been messed with by external sources.  So the name
of an object uniquely identifies a known state - just not a state that
you may want to trust.
 
Furthermore, since the SHA1 signature of a commit refers to the
SHA1 signatures of the tree it is associated with and the signatures
of the parent, a single named commit specifies uniquely a whole set
of history, with full contents.  You can't later fake any step of the
way once you have the name of a commit.
 
So to introduce some real trust in the system, the only thing you need
to do is to digitally sign just 'one' special note, which includes the
name of a top-level commit.  Your digital signature shows others
that you trust that commit, and the immutability of the history of
commits tells others that they can trust the whole history.
 
In other words, you can easily validate a whole archive by just
sending out a single email that tells the people the name (SHA1 hash)
of the top commit, and digitally sign that email using something
like GPG/PGP.
 
To assist in this, git also provides the tag object...
 
Tag Object
~~~~~~~~~~
Git provides the "tag" object to simplify creating, managing and
exchanging symbolic and signed tokens.  The "tag" object at its
simplest simply symbolically identifies another object by containing
the sha1, type and symbolic name.
 
However it can optionally contain additional signature information
(which git doesn't care about as long as there's less than 8k of
it). This can then be verified externally to git.
 
Note that despite the tag features, "git" itself only handles content
integrity; the trust framework (and signature provision and
verification) has to come from outside.
 
A tag is created with link:git-mktag.html[git-mktag],
its data can be accessed by link:git-cat-file.html[git-cat-file],
and the signature can be verified by
link:git-verify-tag.html[git-verify-tag].
 
 
The "index" aka "Current Directory Cache"
-----------------------------------------
The index is a simple binary file, which contains an efficient
representation of a virtual directory content at some random time.  It
does so by a simple array that associates a set of names, dates,
permissions and content (aka "blob") objects together.  The cache is
always kept ordered by name, and names are unique (with a few very
specific rules) at any point in time, but the cache has no long-term
meaning, and can be partially updated at any time.
 
In particular, the index certainly does not need to be consistent with
the current directory contents (in fact, most operations will depend on
different ways to make the index 'not' be consistent with the directory
hierarchy), but it has three very important attributes:
 
'(a) it can re-generate the full state it caches (not just the
directory structure: it contains pointers to the "blob" objects so
that it can regenerate the data too)'
 
As a special case, there is a clear and unambiguous one-way mapping
from a current directory cache to a "tree object", which can be
efficiently created from just the current directory cache without
actually looking at any other data.  So a directory cache at any one
time uniquely specifies one and only one "tree" object (but has
additional data to make it easy to match up that tree object with what
has happened in the directory)
 
'(b) it has efficient methods for finding inconsistencies between that
cached state ("tree object waiting to be instantiated") and the
current state.'
 
'(c) it can additionally efficiently represent information about merge
conflicts between different tree objects, allowing each pathname to be
associated with sufficient information about the trees involved that
you can create a three-way merge between them.'
 
Those are the three ONLY things that the directory cache does.  It's a
cache, and the normal operation is to re-generate it completely from a
known tree object, or update/compare it with a live tree that is being
developed.  If you blow the directory cache away entirely, you generally
haven't lost any information as long as you have the name of the tree
that it described. 
 
At the same time, the index is at the same time also the
staging area for creating new trees, and creating a new tree always
involves a controlled modification of the index file.  In particular,
the index file can have the representation of an intermediate tree that
has not yet been instantiated.  So the index can be thought of as a
write-back cache, which can contain dirty information that has not yet
been written back to the backing store.
 
 
 
The Workflow
------------
Generally, all "git" operations work on the index file. Some operations
work *purely* on the index file (showing the current state of the
index), but most operations move data to and from the index file. Either
from the database or from the working directory. Thus there are four
main combinations: 
 
1) working directory -> index
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
You update the index with information from the working directory with
the link:git-update-index.html[git-update-index] command.  You
generally update the index information by just specifying the filename
you want to update, like so:
 
	git-update-index filename
 
but to avoid common mistakes with filename globbing etc, the command
will not normally add totally new entries or remove old entries,
i.e. it will normally just update existing cache entries.
 
To tell git that yes, you really do realize that certain files no
longer exist in the archive, or that new files should be added, you
should use the `--remove` and `--add` flags respectively.
 
NOTE! A `--remove` flag does 'not' mean that subsequent filenames will
necessarily be removed: if the files still exist in your directory
structure, the index will be updated with their new status, not
removed. The only thing `--remove` means is that update-cache will be
considering a removed file to be a valid thing, and if the file really
does not exist any more, it will update the index accordingly.
 
As a special case, you can also do `git-update-index --refresh`, which
will refresh the "stat" information of each index to match the current
stat information. It will 'not' update the object status itself, and
it will only update the fields that are used to quickly test whether
an object still matches its old backing store object.
 
2) index -> object database
~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
You write your current index file to a "tree" object with the program
 
	git-write-tree
 
that doesn't come with any options - it will just write out the
current index into the set of tree objects that describe that state,
and it will return the name of the resulting top-level tree. You can
use that tree to re-generate the index at any time by going in the
other direction:
 
3) object database -> index
~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
You read a "tree" file from the object database, and use that to
populate (and overwrite - don't do this if your index contains any
unsaved state that you might want to restore later!) your current
index.  Normal operation is just
 
		git-read-tree <sha1 of tree>
 
and your index file will now be equivalent to the tree that you saved
earlier. However, that is only your 'index' file: your working
directory contents have not been modified.
 
4) index -> working directory
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
You update your working directory from the index by "checking out"
files. This is not a very common operation, since normally you'd just
keep your files updated, and rather than write to your working
directory, you'd tell the index files about the changes in your
working directory (i.e. `git-update-index`).
 
However, if you decide to jump to a new version, or check out somebody
else's version, or just restore a previous tree, you'd populate your
index file with read-tree, and then you need to check out the result
with
 
		git-checkout-index filename
 
or, if you want to check out all of the index, use `-a`.
 
NOTE! git-checkout-index normally refuses to overwrite old files, so
if you have an old version of the tree already checked out, you will
need to use the "-f" flag ('before' the "-a" flag or the filename) to
'force' the checkout.
 
 
Finally, there are a few odds and ends which are not purely moving
from one representation to the other:
 
5) Tying it all together
~~~~~~~~~~~~~~~~~~~~~~~~
To commit a tree you have instantiated with "git-write-tree", you'd
create a "commit" object that refers to that tree and the history
behind it - most notably the "parent" commits that preceded it in
history.
 
