1Core GIT Tests 2============== 3 4This directory holds many test scripts for core GIT tools. The 5first part of this short document describes how to run the tests 6and read their output. 7 8When fixing the tools or adding enhancements, you are strongly 9encouraged to add tests in this directory to cover what you are 10trying to fix or enhance. The later part of this short document 11describes how your test scripts should be organized. 12 13 14Running Tests 15------------- 16 17The easiest way to run tests is to say "make". This runs all 18the tests. 19 20 *** t0000-basic.sh *** 21 ok 1 - .git/objects should be empty after git init in an empty repo. 22 ok 2 - .git/objects should have 3 subdirectories. 23 ok 3 - success is reported like this 24 ... 25 ok 43 - very long name in the index handled sanely 26 # fixed 1 known breakage(s) 27 # still have 1 known breakage(s) 28 # passed all remaining 42 test(s) 29 1..43 30 *** t0001-init.sh *** 31 ok 1 - plain 32 ok 2 - plain with GIT_WORK_TREE 33 ok 3 - plain bare 34 35Since the tests all output TAP (see http://testanything.org) they can 36be run with any TAP harness. Here's an example of parallel testing 37powered by a recent version of prove(1): 38 39 $ prove --timer --jobs 15 ./t[0-9]*.sh 40 [19:17:33] ./t0005-signals.sh ................................... ok 36 ms 41 [19:17:33] ./t0022-crlf-rename.sh ............................... ok 69 ms 42 [19:17:33] ./t0024-crlf-archive.sh .............................. ok 154 ms 43 [19:17:33] ./t0004-unwritable.sh ................................ ok 289 ms 44 [19:17:33] ./t0002-gitfile.sh ................................... ok 480 ms 45 ===( 102;0 25/? 6/? 5/? 16/? 1/? 4/? 2/? 1/? 3/? 1... )=== 46 47prove and other harnesses come with a lot of useful options. The 48--state option in particular is very useful: 49 50 # Repeat until no more failures 51 $ prove -j 15 --state=failed,save ./t[0-9]*.sh 52 53You can give DEFAULT_TEST_TARGET=prove on the make command (or define it 54in config.mak) to cause "make test" to run tests under prove. 55GIT_PROVE_OPTS can be used to pass additional options, e.g. 56 57 $ make DEFAULT_TEST_TARGET=prove GIT_PROVE_OPTS='--timer --jobs 16' test 58 59You can also run each test individually from command line, like this: 60 61 $ sh ./t3010-ls-files-killed-modified.sh 62 ok 1 - git update-index --add to add various paths. 63 ok 2 - git ls-files -k to show killed files. 64 ok 3 - validate git ls-files -k output. 65 ok 4 - git ls-files -m to show modified files. 66 ok 5 - validate git ls-files -m output. 67 # passed all 5 test(s) 68 1..5 69 70You can pass --verbose (or -v), --debug (or -d), and --immediate 71(or -i) command line argument to the test, or by setting GIT_TEST_OPTS 72appropriately before running "make". 73 74--verbose:: 75 This makes the test more verbose. Specifically, the 76 command being run and their output if any are also 77 output. 78 79--verbose-only=<pattern>:: 80 Like --verbose, but the effect is limited to tests with 81 numbers matching <pattern>. The number matched against is 82 simply the running count of the test within the file. 83 84--debug:: 85 This may help the person who is developing a new test. 86 It causes the command defined with test_debug to run. 87 The "trash" directory (used to store all temporary data 88 during testing) is not deleted even if there are no 89 failed tests so that you can inspect its contents after 90 the test finished. 91 92--immediate:: 93 This causes the test to immediately exit upon the first 94 failed test. Cleanup commands requested with 95 test_when_finished are not executed if the test failed, 96 in order to keep the state for inspection by the tester 97 to diagnose the bug. 98 99--long-tests:: 100 This causes additional long-running tests to be run (where 101 available), for more exhaustive testing. 102 103--valgrind=<tool>:: 104 Execute all Git binaries under valgrind tool <tool> and exit 105 with status 126 on errors (just like regular tests, this will 106 only stop the test script when running under -i). 107 108 Since it makes no sense to run the tests with --valgrind and 109 not see any output, this option implies --verbose. For 110 convenience, it also implies --tee. 111 112 <tool> defaults to 'memcheck', just like valgrind itself. 113 Other particularly useful choices include 'helgrind' and 114 'drd', but you may use any tool recognized by your valgrind 115 installation. 