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" 35[master (root-commit) 54196cc] initial commit 36 1 files changed, 1 insertions(+), 0 deletions(-) 37 create mode 100644 file.txt 38$ echo 'hello world!' >file.txt 39$ git commit -a -m "add emphasis" 40[master c4d59f3] add emphasis 41 1 files changed, 1 insertions(+), 1 deletions(-) 42------------------------------------------------ 43 44What are the 7 digits of hex that git responded to the commit with? 45 46We saw in part one of the tutorial that commits have names like this. 47It turns out that every object in the git history is stored under 48a 40-digit hex name. That name is the SHA1 hash of the object's 49contents; among other things, this ensures that git will never store 50the same data twice (since identical data is given an identical SHA1 51name), and that the contents of a git object will never change (since 52that would change the object's name as well). The 7 char hex strings 53here are simply the abbreviation of such 40 character long strings. 54Abbreviations can be used everywhere where the 40 character strings 55can be used, so long as they are unambiguous. 56 57It is expected that the content of the commit object you created while 58following the example above generates a different SHA1 hash than 59the one shown above because the commit object records the time when 60it was created and the name of the person performing the commit. 61 62We can ask git about this particular object with the `cat-file` 63command. Don't copy the 40 hex digits from this example but use those 64from your own version. Note that you can shorten it to only a few 65characters to save yourself typing all 40 hex digits: 66 67------------------------------------------------ 68$ git cat-file -t 54196cc2 69commit 70$ git cat-file commit 54196cc2 71tree 92b8b694ffb1675e5975148e1121810081dbdffe 72author J. Bruce Fields <bfields@puzzle.fieldses.org> 1143414668 -0500 73committer J. Bruce Fields <bfields@puzzle.fieldses.org> 1143414668 -0500 74 75initial commit 76------------------------------------------------ 77 78A tree can refer to one or more "blob" objects, each corresponding to 79a file. In addition, a tree can also refer to other tree objects, 80thus creating a directory hierarchy. You can examine the contents of 81any tree using ls-tree (remember that a long enough initial portion 82of the SHA1 will also work): 83 84------------------------------------------------ 85$ git ls-tree 92b8b694 86100644 blob 3b18e512dba79e4c8300dd08aeb37f8e728b8dad file.txt 87------------------------------------------------ 88 89Thus we see that this tree has one file in it. The SHA1 hash is a 90reference to that file's data: 91 92------------------------------------------------ 93$ git cat-file -t 3b18e512 94blob 95------------------------------------------------ 96 97A "blob" is just file data, which we can also examine with cat-file: 98 99------------------------------------------------ 100$ git cat-file blob 3b18e512 101hello world 102------------------------------------------------ 103 104Note that this is the old file data; so the object that git named in 105its response to the initial tree was a tree with a snapshot of the 106directory state that was recorded by the first commit. 107 108All of these objects are stored under their SHA1 names inside the git 109directory: 110 111------------------------------------------------ 112$ find .git/objects/ 113.git/objects/ 114.git/objects/pack 115.git/objects/info 116.git/objects/3b 117.git/objects/3b/18e512dba79e4c8300dd08aeb37f8e728b8dad 118.git/objects/92 119.git/objects/92/b8b694ffb1675e5975148e1121810081dbdffe 120.git/objects/54 121.git/objects/54/196cc2703dc165cbd373a65a4dcf22d50ae7f7 122.git/objects/a0 123.git/objects/a0/423896973644771497bdc03eb99d5281615b51 124.git/objects/d0 125.git/objects/d0/492b368b66bdabf2ac1fd8c92b39d3db916e59 126.git/objects/c4 127.git/objects/c4/d59f390b9cfd4318117afde11d601c1085f241 128------------------------------------------------ 129 130and the contents of these files is just the compressed data plus a 131header identifying their length and their type. The type is either a 132blob, a tree, a commit, or a tag. 133 134The simplest commit to find is the HEAD commit, which we can find 135from .git/HEAD: 136 137------------------------------------------------ 138$ cat .git/HEAD 139ref: refs/heads/master 140------------------------------------------------ 141 142As you can see, this tells us which branch we're currently on, and it 143tells us this by naming a file under the .git directory, which itself 144contains a SHA1 name referring to a commit object, which we can 145examine with cat-file: 146 147------------------------------------------------ 148$ cat .git/refs/heads/master 149c4d59f390b9cfd4318117afde11d601c1085f241 150$ git cat-file -t c4d59f39 151commit 152$ git cat-file commit c4d59f39 153tree d0492b368b66bdabf2ac1fd8c92b39d3db916e59 154parent 54196cc2703dc165cbd373a65a4dcf22d50ae7f7 155author J. Bruce Fields <bfields@puzzle.fieldses.org> 1143418702 -0500 156committer J. Bruce Fields <bfields@puzzle.fieldses.org> 1143418702 -0500 157 158add emphasis 159------------------------------------------------ 160 161The "tree" object here refers to the new state of the tree: 162 163------------------------------------------------ 164$ git ls-tree d0492b36 165100644 blob a0423896973644771497bdc03eb99d5281615b51 file.txt 166$ git cat-file blob a0423896 167hello world! 