Normally a "commit" has one parent: the previous state of the tree
before a certain change was made. However, sometimes it can have two
or more parent commits, in which case we call it a "merge", due to the
fact that such a commit brings together ("merges") two or more
previous states represented by other commits.
 
In other words, while a "tree" represents a particular directory state
of a working directory, a "commit" represents that state in "time",
and explains how we got there.
 
You create a commit object by giving it the tree that describes the
state at the time of the commit, and a list of parents:
 
	git-commit-tree <tree> -p <parent> [-p <parent2> ..]
 
and then giving the reason for the commit on stdin (either through
redirection from a pipe or file, or by just typing it at the tty).
 
git-commit-tree will return the name of the object that represents
that commit, and you should save it away for later use. Normally,
you'd commit a new `HEAD` state, and while git doesn't care where you
save the note about that state, in practice we tend to just write the
result to the file `.git/HEAD`, so that we can always see what the
last committed state was.
 
6) Examining the data
~~~~~~~~~~~~~~~~~~~~~
 
You can examine the data represented in the object database and the
index with various helper tools. For every object, you can use
link:git-cat-file.html[git-cat-file] to examine details about the
object:
 
		git-cat-file -t <objectname>
 
shows the type of the object, and once you have the type (which is
usually implicit in where you find the object), you can use
 
		git-cat-file blob|tree|commit|tag <objectname>
 
to show its contents. NOTE! Trees have binary content, and as a result
there is a special helper for showing that content, called
`git-ls-tree`, which turns the binary content into a more easily
readable form.
 
It's especially instructive to look at "commit" objects, since those
tend to be small and fairly self-explanatory. In particular, if you
follow the convention of having the top commit name in `.git/HEAD`,
you can do
 
		git-cat-file commit $(cat .git/HEAD)
 
to see what the top commit was.
 
7) Merging multiple trees
~~~~~~~~~~~~~~~~~~~~~~~~~
 
Git helps you do a three-way merge, which you can expand to n-way by
repeating the merge procedure arbitrary times until you finally
"commit" the state.  The normal situation is that you'd only do one
three-way merge (two parents), and commit it, but if you like to, you
can do multiple parents in one go.
 
