Documentation / gittutorial.txton commit use -h for synopsis and --help for manpage consistently (87182b1)
   1gittutorial(7)
   2==============
   3
   4NAME
   5----
   6gittutorial - A tutorial introduction to git (for version 1.5.1 or newer)
   7
   8SYNOPSIS
   9--------
  10[verse]
  11git *
  12
  13DESCRIPTION
  14-----------
  15
  16This tutorial explains how to import a new project into git, make
  17changes to it, and share changes with other developers.
  18
  19If you are instead primarily interested in using git to fetch a project,
  20for example, to test the latest version, you may prefer to start with
  21the first two chapters of link:user-manual.html[The Git User's Manual].
  22
  23First, note that you can get documentation for a command such as
  24`git log --graph` with:
  25
  26------------------------------------------------
  27$ man git-log
  28------------------------------------------------
  29
  30or:
  31
  32------------------------------------------------
  33$ git help log
  34------------------------------------------------
  35
  36With the latter, you can use the manual viewer of your choice; see
  37linkgit:git-help[1] for more information.
  38
  39It is a good idea to introduce yourself to git with your name and
  40public email address before doing any operation.  The easiest
  41way to do so is:
  42
  43------------------------------------------------
  44$ git config --global user.name "Your Name Comes Here"
  45$ git config --global user.email you@yourdomain.example.com
  46------------------------------------------------
  47
  48
  49Importing a new project
  50-----------------------
  51
  52Assume you have a tarball project.tar.gz with your initial work.  You
  53can place it under git revision control as follows.
  54
  55------------------------------------------------
  56$ tar xzf project.tar.gz
  57$ cd project
  58$ git init
  59------------------------------------------------
  60
  61Git will reply
  62
  63------------------------------------------------
  64Initialized empty Git repository in .git/
  65------------------------------------------------
  66
  67You've now initialized the working directory--you may notice a new
  68directory created, named ".git".
  69
  70Next, tell git to take a snapshot of the contents of all files under the
  71current directory (note the '.'), with 'git add':
  72
  73------------------------------------------------
  74$ git add .
  75------------------------------------------------
  76
  77This snapshot is now stored in a temporary staging area which git calls
  78the "index".  You can permanently store the contents of the index in the
  79repository with 'git commit':
  80
  81------------------------------------------------
  82$ git commit
  83------------------------------------------------
  84
  85This will prompt you for a commit message.  You've now stored the first
  86version of your project in git.
  87
  88Making changes
  89--------------
  90
  91Modify some files, then add their updated contents to the index:
  92
  93------------------------------------------------
  94$ git add file1 file2 file3
  95------------------------------------------------
  96
  97You are now ready to commit.  You can see what is about to be committed
  98using 'git diff' with the --cached option:
  99
 100------------------------------------------------
 101$ git diff --cached
 102------------------------------------------------
 103
 104(Without --cached, 'git diff' will show you any changes that
 105you've made but not yet added to the index.)  You can also get a brief
 106summary of the situation with 'git status':
 107
 108------------------------------------------------
 109$ git status
 110# On branch master
 111# Changes to be committed:
 112#   (use "git reset HEAD <file>..." to unstage)
 113#
 114#       modified:   file1
 115#       modified:   file2
 116#       modified:   file3
 117#
 118------------------------------------------------
 119
 120If you need to make any further adjustments, do so now, and then add any
 121newly modified content to the index.  Finally, commit your changes with:
 122
 123------------------------------------------------
 124$ git commit
 125------------------------------------------------
 126
 127This will again prompt you for a message describing the change, and then
 128record a new version of the project.
 129
 130Alternatively, instead of running 'git add' beforehand, you can use
 131
 132------------------------------------------------
 133$ git commit -a
 134------------------------------------------------
 135
 136which will automatically notice any modified (but not new) files, add
 137them to the index, and commit, all in one step.
 138
 139A note on commit messages: Though not required, it's a good idea to
 140begin the commit message with a single short (less than 50 character)
 141line summarizing the change, followed by a blank line and then a more
 142thorough description.  Tools that turn commits into email, for
 143example, use the first line on the Subject: line and the rest of the
 144commit in the body.