116 117 As a special case, <tool> can be 'memcheck-fast', which uses 118 memcheck but disables --track-origins. Use this if you are 119 running tests in bulk, to see if there are _any_ memory 120 issues. 121 122 Note that memcheck is run with the option --leak-check=no, 123 as the git process is short-lived and some errors are not 124 interesting. In order to run a single command under the same 125 conditions manually, you should set GIT_VALGRIND to point to 126 the 't/valgrind/' directory and use the commands under 127 't/valgrind/bin/'. 128 129--valgrind-only=<pattern>:: 130 Like --valgrind, but the effect is limited to tests with 131 numbers matching <pattern>. The number matched against is 132 simply the running count of the test within the file. 133 134--tee:: 135 In addition to printing the test output to the terminal, 136 write it to files named 't/test-results/$TEST_NAME.out'. 137 As the names depend on the tests' file names, it is safe to 138 run the tests with this option in parallel. 139 140--with-dashes:: 141 By default tests are run without dashed forms of 142 commands (like git-commit) in the PATH (it only uses 143 wrappers from ../bin-wrappers). Use this option to include 144 the build directory (..) in the PATH, which contains all 145 the dashed forms of commands. This option is currently 146 implied by other options like --valgrind and 147 GIT_TEST_INSTALLED. 148 149--root=<directory>:: 150 Create "trash" directories used to store all temporary data during 151 testing under <directory>, instead of the t/ directory. 152 Using this option with a RAM-based filesystem (such as tmpfs) 153 can massively speed up the test suite. 154 155You can also set the GIT_TEST_INSTALLED environment variable to 156the bindir of an existing git installation to test that installation. 157You still need to have built this git sandbox, from which various 158test-* support programs, templates, and perl libraries are used. 159If your installed git is incomplete, it will silently test parts of 160your built version instead. 161 162When using GIT_TEST_INSTALLED, you can also set GIT_TEST_EXEC_PATH to 163override the location of the dashed-form subcommands (what 164GIT_EXEC_PATH would be used for during normal operation). 165GIT_TEST_EXEC_PATH defaults to `$GIT_TEST_INSTALLED/git --exec-path`. 166 167 168Skipping Tests 169-------------- 170 171In some environments, certain tests have no way of succeeding 172due to platform limitation, such as lack of 'unzip' program, or 173filesystem that do not allow arbitrary sequence of non-NUL bytes 174as pathnames. 175 176You should be able to say something like 177 178 $ GIT_SKIP_TESTS=t9200.8 sh ./t9200-git-cvsexport-commit.sh 179 180and even: 181 182 $ GIT_SKIP_TESTS='t[0-4]??? t91?? t9200.8' make 183 184to omit such tests. The value of the environment variable is a 185SP separated list of patterns that tells which tests to skip, 186and either can match the "t[0-9]{4}" part to skip the whole 187test, or t[0-9]{4} followed by ".$number" to say which 188particular test to skip. 189 190Note that some tests in the existing test suite rely on previous 191test item, so you cannot arbitrarily disable one and expect the 192remainder of test to check what the test originally was intended 193to check. 194 195 196Naming Tests 197------------ 198 199The test files are named as: 200 201 tNNNN-commandname-details.sh 202 203where N is a decimal digit. 204 205First digit tells the family: 206 207 0 - the absolute basics and global stuff 208 1 - the basic commands concerning database 209 2 - the basic commands concerning the working tree 210 3 - the other basic commands (e.g. ls-files) 211 4 - the diff commands 212 5 - the pull and exporting commands 213 6 - the revision tree commands (even e.g. merge-base) 214 7 - the porcelainish commands concerning the working tree 215 8 - the porcelainish commands concerning forensics 216 9 - the git tools 217 218Second digit tells the particular command we are testing. 219 220Third digit (optionally) tells the particular switch or group of switches 221we are testing. 222 223If you create files under t/ directory (i.e. here) that is not 224the top-level test script, never name the file to match the above 225pattern. The Makefile here considers all such files as the 226top-level test script and tries to run all of them. Care is 227especially needed if you are creating a common test library 228file, similar to test-lib.sh, because such a library file may 229not be suitable for standalone execution. 