To do a three-way merge, you need the two sets of "commit" objects
that you want to merge, use those to find the closest common parent (a
third "commit" object), and then use those commit objects to find the
state of the directory ("tree" object) at these points.
 
To get the "base" for the merge, you first look up the common parent
of two commits with
 
		git-merge-base <commit1> <commit2>
 
which will return you the commit they are both based on.  You should
now look up the "tree" objects of those commits, which you can easily
do with (for example)
 
		git-cat-file commit <commitname> | head -1
 
since the tree object information is always the first line in a commit
object.
 
Once you know the three trees you are going to merge (the one
"original" tree, aka the common case, and the two "result" trees, aka
the branches you want to merge), you do a "merge" read into the
index. This will complain if it has to throw away your old index contents, so you should
make sure that you've committed those - in fact you would normally
always do a merge against your last commit (which should thus match
what you have in your current index anyway).
 
To do the merge, do
 
		git-read-tree -m -u <origtree> <yourtree> <targettree>
 
which will do all trivial merge operations for you directly in the
index file, and you can just write the result out with
`git-write-tree`.
 
Historical note.  We did not have `-u` facility when this
section was first written, so we used to warn that
the merge is done in the index file, not in your
working directory, and your working directory will no longer match your
index.
 
 
8) Merging multiple trees, continued
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
Sadly, many merges aren't trivial. If there are files that have
been added.moved or removed, or if both branches have modified the
same file, you will be left with an index tree that contains "merge
entries" in it. Such an index tree can 'NOT' be written out to a tree
object, and you will have to resolve any such merge clashes using
other tools before you can write out the result.
 
You can examine such index state with `git-ls-files --unmerged`
command.  An example:
 
------------------------------------------------
$ git-read-tree -m $orig HEAD $target
$ git-ls-files --unmerged
100644 263414f423d0e4d70dae8fe53fa34614ff3e2860 1	hello.c
100644 06fa6a24256dc7e560efa5687fa84b51f0263c3a 2	hello.c
100644 cc44c73eb783565da5831b4d820c962954019b69 3	hello.c
------------------------------------------------
 
Each line of the `git-ls-files --unmerged` output begins with
the blob mode bits, blob SHA1, 'stage number', and the
filename.  The 'stage number' is git's way to say which tree it
came from: stage 1 corresponds to `$orig` tree, stage 2 `HEAD`
tree, and stage3 `$target` tree.
 
Earlier we said that trivial merges are done inside
`git-read-tree -m`.  For example, if the file did not change
from `$orig` to `HEAD` nor `$target`, or if the file changed
from `$orig` to `HEAD` and `$orig` to `$target` the same way,
obviously the final outcome is what is in `HEAD`.  What the
above example shows is that file `hello.c` was changed from
`$orig` to `HEAD` and `$orig` to `$target` in a different way.
You could resolve this by running your favorite 3-way merge
program, e.g.  `diff3` or `merge`, on the blob objects from
these three stages yourself, like this:
 
------------------------------------------------
$ git-cat-file blob 263414f... >hello.c~1
$ git-cat-file blob 06fa6a2... >hello.c~2
$ git-cat-file blob cc44c73... >hello.c~3
$ merge hello.c~2 hello.c~1 hello.c~3
------------------------------------------------
 
This would leave the merge result in `hello.c~2` file, along
with conflict markers if there are conflicts.  After verifying
the merge result makes sense, you can tell git what the final
merge result for this file is by:
 
	mv -f hello.c~2 hello.c
	git-update-index hello.c
 
When a path is in unmerged state, running `git-update-index` for
that path tells git to mark the path resolved.
 
The above is the description of a git merge at the lowest level,
to help you understand what conceptually happens under the hood.
In practice, nobody, not even git itself, uses three `git-cat-file`
for this.  There is `git-merge-index` program that extracts the
stages to temporary files and calls a `merge` script on it
 
	git-merge-index git-merge-one-file hello.c
 
and that is what higher level `git resolve` is implemented with.