168------------------------------------------------ 169 170and the "parent" object refers to the previous commit: 171 172------------------------------------------------ 173$ git cat-file commit 54196cc2 174tree 92b8b694ffb1675e5975148e1121810081dbdffe 175author J. Bruce Fields <bfields@puzzle.fieldses.org> 1143414668 -0500 176committer J. Bruce Fields <bfields@puzzle.fieldses.org> 1143414668 -0500 177 178initial commit 179------------------------------------------------ 180 181The tree object is the tree we examined first, and this commit is 182unusual in that it lacks any parent. 183 184Most commits have only one parent, but it is also common for a commit 185to have multiple parents. In that case the commit represents a 186merge, with the parent references pointing to the heads of the merged 187branches. 188 189Besides blobs, trees, and commits, the only remaining type of object 190is a "tag", which we won't discuss here; refer to linkgit:git-tag[1] 191for details. 192 193So now we know how git uses the object database to represent a 194project's history: 195 196 * "commit" objects refer to "tree" objects representing the 197 snapshot of a directory tree at a particular point in the 198 history, and refer to "parent" commits to show how they're 199 connected into the project history. 200 * "tree" objects represent the state of a single directory, 201 associating directory names to "blob" objects containing file 202 data and "tree" objects containing subdirectory information. 203 * "blob" objects contain file data without any other structure. 204 * References to commit objects at the head of each branch are 205 stored in files under .git/refs/heads/. 206 * The name of the current branch is stored in .git/HEAD. 207 208Note, by the way, that lots of commands take a tree as an argument. 209But as we can see above, a tree can be referred to in many different 210ways--by the SHA1 name for that tree, by the name of a commit that 211refers to the tree, by the name of a branch whose head refers to that 212tree, etc.--and most such commands can accept any of these names. 213 214In command synopses, the word "tree-ish" is sometimes used to 215designate such an argument. 216 217The index file 218-------------- 219 220The primary tool we've been using to create commits is `git-commit 221-a`, which creates a commit including every change you've made to 222your working tree. But what if you want to commit changes only to 223certain files? Or only certain changes to certain files? 224 225If we look at the way commits are created under the cover, we'll see 226that there are more flexible ways creating commits. 227 228Continuing with our test-project, let's modify file.txt again: 229 230------------------------------------------------ 231$ echo "hello world, again" >>file.txt 232------------------------------------------------ 233 234but this time instead of immediately making the commit, let's take an 235intermediate step, and ask for diffs along the way to keep track of 236what's happening: 237 238------------------------------------------------ 239$ git diff 240--- a/file.txt 241+++ b/file.txt 242@@ -1 +1,2 @@ 243 hello world! 244+hello world, again 245$ git add file.txt 246$ git diff 247------------------------------------------------ 248 249The last diff is empty, but no new commits have been made, and the 250head still doesn't contain the new line: 251 252------------------------------------------------ 253$ git diff HEAD 254diff --git a/file.txt b/file.txt 255index a042389..513feba 100644 256--- a/file.txt 257+++ b/file.txt 258@@ -1 +1,2 @@ 259 hello world! 260+hello world, again 261------------------------------------------------ 262 263So 'git diff' is comparing against something other than the head. 264The thing that it's comparing against is actually the index file, 265which is stored in .git/index in a binary format, but whose contents 266we can examine with ls-files: 267 268------------------------------------------------ 269$ git ls-files --stage 270100644 513feba2e53ebbd2532419ded848ba19de88ba00 0 file.txt 271$ git cat-file -t 513feba2 272blob 273$ git cat-file blob 513feba2 274hello world! 275hello world, again 276------------------------------------------------ 277 278So what our 'git add' did was store a new blob and then put 279a reference to it in the index file. If we modify the file again, 280we'll see that the new modifications are reflected in the 'git diff' 281output: 282 283------------------------------------------------ 284$ echo 'again?' >>file.txt 285$ git diff 286index 513feba..ba3da7b 100644 287--- a/file.txt 288+++ b/file.txt 289@@ -1,2 +1,3 @@ 290 hello world! 