230 231 232Writing Tests 233------------- 234 235The test script is written as a shell script. It should start 236with the standard "#!/bin/sh" with copyright notices, and an 237assignment to variable 'test_description', like this: 238 239 #!/bin/sh 240 # 241 # Copyright (c) 2005 Junio C Hamano 242 # 243 244 test_description='xxx test (option --frotz) 245 246 This test registers the following structure in the cache 247 and tries to run git-ls-files with option --frotz.' 248 249 250Source 'test-lib.sh' 251-------------------- 252 253After assigning test_description, the test script should source 254test-lib.sh like this: 255 256 . ./test-lib.sh 257 258This test harness library does the following things: 259 260 - If the script is invoked with command line argument --help 261 (or -h), it shows the test_description and exits. 262 263 - Creates an empty test directory with an empty .git/objects database 264 and chdir(2) into it. This directory is 't/trash 265 directory.$test_name_without_dotsh', with t/ subject to change by 266 the --root option documented above. 267 268 - Defines standard test helper functions for your scripts to 269 use. These functions are designed to make all scripts behave 270 consistently when command line arguments --verbose (or -v), 271 --debug (or -d), and --immediate (or -i) is given. 272 273Do's, don'ts & things to keep in mind 274------------------------------------- 275 276Here are a few examples of things you probably should and shouldn't do 277when writing tests. 278 279Do: 280 281 - Put all code inside test_expect_success and other assertions. 282 283 Even code that isn't a test per se, but merely some setup code 284 should be inside a test assertion. 285 286 - Chain your test assertions 287 288 Write test code like this: 289 290 git merge foo && 291 git push bar && 292 test ... 293 294 Instead of: 295 296 git merge hla 297 git push gh 298 test ... 299 300 That way all of the commands in your tests will succeed or fail. If 301 you must ignore the return value of something, consider using a 302 helper function (e.g. use sane_unset instead of unset, in order 303 to avoid unportable return value for unsetting a variable that was 304 already unset), or prepending the command with test_might_fail or 305 test_must_fail. 306 307 - Check the test coverage for your tests. See the "Test coverage" 308 below. 309 310 Don't blindly follow test coverage metrics; if a new function you added 311 doesn't have any coverage, then you're probably doing something wrong, 312 but having 100% coverage doesn't necessarily mean that you tested 313 everything. 314 315 Tests that are likely to smoke out future regressions are better 316 than tests that just inflate the coverage metrics. 317 318 - When a test checks for an absolute path that a git command generated, 319 construct the expected value using $(pwd) rather than $PWD, 320 $TEST_DIRECTORY, or $TRASH_DIRECTORY. It makes a difference on 321 Windows, where the shell (MSYS bash) mangles absolute path names. 322 For details, see the commit message of 4114156ae9. 323 324Don't: 325 326 - exit() within a <script> part. 327 328 The harness will catch this as a programming error of the test. 329 Use test_done instead if you need to stop the tests early (see 330 "Skipping tests" below). 331 332 - use '! git cmd' when you want to make sure the git command exits 333 with failure in a controlled way by calling "die()". Instead, 334 use 'test_must_fail git cmd'. This will signal a failure if git 335 dies in an unexpected way (e.g. segfault). 336 337 - use perl without spelling it as "$PERL_PATH". This is to help our 338 friends on Windows where the platform Perl often adds CR before 339 the end of line, and they bundle Git with a version of Perl that 340 does not do so, whose path is specified with $PERL_PATH. 341 342 - use sh without spelling it as "$SHELL_PATH", when the script can 343 be misinterpreted by broken platform shell (e.g. Solaris). 344 345 - chdir around in tests. It is not sufficient to chdir to 346 somewhere and then chdir back to the original location later in 347 the test, as any intermediate step can fail and abort the test, 348 causing the next test to start in an unexpected directory. Do so 349 inside a subshell if necessary. 