1gittutorial-2(7) 2================ 3 4NAME 5---- 6gittutorial-2 - A tutorial introduction to git: part two 7 8SYNOPSIS 9-------- 10git * 11 12DESCRIPTION 13----------- 14 15You should work through linkgit:gittutorial[7] before reading this tutorial. 16 17The goal of this tutorial is to introduce two fundamental pieces of 18git's architecture--the object database and the index file--and to 19provide the reader with everything necessary to understand the rest 20of the git documentation. 21 22The git object database 23----------------------- 24 25Let's start a new project and create a small amount of history: 26 27------------------------------------------------ 28$ mkdir test-project 29$ cd test-project 30$ git init 31Initialized empty Git repository in .git/ 32$ echo 'hello world' > file.txt 33$ git add . 34$ git commit -a -m "initial commit" 35Created root-commit 54196cc (initial commit) on master 36 create mode 100644 file.txt 37$ echo 'hello world!' >file.txt 38$ git commit -a -m "add emphasis" 39Created c4d59f3 (add emphasis) on master 40------------------------------------------------ 41 42What are the 7 digits of hex that git responded to the commit with? 43 44We saw in part one of the tutorial that commits have names like this. 45It turns out that every object in the git history is stored under 46a 40-digit hex name. That name is the SHA1 hash of the object's 47contents; among other things, this ensures that git will never store 48the same data twice (since identical data is given an identical SHA1 49name), and that the contents of a git object will never change (since 50that would change the object's name as well). The 7 char hex strings 51here are simply the abbreviation of such 40 character long strings. 52Abbreviations can be used everywhere where the 40 character strings 53can be used, so long as they are unambiguous. 54 55It is expected that the content of the commit object you created while 56following the example above generates a different SHA1 hash than 57the one shown above because the commit object records the time when 58it was created and the name of the person performing the commit. 59 60We can ask git about this particular object with the `cat-file` 61command. Don't copy the 40 hex digits from this example but use those 62from your own version. Note that you can shorten it to only a few 63characters to save yourself typing all 40 hex digits: 64 65------------------------------------------------ 66$ git cat-file -t 54196cc2 67commit 68$ git cat-file commit 54196cc2 69tree 92b8b694ffb1675e5975148e1121810081dbdffe 70author J. Bruce Fields <bfields@puzzle.fieldses.org> 1143414668 -0500 71committer J. Bruce Fields <bfields@puzzle.fieldses.org> 1143414668 -0500 72 73initial commit 74------------------------------------------------ 75 76A tree can refer to one or more "blob" objects, each corresponding to 77a file. In addition, a tree can also refer to other tree objects, 78thus creating a directory hierarchy. You can examine the contents of 79any tree using ls-tree (remember that a long enough initial portion 80of the SHA1 will also work): 81 82------------------------------------------------ 83$ git ls-tree 92b8b694 84100644 blob 3b18e512dba79e4c8300dd08aeb37f8e728b8dad file.txt 85------------------------------------------------ 86 87Thus we see that this tree has one file in it. The SHA1 hash is a 88reference to that file's data: 89 90------------------------------------------------ 91$ git cat-file -t 3b18e512 92blob 93------------------------------------------------ 94 95A "blob" is just file data, which we can also examine with cat-file: 96 97------------------------------------------------ 98$ git cat-file blob 3b18e512 99hello world 100------------------------------------------------ 101 102Note that this is the old file data; so the object that git named in 103its response to the initial tree was a tree with a snapshot of the 104directory state that was recorded by the first commit. 105 106All of these objects are stored under their SHA1 names inside the git 107directory: 108 109------------------------------------------------ 110$ find .git/objects/ 111.git/objects/ 112.git/objects/pack 113.git/objects/info 114.git/objects/3b 115.git/objects/3b/18e512dba79e4c8300dd08aeb37f8e728b8dad 116.git/objects/92 117.git/objects/92/b8b694ffb1675e5975148e1121810081dbdffe 118.git/objects/54 119.git/objects/54/196cc2703dc165cbd373a65a4dcf22d50ae7f7 120.git/objects/a0 121.git/objects/a0/423896973644771497bdc03eb99d5281615b51 122.git/objects/d0 123.git/objects/d0/492b368b66bdabf2ac1fd8c92b39d3db916e59 124.git/objects/c4 125.git/objects/c4/d59f390b9cfd4318117afde11d601c1085f241 126------------------------------------------------ 127 128and the contents of these files is just the compressed data plus a 129header identifying their length and their type. The type is either a 130blob, a tree, a commit, or a tag. 131 132The simplest commit to find is the HEAD commit, which we can find 133from .git/HEAD: 134 135------------------------------------------------ 136$ cat .git/HEAD 137ref: refs/heads/master 138------------------------------------------------ 139 140As you can see, this tells us which branch we're currently on, and it 141tells us