 145
 146Git tracks content not files
 147----------------------------
 148
 149Many revision control systems provide an `add` command that tells the
 150system to start tracking changes to a new file.  Git's `add` command
 151does something simpler and more powerful: 'git add' is used both for new
 152and newly modified files, and in both cases it takes a snapshot of the
 153given files and stages that content in the index, ready for inclusion in
 154the next commit.
 155
 156Viewing project history
 157-----------------------
 158
 159At any point you can view the history of your changes using
 160
 161------------------------------------------------
 162$ git log
 163------------------------------------------------
 164
 165If you also want to see complete diffs at each step, use
 166
 167------------------------------------------------
 168$ git log -p
 169------------------------------------------------
 170
 171Often the overview of the change is useful to get a feel of
 172each step
 173
 174------------------------------------------------
 175$ git log --stat --summary
 176------------------------------------------------
 177
 178Managing branches
 179-----------------
 180
 181A single git repository can maintain multiple branches of
 182development.  To create a new branch named "experimental", use
 183
 184------------------------------------------------
 185$ git branch experimental
 186------------------------------------------------
 187
 188If you now run
 189
 190------------------------------------------------
 191$ git branch
 192------------------------------------------------
 193
 194you'll get a list of all existing branches:
 195
 196------------------------------------------------
 197  experimental
 198* master
 199------------------------------------------------
 200
 201The "experimental" branch is the one you just created, and the
 202"master" branch is a default branch that was created for you
 203automatically.  The asterisk marks the branch you are currently on;
 204type
 205
 206------------------------------------------------
 207$ git checkout experimental
 208------------------------------------------------
 209
 210to switch to the experimental branch.  Now edit a file, commit the
 211change, and switch back to the master branch:
 212
 213------------------------------------------------
 214(edit file)
 215$ git commit -a
 216$ git checkout master
 217------------------------------------------------
 218
 219Check that the change you made is no longer visible, since it was
 220made on the experimental branch and you're back on the master branch.
 221
 222You can make a different change on the master branch:
 223
 224------------------------------------------------
 225(edit file)
 226$ git commit -a
 227------------------------------------------------
 228
 229at this point the two branches have diverged, with different changes
 230made in each.  To merge the changes made in experimental into master, run
 231
 232------------------------------------------------
 233$ git merge experimental
 234------------------------------------------------
 235
 236If the changes don't conflict, you're done.  If there are conflicts,
 237markers will be left in the problematic files showing the conflict;
 238
 239------------------------------------------------
 240$ git diff
 241------------------------------------------------
 242
 243will show this.  Once you've edited the files to resolve the
 244conflicts,
 245
 246------------------------------------------------
 247$ git commit -a
 248------------------------------------------------
 249
 250will commit the result of the merge. Finally,
 251
 252------------------------------------------------
 253$ gitk
 254------------------------------------------------
 255
 256will show a nice graphical representation of the resulting history.
 257
 258At this point you could delete the experimental branch with
 259
 260------------------------------------------------
 261$ git branch -d experimental
 262------------------------------------------------
 263
 264This command ensures that the changes in the experimental branch are
 265already in the current branch.
 266
 267If you develop on a branch crazy-idea, then regret it, you can always
 268delete the branch with
 269
 270-------------------------------------
 271$ git branch -D crazy-idea
 272-------------------------------------
 273
 274Branches are cheap and easy, so this is a good way to try something
 275out.
 276
 277Using git for collaboration
 278---------------------------
 279
 280Suppose that Alice has started a new project with a git repository in
 281/home/alice/project, and that Bob, who has a home directory on the
 282same machine, wants to contribute.
 283
 284Bob begins with:
 285
 286------------------------------------------------
 287bob$ git clone /home/alice/project myrepo
 288------------------------------------------------
 289
 290This creates a new directory "myrepo" containing a clone of Alice's
 291repository.  The clone is on an equal footing with the original
 292project, possessing its own copy of the original project's history.
 293
 294Bob then makes some changes and commits them:
 295
 296------------------------------------------------
 297(edit files)
 298bob$ git commit -a
 299(repeat as necessary)
 300------------------------------------------------
 301
 302When he's ready, he tells Alice to pull changes from the repository
 303at /home/bob/myrepo.  She does this with:
 304
 305------------------------------------------------
 306alice$ cd /home/alice/project
 307alice$ git pull /home/bob/myrepo master
 308------------------------------------------------
 309
 310This merges the changes from Bob's "master" branch into Alice's
 311current branch.  If Alice has made her own changes in the meantime,
 312then she may need to manually fix any conflicts.
