tests: add a special setup where stash.useBuiltin is off
[git/git.git] / t / README
1 Core GIT Tests
2 ==============
3
4 This directory holds many test scripts for core GIT tools. The
5 first part of this short document describes how to run the tests
6 and read their output.
7
8 When fixing the tools or adding enhancements, you are strongly
9 encouraged to add tests in this directory to cover what you are
10 trying to fix or enhance. The later part of this short document
11 describes how your test scripts should be organized.
12
13
14 Running Tests
15 -------------
16
17 The easiest way to run tests is to say "make". This runs all
18 the 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
35 Since the tests all output TAP (see http://testanything.org) they can
36 be run with any TAP harness. Here's an example of parallel testing
37 powered 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
47 prove 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
53 You can give DEFAULT_TEST_TARGET=prove on the make command (or define it
54 in config.mak) to cause "make test" to run tests under prove.
55 GIT_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
59 You 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
70 You 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
72 appropriately before running "make".
73
74 -v::
75 --verbose::
76 This makes the test more verbose. Specifically, the
77 command being run and their output if any are also
78 output.
79
80 --verbose-only=<pattern>::
81 Like --verbose, but the effect is limited to tests with
82 numbers matching <pattern>. The number matched against is
83 simply the running count of the test within the file.
84
85 -x::
86 Turn on shell tracing (i.e., `set -x`) during the tests
87 themselves. Implies `--verbose`.
88 Ignored in test scripts that set the variable 'test_untraceable'
89 to a non-empty value, unless it's run with a Bash version
90 supporting BASH_XTRACEFD, i.e. v4.1 or later.
91
92 -d::
93 --debug::
94 This may help the person who is developing a new test.
95 It causes the command defined with test_debug to run.
96 The "trash" directory (used to store all temporary data
97 during testing) is not deleted even if there are no
98 failed tests so that you can inspect its contents after
99 the test finished.
100
101 -i::
102 --immediate::
103 This causes the test to immediately exit upon the first
104 failed test. Cleanup commands requested with
105 test_when_finished are not executed if the test failed,
106 in order to keep the state for inspection by the tester
107 to diagnose the bug.
108
109 -l::
110 --long-tests::
111 This causes additional long-running tests to be run (where
112 available), for more exhaustive testing.
113
114 -r::
115 --run=<test-selector>::
116 Run only the subset of tests indicated by
117 <test-selector>. See section "Skipping Tests" below for
118 <test-selector> syntax.
119
120 --valgrind=<tool>::
121 Execute all Git binaries under valgrind tool <tool> and exit
122 with status 126 on errors (just like regular tests, this will
123 only stop the test script when running under -i).
124
125 Since it makes no sense to run the tests with --valgrind and
126 not see any output, this option implies --verbose. For
127 convenience, it also implies --tee.
128
129 <tool> defaults to 'memcheck', just like valgrind itself.
130 Other particularly useful choices include 'helgrind' and
131 'drd', but you may use any tool recognized by your valgrind
132 installation.
133
134 As a special case, <tool> can be 'memcheck-fast', which uses
135 memcheck but disables --track-origins. Use this if you are
136 running tests in bulk, to see if there are _any_ memory
137 issues.
138
139 Note that memcheck is run with the option --leak-check=no,
140 as the git process is short-lived and some errors are not
141 interesting. In order to run a single command under the same
142 conditions manually, you should set GIT_VALGRIND to point to
143 the 't/valgrind/' directory and use the commands under
144 't/valgrind/bin/'.
145
146 --valgrind-only=<pattern>::
147 Like --valgrind, but the effect is limited to tests with
148 numbers matching <pattern>. The number matched against is
149 simply the running count of the test within the file.
150
151 --tee::
152 In addition to printing the test output to the terminal,
153 write it to files named 't/test-results/$TEST_NAME.out'.
154 As the names depend on the tests' file names, it is safe to
155 run the tests with this option in parallel.
156
157 -V::
158 --verbose-log::
159 Write verbose output to the same logfile as `--tee`, but do
160 _not_ write it to stdout. Unlike `--tee --verbose`, this option
161 is safe to use when stdout is being consumed by a TAP parser
162 like `prove`. Implies `--tee` and `--verbose`.