350 351 - Break the TAP output 352 353 The raw output from your test may be interpreted by a TAP harness. TAP 354 harnesses will ignore everything they don't know about, but don't step 355 on their toes in these areas: 356 357 - Don't print lines like "$x..$y" where $x and $y are integers. 358 359 - Don't print lines that begin with "ok" or "not ok". 360 361 TAP harnesses expect a line that begins with either "ok" and "not 362 ok" to signal a test passed or failed (and our harness already 363 produces such lines), so your script shouldn't emit such lines to 364 their output. 365 366 You can glean some further possible issues from the TAP grammar 367 (see http://search.cpan.org/perldoc?TAP::Parser::Grammar#TAP_Grammar) 368 but the best indication is to just run the tests with prove(1), 369 it'll complain if anything is amiss. 370 371Keep in mind: 372 373 - Inside <script> part, the standard output and standard error 374 streams are discarded, and the test harness only reports "ok" or 375 "not ok" to the end user running the tests. Under --verbose, they 376 are shown to help debugging the tests. 377 378 379Skipping tests 380-------------- 381 382If you need to skip tests you should do so by using the three-arg form 383of the test_* functions (see the "Test harness library" section 384below), e.g.: 385 386 test_expect_success PERL 'I need Perl' ' 387 "$PERL_PATH" -e "hlagh() if unf_unf()" 388 ' 389 390The advantage of skipping tests like this is that platforms that don't 391have the PERL and other optional dependencies get an indication of how 392many tests they're missing. 393 394If the test code is too hairy for that (i.e. does a lot of setup work 395outside test assertions) you can also skip all remaining tests by 396setting skip_all and immediately call test_done: 397 398 if ! test_have_prereq PERL 399 then 400 skip_all='skipping perl interface tests, perl not available' 401 test_done 402 fi 403 404The string you give to skip_all will be used as an explanation for why 405the test was skipped. 406 407End with test_done 408------------------ 409 410Your script will be a sequence of tests, using helper functions 411from the test harness library. At the end of the script, call 412'test_done'. 413 414 415Test harness library 416-------------------- 417 418There are a handful helper functions defined in the test harness 419library for your script to use. 420 421 - test_expect_success [<prereq>] <message> <script> 422 423 Usually takes two strings as parameters, and evaluates the 424 <script>. If it yields success, test is considered 425 successful. <message> should state what it is testing. 426 427 Example: 428 429 test_expect_success \ 430 'git-write-tree should be able to write an empty tree.' \ 431 'tree=$(git-write-tree)' 432 433 If you supply three parameters the first will be taken to be a 434 prerequisite; see the test_set_prereq and test_have_prereq 435 documentation below: 436 437 test_expect_success TTY 'git --paginate rev-list uses a pager' \ 438 ' ... ' 439 440 You can also supply a comma-separated list of prerequisites, in the 441 rare case where your test depends on more than one: 442 443 test_expect_success PERL,PYTHON 'yo dawg' \ 444 ' test $(perl -E 'print eval "1 +" . qx[python -c "print 2"]') == "4" ' 445 446 - test_expect_failure [<prereq>] <message> <script> 447 448 This is NOT the opposite of test_expect_success, but is used 449 to mark a test that demonstrates a known breakage. Unlike 450 the usual test_expect_success tests, which say "ok" on 451 success and "FAIL" on failure, this will say "FIXED" on 452 success and "still broken" on failure. Failures from these 453 tests won't cause -i (immediate) to stop. 454 455 Like test_expect_success this function can optionally use a three 456 argument invocation with a prerequisite as the first argument. 457 458 - test_debug <script> 459 460 This takes a single argument, <script>, and evaluates it only 461 when the test script is started with --debug command line 462 argument. This is primarily meant for use during the 463 development of a new test script. 464 465 - test_done 466 467 Your test script must have test_done at the end. Its purpose 468 is to summarize successes and failures in the test script and 469 exit with an appropriate error code. 470 471 - test_tick 472 473 Make commit and tag names consistent by setting the author and 474 committer times to defined state. Subsequent calls will 475 advance the times by a fixed amount. 