this by naming a file under the .git directory, which itself 142contains a SHA1 name referring to a commit object, which we can 143examine with cat-file: 144 145------------------------------------------------ 146$ cat .git/refs/heads/master 147c4d59f390b9cfd4318117afde11d601c1085f241 148$ git cat-file -t c4d59f39 149commit 150$ git cat-file commit c4d59f39 151tree d0492b368b66bdabf2ac1fd8c92b39d3db916e59 152parent 54196cc2703dc165cbd373a65a4dcf22d50ae7f7 153author J. Bruce Fields <bfields@puzzle.fieldses.org> 1143418702 -0500 154committer J. Bruce Fields <bfields@puzzle.fieldses.org> 1143418702 -0500 155 156add emphasis 157------------------------------------------------ 158 159The "tree" object here refers to the new state of the tree: 160 161------------------------------------------------ 162$ git ls-tree d0492b36 163100644 blob a0423896973644771497bdc03eb99d5281615b51 file.txt 164$ git cat-file blob a0423896 165hello world! 166------------------------------------------------ 167 168and the "parent" object refers to the previous commit: 169 170------------------------------------------------ 171$ git cat-file commit 54196cc2 172tree 92b8b694ffb1675e5975148e1121810081dbdffe 173author J. Bruce Fields <bfields@puzzle.fieldses.org> 1143414668 -0500 174committer J. Bruce Fields <bfields@puzzle.fieldses.org> 1143414668 -0500 175 176initial commit 177------------------------------------------------ 178 179The tree object is the tree we examined first, and this commit is 180unusual in that it lacks any parent. 181 182Most commits have only one parent, but it is also common for a commit 183to have multiple parents. In that case the commit represents a 184merge, with the parent references pointing to the heads of the merged 185branches. 186 187Besides blobs, trees, and commits, the only remaining type of object 188is a "tag", which we won't discuss here; refer to linkgit:git-tag[1] 189for details. 190 191So now we know how git uses the object database to represent a 192project's history: 193 194 * "commit" objects refer to "tree" objects representing the 195 snapshot of a directory tree at a particular point in the 196 history, and refer to "parent" commits to show how they're 197 connected into the project history. 198 * "tree" objects represent the state of a single directory, 199 associating directory names to "blob" objects containing file 200 data and "tree" objects containing subdirectory information. 201 * "blob" objects contain file data without any other structure. 202 * References to commit objects at the head of each branch are 203 stored in files under .git/refs/heads/. 204 * The name of the current branch is stored in .git/HEAD. 205 206Note, by the way, that lots of commands take a tree as an argument. 207But as we can see above, a tree can be referred to in many different 208ways--by the SHA1 name for that tree, by the name of a commit that 209refers to the tree, by the name of a branch whose head refers to that 210tree, etc.--and most such commands can accept any of these names. 211 212In command synopses, the word "tree-ish" is sometimes used to 213designate such an argument. 214 215The index file 216-------------- 217 218The primary tool we've been using to create commits is `git-commit 219-a`, which creates a commit including every change you've made to 220your working tree. But what if you want to commit changes only to 221certain files? Or only certain changes to certain files? 222 223If we look at the way commits are created under the cover, we'll see 224that there are more flexible ways creating commits. 225 226Continuing with our test-project, let's modify file.txt again: 227 228------------------------------------------------ 229$ echo "hello world, again" >>file.txt 230------------------------------------------------ 231 232but this time instead of immediately making the commit, let's take an 233intermediate step, and ask for diffs along the way to keep track of 234what's happening: 235 236------------------------------------------------ 237$ git diff 238--- a/file.txt 239+++ b/file.txt 240@@ -1 +1,2 @@ 241 hello world! 242+hello world, again 243$ git add file.txt 244$ git diff 245------------------------------------------------ 246 247The last diff is empty, but no new commits have been made, and the 248head still doesn't contain the new line: 249 250------------------------------------------------ 251$ git diff HEAD 252diff --git a/file.txt b/file.txt 253index a042389..513feba 100644 254--- a/file.txt 255+++ b/file.txt 256@@ -1 +1,2 @@ 257 hello world! 258+hello world, again 259------------------------------------------------ 260 261So 'git-diff' is comparing against something other than the head. 262The thing that it's comparing against is actually the index file, 263which is stored in .git/index in a binary format, but whose contents 264we can examine with ls-files: 265 266------------------------------------------------ 267$ git ls-files --stage 268100644 513feba2e53ebbd2532419ded848ba19de88ba00 0 file.txt 269$ git cat-file -t 513feba2 270blob 271$ git cat-file blob 513feba2 272hello world! 