163
164 --with-dashes::
165 By default tests are run without dashed forms of
166 commands (like git-commit) in the PATH (it only uses
167 wrappers from ../bin-wrappers). Use this option to include
168 the build directory (..) in the PATH, which contains all
169 the dashed forms of commands. This option is currently
170 implied by other options like --valgrind and
171 GIT_TEST_INSTALLED.
172
173 --root=<directory>::
174 Create "trash" directories used to store all temporary data during
175 testing under <directory>, instead of the t/ directory.
176 Using this option with a RAM-based filesystem (such as tmpfs)
177 can massively speed up the test suite.
178
179 --chain-lint::
180 --no-chain-lint::
181 If --chain-lint is enabled, the test harness will check each
182 test to make sure that it properly "&&-chains" all commands (so
183 that a failure in the middle does not go unnoticed by the final
184 exit code of the test). This check is performed in addition to
185 running the tests themselves. You may also enable or disable
186 this feature by setting the GIT_TEST_CHAIN_LINT environment
187 variable to "1" or "0", respectively.
188
189 You can also set the GIT_TEST_INSTALLED environment variable to
190 the bindir of an existing git installation to test that installation.
191 You still need to have built this git sandbox, from which various
192 test-* support programs, templates, and perl libraries are used.
193 If your installed git is incomplete, it will silently test parts of
194 your built version instead.
195
196 When using GIT_TEST_INSTALLED, you can also set GIT_TEST_EXEC_PATH to
197 override the location of the dashed-form subcommands (what
198 GIT_EXEC_PATH would be used for during normal operation).
199 GIT_TEST_EXEC_PATH defaults to `$GIT_TEST_INSTALLED/git --exec-path`.
200
201
202 Skipping Tests
203 --------------
204
205 In some environments, certain tests have no way of succeeding
206 due to platform limitation, such as lack of 'unzip' program, or
207 filesystem that do not allow arbitrary sequence of non-NUL bytes
208 as pathnames.
209
210 You should be able to say something like
211
212 $ GIT_SKIP_TESTS=t9200.8 sh ./t9200-git-cvsexport-commit.sh
213
214 and even:
215
216 $ GIT_SKIP_TESTS='t[0-4]??? t91?? t9200.8' make
217
218 to omit such tests. The value of the environment variable is a
219 SP separated list of patterns that tells which tests to skip,
220 and either can match the "t[0-9]{4}" part to skip the whole
221 test, or t[0-9]{4} followed by ".$number" to say which
222 particular test to skip.
223
224 For an individual test suite --run could be used to specify that
225 only some tests should be run or that some tests should be
226 excluded from a run.
227
228 The argument for --run is a list of individual test numbers or
229 ranges with an optional negation prefix that define what tests in
230 a test suite to include in the run. A range is two numbers
231 separated with a dash and matches a range of tests with both ends
232 been included. You may omit the first or the second number to
233 mean "from the first test" or "up to the very last test"
234 respectively.
235
236 Optional prefix of '!' means that the test or a range of tests
237 should be excluded from the run.
238
239 If --run starts with an unprefixed number or range the initial
240 set of tests to run is empty. If the first item starts with '!'
241 all the tests are added to the initial set. After initial set is
242 determined every test number or range is added or excluded from
243 the set one by one, from left to right.
244
245 Individual numbers or ranges could be separated either by a space
246 or a comma.
247
248 For example, to run only tests up to a specific test (21), one
249 could do this:
250
251 $ sh ./t9200-git-cvsexport-commit.sh --run='1-21'
252
253 or this:
254
255 $ sh ./t9200-git-cvsexport-commit.sh --run='-21'
256
257 Common case is to run several setup tests (1, 2, 3) and then a
258 specific test (21) that relies on that setup:
259
260 $ sh ./t9200-git-cvsexport-commit.sh --run='1 2 3 21'
261
262 or:
263
264 $ sh ./t9200-git-cvsexport-commit.sh --run=1,2,3,21
265
266 or:
267
268 $ sh ./t9200-git-cvsexport-commit.sh --run='-3 21'
269
270 As noted above, the test set is built by going through the items
271 from left to right, so this:
272
273 $ sh ./t9200-git-cvsexport-commit.sh --run='1-4 !3'
274
275 will run tests 1, 2, and 4. Items that come later have higher
276 precedence. It means that this:
277
278 $ sh ./t9200-git-cvsexport-commit.sh --run='!3 1-4'
279
280 would just run tests from 1 to 4, including 3.