291 hello world, again 292+again? 293------------------------------------------------ 294 295With the right arguments, 'git diff' can also show us the difference 296between the working directory and the last commit, or between the 297index and the last commit: 298 299------------------------------------------------ 300$ git diff HEAD 301diff --git a/file.txt b/file.txt 302index a042389..ba3da7b 100644 303--- a/file.txt 304+++ b/file.txt 305@@ -1 +1,3 @@ 306 hello world! 307+hello world, again 308+again? 309$ git diff --cached 310diff --git a/file.txt b/file.txt 311index a042389..513feba 100644 312--- a/file.txt 313+++ b/file.txt 314@@ -1 +1,2 @@ 315 hello world! 316+hello world, again 317------------------------------------------------ 318 319At any time, we can create a new commit using 'git commit' (without 320the "-a" option), and verify that the state committed only includes the 321changes stored in the index file, not the additional change that is 322still only in our working tree: 323 324------------------------------------------------ 325$ git commit -m "repeat" 326$ git diff HEAD 327diff --git a/file.txt b/file.txt 328index 513feba..ba3da7b 100644 329--- a/file.txt 330+++ b/file.txt 331@@ -1,2 +1,3 @@ 332 hello world! 333 hello world, again 334+again? 335------------------------------------------------ 336 337So by default 'git commit' uses the index to create the commit, not 338the working tree; the "-a" option to commit tells it to first update 339the index with all changes in the working tree. 340 341Finally, it's worth looking at the effect of 'git add' on the index 342file: 343 344------------------------------------------------ 345$ echo "goodbye, world" >closing.txt 346$ git add closing.txt 347------------------------------------------------ 348 349The effect of the 'git add' was to add one entry to the index file: 350 351------------------------------------------------ 352$ git ls-files --stage 353100644 8b9743b20d4b15be3955fc8d5cd2b09cd2336138 0 closing.txt 354100644 513feba2e53ebbd2532419ded848ba19de88ba00 0 file.txt 355------------------------------------------------ 356 357And, as you can see with cat-file, this new entry refers to the 358current contents of the file: 359 360------------------------------------------------ 361$ git cat-file blob 8b9743b2 362goodbye, world 363------------------------------------------------ 364 365The "status" command is a useful way to get a quick summary of the 366situation: 367 368------------------------------------------------ 369$ git status 370# On branch master 371# Changes to be committed: 372# (use "git reset HEAD <file>..." to unstage) 373# 374# new file: closing.txt 375# 376# Changes not staged for commit: 377# (use "git add <file>..." to update what will be committed) 378# 379# modified: file.txt 380# 381------------------------------------------------ 382 383Since the current state of closing.txt is cached in the index file, 384it is listed as "Changes to be committed". Since file.txt has 385changes in the working directory that aren't reflected in the index, 386it is marked "changed but not updated". At this point, running "git 387commit" would create a commit that added closing.txt (with its new 388contents), but that didn't modify file.txt. 389 390Also, note that a bare `git diff` shows the changes to file.txt, but 391not the addition of closing.txt, because the version of closing.txt 392in the index file is identical to the one in the working directory. 393 394In addition to being the staging area for new commits, the index file 395is also populated from the object database when checking out a 396branch, and is used to hold the trees involved in a merge operation. 397See linkgit:gitcore-tutorial[7] and the relevant man 398pages for details. 399 400What next? 401---------- 402 403At this point you should know everything necessary to read the man 404pages for any of the git commands; one good place to start would be 405with the commands mentioned in link:everyday.html[Everyday git]. You 406should be able to find any unknown jargon in linkgit:gitglossary[7]. 407 408The link:user-manual.html[Git User's Manual] provides a more 409comprehensive introduction to git. 410 411linkgit:gitcvs-migration[7] explains how to 412import a CVS repository into git, and shows how to use git in a 413CVS-like way. 414 415For some interesting examples of git use, see the 416link:howto-index.html[howtos]. 417 418For git developers, linkgit:gitcore-tutorial[7] goes 419into detail on the lower-level git mechanisms involved in, for 420example, creating a new commit. 421 422SEE ALSO 423-------- 424linkgit:gittutorial[7], 425linkgit:gitcvs-migration[7], 426linkgit:gitcore-tutorial[7], 427linkgit:gitglossary[7], 428linkgit:git-help[1], 429link:everyday.html[Everyday git], 430link:user-manual.html[The Git User's Manual] 431 432GIT 433--- 434Part of the linkgit:git[1] suite.