 313
 314The "pull" command thus performs two operations: it fetches changes
 315from a remote branch, then merges them into the current branch.
 316
 317Note that in general, Alice would want her local changes committed before
 318initiating this "pull".  If Bob's work conflicts with what Alice did since
 319their histories forked, Alice will use her working tree and the index to
 320resolve conflicts, and existing local changes will interfere with the
 321conflict resolution process (git will still perform the fetch but will
 322refuse to merge --- Alice will have to get rid of her local changes in
 323some way and pull again when this happens).
 324
 325Alice can peek at what Bob did without merging first, using the "fetch"
 326command; this allows Alice to inspect what Bob did, using a special
 327symbol "FETCH_HEAD", in order to determine if he has anything worth
 328pulling, like this:
 329
 330------------------------------------------------
 331alice$ git fetch /home/bob/myrepo master
 332alice$ git log -p HEAD..FETCH_HEAD
 333------------------------------------------------
 334
 335This operation is safe even if Alice has uncommitted local changes.
 336The range notation "HEAD..FETCH_HEAD" means "show everything that is reachable
 337from the FETCH_HEAD but exclude anything that is reachable from HEAD".
 338Alice already knows everything that leads to her current state (HEAD),
 339and reviews what Bob has in his state (FETCH_HEAD) that she has not
 340seen with this command.
 341
 342If Alice wants to visualize what Bob did since their histories forked
 343she can issue the following command:
 344
 345------------------------------------------------
 346$ gitk HEAD..FETCH_HEAD
 347------------------------------------------------
 348
 349This uses the same two-dot range notation we saw earlier with 'git log'.
 350
 351Alice may want to view what both of them did since they forked.
 352She can use three-dot form instead of the two-dot form:
 353
 354------------------------------------------------
 355$ gitk HEAD...FETCH_HEAD
 356------------------------------------------------
 357
 358This means "show everything that is reachable from either one, but
 359exclude anything that is reachable from both of them".
 360
 361Please note that these range notation can be used with both gitk
 362and "git log".
 363
 364After inspecting what Bob did, if there is nothing urgent, Alice may
 365decide to continue working without pulling from Bob.  If Bob's history
 366does have something Alice would immediately need, Alice may choose to
 367stash her work-in-progress first, do a "pull", and then finally unstash
 368her work-in-progress on top of the resulting history.
 369
 370When you are working in a small closely knit group, it is not
 371unusual to interact with the same repository over and over
 372again.  By defining 'remote' repository shorthand, you can make
 373it easier:
 374
 375------------------------------------------------
 376alice$ git remote add bob /home/bob/myrepo
 377------------------------------------------------
 378
 379With this, Alice can perform the first part of the "pull" operation
 380alone using the 'git fetch' command without merging them with her own
 381branch, using:
 382
 383-------------------------------------
 384alice$ git fetch bob
 385-------------------------------------
 386
 387Unlike the longhand form, when Alice fetches from Bob using a
 388remote repository shorthand set up with 'git remote', what was
 389fetched is stored in a remote-tracking branch, in this case
 390`bob/master`.  So after this:
 391
 392-------------------------------------
 393alice$ git log -p master..bob/master
 394-------------------------------------
 395
 396shows a list of all the changes that Bob made since he branched from
 397Alice's master branch.
 398
 399After examining those changes, Alice
 400could merge the changes into her master branch:
 401
 402-------------------------------------
 403alice$ git merge bob/master
 404-------------------------------------
 405
 406This `merge` can also be done by 'pulling from her own remote-tracking
 407branch', like this:
 408
 409-------------------------------------
 410alice$ git pull . remotes/bob/master
 411-------------------------------------
 412
 413Note that git pull always merges into the current branch,
 414regardless of what else is given on the command line.
 415
 416Later, Bob can update his repo with Alice's latest changes using
 417
 418-------------------------------------
 419bob$ git pull
 420-------------------------------------
 421
 422Note that he doesn't need to give the path to Alice's repository;
 423when Bob cloned Alice's repository, git stored the location of her
 424repository in the repository configuration, and that location is
 425used for pulls:
 426
 427-------------------------------------
 428bob$ git config --get remote.origin.url
 429/home/alice/project
 430-------------------------------------
 431
 432(The complete configuration created by 'git clone' is visible using
 433`git config -l`, and the linkgit:git-config[1] man page
 434explains the meaning of each option.)