281
282 You may use negation with ranges. The following will run all
283 test in the test suite except from 7 up to 11:
284
285 $ sh ./t9200-git-cvsexport-commit.sh --run='!7-11'
286
287 Some tests in a test suite rely on the previous tests performing
288 certain actions, specifically some tests are designated as
289 "setup" test, so you cannot _arbitrarily_ disable one test and
290 expect the rest to function correctly.
291
292 --run is mostly useful when you want to focus on a specific test
293 and know what setup is needed for it. Or when you want to run
294 everything up to a certain test.
295
296
297 Running tests with special setups
298 ---------------------------------
299
300 The whole test suite could be run to test some special features
301 that cannot be easily covered by a few specific test cases. These
302 could be enabled by running the test suite with correct GIT_TEST_
303 environment set.
304
305 GIT_TEST_GETTEXT_POISON=<non-empty?> turns all strings marked for
306 translation into gibberish if non-empty (think "test -n"). Used for
307 spotting those tests that need to be marked with a C_LOCALE_OUTPUT
308 prerequisite when adding more strings for translation. See "Testing
309 marked strings" in po/README for details.
310
311 GIT_TEST_SPLIT_INDEX=<boolean> forces split-index mode on the whole
312 test suite. Accept any boolean values that are accepted by git-config.
313
314 GIT_TEST_FULL_IN_PACK_ARRAY=<boolean> exercises the uncommon
315 pack-objects code path where there are more than 1024 packs even if
316 the actual number of packs in repository is below this limit. Accept
317 any boolean values that are accepted by git-config.
318
319 GIT_TEST_OE_SIZE=<n> exercises the uncommon pack-objects code path
320 where we do not cache object size in memory and read it from existing
321 packs on demand. This normally only happens when the object size is
322 over 2GB. This variable forces the code path on any object larger than
323 <n> bytes.
324
325 GIT_TEST_OE_DELTA_SIZE=<n> exercises the uncommon pack-objects code
326 path where deltas larger than this limit require extra memory
327 allocation for bookkeeping.
328
329 GIT_TEST_VALIDATE_INDEX_CACHE_ENTRIES=<boolean> checks that cache-tree
330 records are valid when the index is written out or after a merge. This
331 is mostly to catch missing invalidation. Default is true.
332
333 GIT_TEST_COMMIT_GRAPH=<boolean>, when true, forces the commit-graph to
334 be written after every 'git commit' command, and overrides the
335 'core.commitGraph' setting to true.
336
337 GIT_TEST_FSMONITOR=$PWD/t7519/fsmonitor-all exercises the fsmonitor
338 code path for utilizing a file system monitor to speed up detecting
339 new or changed files.
340
341 GIT_TEST_INDEX_VERSION=<n> exercises the index read/write code path
342 for the index version specified. Can be set to any valid version
343 (currently 2, 3, or 4).
344
345 GIT_TEST_PRELOAD_INDEX=<boolean> exercises the preload-index code path
346 by overriding the minimum number of cache entries required per thread.
347
348 GIT_TEST_REBASE_USE_BUILTIN=<boolean>, when false, disables the
349 builtin version of git-rebase. See 'rebase.useBuiltin' in
350 git-config(1).
351
352 GIT_TEST_STASH_USE_BUILTIN=<boolean>, when false, disables the
353 built-in version of git-stash. See 'stash.useBuiltin' in
354 git-config(1).
355
356 GIT_TEST_INDEX_THREADS=<n> enables exercising the multi-threaded loading
357 of the index for the whole test suite by bypassing the default number of
358 cache entries and thread minimums. Setting this to 1 will make the
359 index loading single threaded.