476 477 - test_commit <message> [<filename> [<contents>]] 478 479 Creates a commit with the given message, committing the given 480 file with the given contents (default for both is to reuse the 481 message string), and adds a tag (again reusing the message 482 string as name). Calls test_tick to make the SHA-1s 483 reproducible. 484 485 - test_merge <message> <commit-or-tag> 486 487 Merges the given rev using the given message. Like test_commit, 488 creates a tag and calls test_tick before committing. 489 490 - test_set_prereq <prereq> 491 492 Set a test prerequisite to be used later with test_have_prereq. The 493 test-lib will set some prerequisites for you, see the 494 "Prerequisites" section below for a full list of these. 495 496 Others you can set yourself and use later with either 497 test_have_prereq directly, or the three argument invocation of 498 test_expect_success and test_expect_failure. 499 500 - test_have_prereq <prereq> 501 502 Check if we have a prerequisite previously set with 503 test_set_prereq. The most common use of this directly is to skip 504 all the tests if we don't have some essential prerequisite: 505 506 if ! test_have_prereq PERL 507 then 508 skip_all='skipping perl interface tests, perl not available' 509 test_done 510 fi 511 512 - test_external [<prereq>] <message> <external> <script> 513 514 Execute a <script> with an <external> interpreter (like perl). This 515 was added for tests like t9700-perl-git.sh which do most of their 516 work in an external test script. 517 518 test_external \ 519 'GitwebCache::*FileCache*' \ 520 "$PERL_PATH" "$TEST_DIRECTORY"/t9503/test_cache_interface.pl 521 522 If the test is outputting its own TAP you should set the 523 test_external_has_tap variable somewhere before calling the first 524 test_external* function. See t9700-perl-git.sh for an example. 525 526 # The external test will outputs its own plan 527 test_external_has_tap=1 528 529 - test_external_without_stderr [<prereq>] <message> <external> <script> 530 531 Like test_external but fail if there's any output on stderr, 532 instead of checking the exit code. 533 534 test_external_without_stderr \ 535 'Perl API' \ 536 "$PERL_PATH" "$TEST_DIRECTORY"/t9700/test.pl 537 538 - test_expect_code <exit-code> <command> 539 540 Run a command and ensure that it exits with the given exit code. 541 For example: 542 543 test_expect_success 'Merge with d/f conflicts' ' 544 test_expect_code 1 git merge "merge msg" B master 545 ' 546 547 - test_must_fail <git-command> 548 549 Run a git command and ensure it fails in a controlled way. Use 550 this instead of "! <git-command>". When git-command dies due to a 551 segfault, test_must_fail diagnoses it as an error; "! <git-command>" 552 treats it as just another expected failure, which would let such a 553 bug go unnoticed. 554 555 - test_might_fail <git-command> 556 557 Similar to test_must_fail, but tolerate success, too. Use this 558 instead of "<git-command> || :" to catch failures due to segv. 559 560 - test_cmp <expected> <actual> 561 562 Check whether the content of the <actual> file matches the 563 <expected> file. This behaves like "cmp" but produces more 564 helpful output when the test is run with "-v" option. 565 566 - test_line_count (= | -lt | -ge | ...) <length> <file> 567 568 Check whether a file has the length it is expected to. 569 570 - test_path_is_file <path> [<diagnosis>] 571 test_path_is_dir <path> [<diagnosis>] 572 test_path_is_missing <path> [<diagnosis>] 573 574 Check if the named path is a file, if the named path is a 575 directory, or if the named path does not exist, respectively, 576 and fail otherwise, showing the <diagnosis> text. 577 578 - test_when_finished <script> 579 580 Prepend <script> to a list of commands to run to clean up 581 at the end of the current test. If some clean-up command 582 fails, the test will not pass. 583 584 Example: 585 586 test_expect_success 'branch pointing to non-commit' ' 587 git rev-parse HEAD^{tree} >.git/refs/heads/invalid && 588 test_when_finished "git update-ref -d refs/heads/invalid" && 589 ... 