 435
 436Git also keeps a pristine copy of Alice's master branch under the
 437name "origin/master":
 438
 439-------------------------------------
 440bob$ git branch -r
 441  origin/master
 442-------------------------------------
 443
 444If Bob later decides to work from a different host, he can still
 445perform clones and pulls using the ssh protocol:
 446
 447-------------------------------------
 448bob$ git clone alice.org:/home/alice/project myrepo
 449-------------------------------------
 450
 451Alternatively, git has a native protocol, or can use rsync or http;
 452see linkgit:git-pull[1] for details.
 453
 454Git can also be used in a CVS-like mode, with a central repository
 455that various users push changes to; see linkgit:git-push[1] and
 456linkgit:gitcvs-migration[7].
 457
 458Exploring history
 459-----------------
 460
 461Git history is represented as a series of interrelated commits.  We
 462have already seen that the 'git log' command can list those commits.
 463Note that first line of each git log entry also gives a name for the
 464commit:
 465
 466-------------------------------------
 467$ git log
 468commit c82a22c39cbc32576f64f5c6b3f24b99ea8149c7
 469Author: Junio C Hamano <junkio@cox.net>
 470Date:   Tue May 16 17:18:22 2006 -0700
 471
 472    merge-base: Clarify the comments on post processing.
 473-------------------------------------
 474
 475We can give this name to 'git show' to see the details about this
 476commit.
 477
 478-------------------------------------
 479$ git show c82a22c39cbc32576f64f5c6b3f24b99ea8149c7
 480-------------------------------------
 481
 482But there are other ways to refer to commits.  You can use any initial
 483part of the name that is long enough to uniquely identify the commit:
 484
 485-------------------------------------
 486$ git show c82a22c39c   # the first few characters of the name are
 487                        # usually enough
 488$ git show HEAD         # the tip of the current branch
 489$ git show experimental # the tip of the "experimental" branch
 490-------------------------------------
 491
 492Every commit usually has one "parent" commit
 493which points to the previous state of the project:
 494
 495-------------------------------------
 496$ git show HEAD^  # to see the parent of HEAD
 497$ git show HEAD^^ # to see the grandparent of HEAD
 498$ git show HEAD~4 # to see the great-great grandparent of HEAD
 499-------------------------------------
 500
 501Note that merge commits may have more than one parent:
 502
 503-------------------------------------
 504$ git show HEAD^1 # show the first parent of HEAD (same as HEAD^)
 505$ git show HEAD^2 # show the second parent of HEAD
 506-------------------------------------
 507
 508You can also give commits names of your own; after running
 509
 510-------------------------------------
 511$ git tag v2.5 1b2e1d63ff
 512-------------------------------------
 513
 514you can refer to 1b2e1d63ff by the name "v2.5".  If you intend to
 515share this name with other people (for example, to identify a release
 516version), you should create a "tag" object, and perhaps sign it; see
 517linkgit:git-tag[1] for details.
 518
 519Any git command that needs to know a commit can take any of these
 520names.  For example:
 521
 522-------------------------------------
 523$ git diff v2.5 HEAD     # compare the current HEAD to v2.5
 524$ git branch stable v2.5 # start a new branch named "stable" based
 525                         # at v2.5
 526$ git reset --hard HEAD^ # reset your current branch and working
 527                         # directory to its state at HEAD^
 528-------------------------------------
 529
 530Be careful with that last command: in addition to losing any changes
 531in the working directory, it will also remove all later commits from
 532this branch.  If this branch is the only branch containing those
 533commits, they will be lost.  Also, don't use 'git reset' on a
 534publicly-visible branch that other developers pull from, as it will
 535force needless merges on other developers to clean up the history.
 536If you need to undo changes that you have pushed, use 'git revert'
 537instead.
 538
 539The 'git grep' command can search for strings in any version of your
 540project, so
 541
 542-------------------------------------
 543$ git grep "hello" v2.5
 544-------------------------------------
 545
 546searches for all occurrences of "hello" in v2.5.
 547
 548If you leave out the commit name, 'git grep' will search any of the
 549files it manages in your current directory.  So
 550
 551-------------------------------------
 552$ git grep "hello"
 553-------------------------------------
 554
 555is a quick way to search just the files that are tracked by git.