360
361 GIT_TEST_MULTI_PACK_INDEX=<boolean>, when true, forces the multi-pack-
362 index to be written after every 'git repack' command, and overrides the
363 'core.multiPackIndex' setting to true.
364
365 Naming Tests
366 ------------
367
368 The test files are named as:
369
370 tNNNN-commandname-details.sh
371
372 where N is a decimal digit.
373
374 First digit tells the family:
375
376 0 - the absolute basics and global stuff
377 1 - the basic commands concerning database
378 2 - the basic commands concerning the working tree
379 3 - the other basic commands (e.g. ls-files)
380 4 - the diff commands
381 5 - the pull and exporting commands
382 6 - the revision tree commands (even e.g. merge-base)
383 7 - the porcelainish commands concerning the working tree
384 8 - the porcelainish commands concerning forensics
385 9 - the git tools
386
387 Second digit tells the particular command we are testing.
388
389 Third digit (optionally) tells the particular switch or group of switches
390 we are testing.
391
392 If you create files under t/ directory (i.e. here) that is not
393 the top-level test script, never name the file to match the above
394 pattern. The Makefile here considers all such files as the
395 top-level test script and tries to run all of them. Care is
396 especially needed if you are creating a common test library
397 file, similar to test-lib.sh, because such a library file may
398 not be suitable for standalone execution.
399
400
401 Writing Tests
402 -------------
403
404 The test script is written as a shell script. It should start
405 with the standard "#!/bin/sh", and an
406 assignment to variable 'test_description', like this:
407
408 #!/bin/sh
409
410 test_description='xxx test (option --frotz)
411
412 This test registers the following structure in the cache
413 and tries to run git-ls-files with option --frotz.'
414
415
416 Source 'test-lib.sh'
417 --------------------
418
419 After assigning test_description, the test script should source
420 test-lib.sh like this:
421
422 . ./test-lib.sh
423
424 This test harness library does the following things:
425
426 - If the script is invoked with command line argument --help
427 (or -h), it shows the test_description and exits.
428
429 - Creates an empty test directory with an empty .git/objects database
430 and chdir(2) into it. This directory is 't/trash
431 directory.$test_name_without_dotsh', with t/ subject to change by
432 the --root option documented above.
433
434 - Defines standard test helper functions for your scripts to
435 use. These functions are designed to make all scripts behave
436 consistently when command line arguments --verbose (or -v),
437 --debug (or -d), and --immediate (or -i) is given.
438
439 Do's & don'ts
440 -------------
441
442 Here are a few examples of things you probably should and shouldn't do
443 when writing tests.
444
445 Here are the "do's:"
446
447 - Put all code inside test_expect_success and other assertions.
448
449 Even code that isn't a test per se, but merely some setup code
450 should be inside a test assertion.
451
452 - Chain your test assertions
453
454 Write test code like this:
455
456 git merge foo &&
457 git push bar &&
458 test ...
459
460 Instead of:
461
462 git merge hla
463 git push gh
464 test ...
465
466 That way all of the commands in your tests will succeed or fail. If
467 you must ignore the return value of something, consider using a
468 helper function (e.g. use sane_unset instead of unset, in order
469 to avoid unportable return value for unsetting a variable that was
470 already unset), or prepending the command with test_might_fail or
471 test_must_fail.
472
473 - Check the test coverage for your tests. See the "Test coverage"
474 below.
475
476 Don't blindly follow test coverage metrics; if a new function you added
477 doesn't have any coverage, then you're probably doing something wrong,
478 but having 100% coverage doesn't necessarily mean that you tested
479 everything.
480
481 Tests that are likely to smoke out future regressions are better
482 than tests that just inflate the coverage metrics.
483
484 - When a test checks for an absolute path that a git command generated,
485 construct the expected value using $(pwd) rather than $PWD,
486 $TEST_DIRECTORY, or $TRASH_DIRECTORY. It makes a difference on
487 Windows, where the shell (MSYS bash) mangles absolute path names.
488 For details, see the commit message of 4114156ae9.
489
490 - Remember that inside the <script> part, the standard output and
491 standard error streams are discarded, and the test harness only
492 reports "ok" or "not ok" to the end user running the tests. Under
493 --verbose, they are shown to help debug the tests.