273hello world, again 274------------------------------------------------ 275 276So what our 'git-add' did was store a new blob and then put 277a reference to it in the index file. If we modify the file again, 278we'll see that the new modifications are reflected in the 'git-diff' 279output: 280 281------------------------------------------------ 282$ echo 'again?' >>file.txt 283$ git diff 284index 513feba..ba3da7b 100644 285--- a/file.txt 286+++ b/file.txt 287@@ -1,2 +1,3 @@ 288 hello world! 289 hello world, again 290+again? 291------------------------------------------------ 292 293With the right arguments, 'git-diff' can also show us the difference 294between the working directory and the last commit, or between the 295index and the last commit: 296 297------------------------------------------------ 298$ git diff HEAD 299diff --git a/file.txt b/file.txt 300index a042389..ba3da7b 100644 301--- a/file.txt 302+++ b/file.txt 303@@ -1 +1,3 @@ 304 hello world! 305+hello world, again 306+again? 307$ git diff --cached 308diff --git a/file.txt b/file.txt 309index a042389..513feba 100644 310--- a/file.txt 311+++ b/file.txt 312@@ -1 +1,2 @@ 313 hello world! 314+hello world, again 315------------------------------------------------ 316 317At any time, we can create a new commit using 'git-commit' (without 318the "-a" option), and verify that the state committed only includes the 319changes stored in the index file, not the additional change that is 320still only in our working tree: 321 322------------------------------------------------ 323$ git commit -m "repeat" 324$ git diff HEAD 325diff --git a/file.txt b/file.txt 326index 513feba..ba3da7b 100644 327--- a/file.txt 328+++ b/file.txt 329@@ -1,2 +1,3 @@ 330 hello world! 331 hello world, again 332+again? 333------------------------------------------------ 334 335So by default 'git-commit' uses the index to create the commit, not 336the working tree; the "-a" option to commit tells it to first update 337the index with all changes in the working tree. 338 339Finally, it's worth looking at the effect of 'git-add' on the index 340file: 341 342------------------------------------------------ 343$ echo "goodbye, world" >closing.txt 344$ git add closing.txt 345------------------------------------------------ 346 347The effect of the 'git-add' was to add one entry to the index file: 348 349------------------------------------------------ 350$ git ls-files --stage 351100644 8b9743b20d4b15be3955fc8d5cd2b09cd2336138 0 closing.txt 352100644 513feba2e53ebbd2532419ded848ba19de88ba00 0 file.txt 353------------------------------------------------ 354 355And, as you can see with cat-file, this new entry refers to the 356current contents of the file: 357 358------------------------------------------------ 359$ git cat-file blob 8b9743b2 360goodbye, world 361------------------------------------------------ 362 363The "status" command is a useful way to get a quick summary of the 364situation: 365 366------------------------------------------------ 367$ git status 368# On branch master 369# Changes to be committed: 370# (use "git reset HEAD <file>..." to unstage) 371# 372# new file: closing.txt 373# 374# Changed but not updated: 375# (use "git add <file>..." to update what will be committed) 376# 377# modified: file.txt 378# 379------------------------------------------------ 380 381Since the current state of closing.txt is cached in the index file, 382it is listed as "Changes to be committed". Since file.txt has 383changes in the working directory that aren't reflected in the index, 384it is marked "changed but not updated". At this point, running "git 385commit" would create a commit that added closing.txt (with its new 386contents), but that didn't modify file.txt. 387 388Also, note that a bare `git diff` shows the changes to file.txt, but 389not the addition of closing.txt, because the version of closing.txt 390in the index file is identical to the one in the working directory. 391 392In addition to being the staging area for new commits, the index file 393is also populated from the object database when checking out a 394branch, and is used to hold the trees involved in a merge operation. 395See linkgit:gitcore-tutorial[7] and the relevant man 396pages for details. 397 398What next? 399---------- 400 401At this point you should know everything necessary to read the man 402pages for any of the git commands; one good place to start would be 403with the commands mentioned in link:everyday.html[Everyday git]. You 404should be able to find any unknown jargon in linkgit:gitglossary[7]. 405 406The link:user-manual.html[Git User's Manual] provides a more 407comprehensive introduction to git. 408 409linkgit:gitcvs-migration[7] explains how to 410import a CVS repository into git, and shows how to use git in a 411CVS-like way. 412 413For some interesting examples of git use, see the 414link:howto-index.html[howtos]. 415 416For git developers, linkgit:gitcore-tutorial[7] goes 417into detail on the lower-level git mechanisms involved in, for 418example, creating a new commit. 419 420SEE ALSO 421-------- 422linkgit:gittutorial[7], 423linkgit:gitcvs-migration[7], 424linkgit:gitcore-tutorial[7], 425linkgit:gitglossary[7], 426link:everyday.html[Everyday git], 427link:user-manual.html[The Git User's Manual] 428 429GIT 430--- 431Part of the linkgit:git[1] suite.