590 ' 591 592 - test_pause 593 594 This command is useful for writing and debugging tests and must be 595 removed before submitting. It halts the execution of the test and 596 spawns a shell in the trash directory. Exit the shell to continue 597 the test. Example: 598 599 test_expect_success 'test' ' 600 git do-something >actual && 601 test_pause && 602 test_cmp expected actual 603 ' 604 605Prerequisites 606------------- 607 608These are the prerequisites that the test library predefines with 609test_have_prereq. 610 611See the prereq argument to the test_* functions in the "Test harness 612library" section above and the "test_have_prereq" function for how to 613use these, and "test_set_prereq" for how to define your own. 614 615 - PERL & PYTHON 616 617 Git wasn't compiled with NO_PERL=YesPlease or 618 NO_PYTHON=YesPlease. Wrap any tests that need Perl or Python in 619 these. 620 621 - POSIXPERM 622 623 The filesystem supports POSIX style permission bits. 624 625 - BSLASHPSPEC 626 627 Backslashes in pathspec are not directory separators. This is not 628 set on Windows. See 6fd1106a for details. 629 630 - EXECKEEPSPID 631 632 The process retains the same pid across exec(2). See fb9a2bea for 633 details. 634 635 - PIPE 636 637 The filesystem we're on supports creation of FIFOs (named pipes) 638 via mkfifo(1). 639 640 - SYMLINKS 641 642 The filesystem we're on supports symbolic links. E.g. a FAT 643 filesystem doesn't support these. See 704a3143 for details. 644 645 - SANITY 646 647 Test is not run by root user, and an attempt to write to an 648 unwritable file is expected to fail correctly. 649 650 - LIBPCRE 651 652 Git was compiled with USE_LIBPCRE=YesPlease. Wrap any tests 653 that use git-grep --perl-regexp or git-grep -P in these. 654 655 - CASE_INSENSITIVE_FS 656 657 Test is run on a case insensitive file system. 658 659 - UTF8_NFD_TO_NFC 660 661 Test is run on a filesystem which converts decomposed utf-8 (nfd) 662 to precomposed utf-8 (nfc). 663 664Tips for Writing Tests 665---------------------- 666 667As with any programming projects, existing programs are the best 668source of the information. However, do _not_ emulate 669t0000-basic.sh when writing your tests. The test is special in 670that it tries to validate the very core of GIT. For example, it 671knows that there will be 256 subdirectories under .git/objects/, 672and it knows that the object ID of an empty tree is a certain 67340-byte string. This is deliberately done so in t0000-basic.sh 674because the things the very basic core test tries to achieve is 675to serve as a basis for people who are changing the GIT internal 676drastically. For these people, after making certain changes, 677not seeing failures from the basic test _is_ a failure. And 678such drastic changes to the core GIT that even changes these 679otherwise supposedly stable object IDs should be accompanied by 680an update to t0000-basic.sh. 681 682However, other tests that simply rely on basic parts of the core 683GIT working properly should not have that level of intimate 684knowledge of the core GIT internals. If all the test scripts 685hardcoded the object IDs like t0000-basic.sh does, that defeats 686the purpose of t0000-basic.sh, which is to isolate that level of 687validation in one place. Your test also ends up needing 688updating when such a change to the internal happens, so do _not_ 689do it and leave the low level of validation to t0000-basic.sh. 690 691Test coverage 692------------- 693 694You can use the coverage tests to find code paths that are not being 695used or properly exercised yet. 696 697To do that, run the coverage target at the top-level (not in the t/ 698directory): 699 700 make coverage 701 702That'll compile Git with GCC's coverage arguments, and generate a test 703report with gcov after the tests finish. Running the coverage tests 704can take a while, since running the tests in parallel is incompatible 705with GCC's coverage mode. 706 707After the tests have run you can generate a list of untested 708functions: 709 710 make coverage-untested-functions 711 712You can also generate a detailed per-file HTML report using the 713Devel::Cover module. To install it do: 714 715 # On Debian or Ubuntu: 716 sudo aptitude install libdevel-cover-perl 717 718 # From the CPAN with cpanminus 719 curl -L http://cpanmin.us | perl - --sudo --self-upgrade 720 cpanm --sudo Devel::Cover 721 722Then, at the top-level: 723 724 make cover_db_html 725 726That'll generate a detailed cover report in the "cover_db_html" 727directory, which you can then copy to a webserver, or inspect locally 728in a browser.