 556
 557Many git commands also take sets of commits, which can be specified
 558in a number of ways.  Here are some examples with 'git log':
 559
 560-------------------------------------
 561$ git log v2.5..v2.6            # commits between v2.5 and v2.6
 562$ git log v2.5..                # commits since v2.5
 563$ git log --since="2 weeks ago" # commits from the last 2 weeks
 564$ git log v2.5.. Makefile       # commits since v2.5 which modify
 565                                # Makefile
 566-------------------------------------
 567
 568You can also give 'git log' a "range" of commits where the first is not
 569necessarily an ancestor of the second; for example, if the tips of
 570the branches "stable" and "master" diverged from a common
 571commit some time ago, then
 572
 573-------------------------------------
 574$ git log stable..master
 575-------------------------------------
 576
 577will list commits made in the master branch but not in the
 578stable branch, while
 579
 580-------------------------------------
 581$ git log master..stable
 582-------------------------------------
 583
 584will show the list of commits made on the stable branch but not
 585the master branch.
 586
 587The 'git log' command has a weakness: it must present commits in a
 588list.  When the history has lines of development that diverged and
 589then merged back together, the order in which 'git log' presents
 590those commits is meaningless.
 591
 592Most projects with multiple contributors (such as the Linux kernel,
 593or git itself) have frequent merges, and 'gitk' does a better job of
 594visualizing their history.  For example,
 595
 596-------------------------------------
 597$ gitk --since="2 weeks ago" drivers/
 598-------------------------------------
 599
 600allows you to browse any commits from the last 2 weeks of commits
 601that modified files under the "drivers" directory.  (Note: you can
 602adjust gitk's fonts by holding down the control key while pressing
 603"-" or "+".)
 604
 605Finally, most commands that take filenames will optionally allow you
 606to precede any filename by a commit, to specify a particular version
 607of the file:
 608
 609-------------------------------------
 610$ git diff v2.5:Makefile HEAD:Makefile.in
 611-------------------------------------
 612
 613You can also use 'git show' to see any such file:
 614
 615-------------------------------------
 616$ git show v2.5:Makefile
 617-------------------------------------
 618
 619Next Steps
 620----------
 621
 622This tutorial should be enough to perform basic distributed revision
 623control for your projects.  However, to fully understand the depth
 624and power of git you need to understand two simple ideas on which it
 625is based:
 626
 627  * The object database is the rather elegant system used to
 628    store the history of your project--files, directories, and
 629    commits.
 630
 631  * The index file is a cache of the state of a directory tree,
 632    used to create commits, check out working directories, and
 633    hold the various trees involved in a merge.
 634
 635Part two of this tutorial explains the object
 636database, the index file, and a few other odds and ends that you'll
 637need to make the most of git. You can find it at linkgit:gittutorial-2[7].
 638
 639If you don't want to continue with that right away, a few other
 640digressions that may be interesting at this point are:
 641
 642  * linkgit:git-format-patch[1], linkgit:git-am[1]: These convert
 643    series of git commits into emailed patches, and vice versa,
 644    useful for projects such as the Linux kernel which rely heavily
 645    on emailed patches.
 646
 647  * linkgit:git-bisect[1]: When there is a regression in your
 648    project, one way to track down the bug is by searching through
 649    the history to find the exact commit that's to blame.  Git bisect
 650    can help you perform a binary search for that commit.  It is
 651    smart enough to perform a close-to-optimal search even in the
 652    case of complex non-linear history with lots of merged branches.
 653
 654  * linkgit:gitworkflows[7]: Gives an overview of recommended
 655    workflows.
 656
 657  * link:everyday.html[Everyday GIT with 20 Commands Or So]
 658
 659  * linkgit:gitcvs-migration[7]: Git for CVS users.
 660
 661SEE ALSO
 662--------
 663linkgit:gittutorial-2[7],
 664linkgit:gitcvs-migration[7],
 665linkgit:gitcore-tutorial[7],
 666linkgit:gitglossary[7],
 667linkgit:git-help[1],
 668linkgit:gitworkflows[7],
 669link:everyday.html[Everyday git],
 670link:user-manual.html[The Git User's Manual]
 671
 672GIT
 673---
 674Part of the linkgit:git[1] suite.