494
495 And here are the "don'ts:"
496
497 - Don't exit() within a <script> part.
498
499 The harness will catch this as a programming error of the test.
500 Use test_done instead if you need to stop the tests early (see
501 "Skipping tests" below).
502
503 - Don't use '! git cmd' when you want to make sure the git command
504 exits with failure in a controlled way by calling "die()". Instead,
505 use 'test_must_fail git cmd'. This will signal a failure if git
506 dies in an unexpected way (e.g. segfault).
507
508 On the other hand, don't use test_must_fail for running regular
509 platform commands; just use '! cmd'. We are not in the business
510 of verifying that the world given to us sanely works.
511
512 - Don't feed the output of a git command to a pipe, as in:
513
514 git -C repo ls-files |
515 xargs -n 1 basename |
516 grep foo
517
518 which will discard git's exit code and may mask a crash. In the
519 above example, all exit codes are ignored except grep's.
520
521 Instead, write the output of that command to a temporary
522 file with ">" or assign it to a variable with "x=$(git ...)" rather
523 than pipe it.
524
525 - Don't use command substitution in a way that discards git's exit
526 code. When assigning to a variable, the exit code is not discarded,
527 e.g.:
528
529 x=$(git cat-file -p $sha) &&
530 ...
531
532 is OK because a crash in "git cat-file" will cause the "&&" chain
533 to fail, but:
534
535 test "refs/heads/foo" = "$(git symbolic-ref HEAD)"
536
537 is not OK and a crash in git could go undetected.
538
539 - Don't use perl without spelling it as "$PERL_PATH". This is to help
540 our friends on Windows where the platform Perl often adds CR before
541 the end of line, and they bundle Git with a version of Perl that
542 does not do so, whose path is specified with $PERL_PATH. Note that we
543 provide a "perl" function which uses $PERL_PATH under the hood, so
544 you do not need to worry when simply running perl in the test scripts
545 (but you do, for example, on a shebang line or in a sub script
546 created via "write_script").
547
548 - Don't use sh without spelling it as "$SHELL_PATH", when the script
549 can be misinterpreted by broken platform shell (e.g. Solaris).
550
551 - Don't chdir around in tests. It is not sufficient to chdir to
552 somewhere and then chdir back to the original location later in
553 the test, as any intermediate step can fail and abort the test,
554 causing the next test to start in an unexpected directory. Do so
555 inside a subshell if necessary.
556
557 - Don't save and verify the standard error of compound commands, i.e.
558 group commands, subshells, and shell functions (except test helper
559 functions like 'test_must_fail') like this:
560
561 ( cd dir && git cmd ) 2>error &&
562 test_cmp expect error
563
564 When running the test with '-x' tracing, then the trace of commands
565 executed in the compound command will be included in standard error
566 as well, quite possibly throwing off the subsequent checks examining
567 the output. Instead, save only the relevant git command's standard
568 error:
569
570 ( cd dir && git cmd 2>../error ) &&
571 test_cmp expect error
572
573 - Don't break the TAP output
574
575 The raw output from your test may be interpreted by a TAP harness. TAP
576 harnesses will ignore everything they don't know about, but don't step
577 on their toes in these areas:
578
579 - Don't print lines like "$x..$y" where $x and $y are integers.
580
581 - Don't print lines that begin with "ok" or "not ok".
582
583 TAP harnesses expect a line that begins with either "ok" and "not
584 ok" to signal a test passed or failed (and our harness already
585 produces such lines), so your script shouldn't emit such lines to
586 their output.
587
588 You can glean some further possible issues from the TAP grammar
589 (see https://metacpan.org/pod/TAP::Parser::Grammar#TAP-GRAMMAR)
590 but the best indication is to just run the tests with prove(1),
591 it'll complain if anything is amiss.
592
593
594 Skipping tests
595 --------------
596
597 If you need to skip tests you should do so by using the three-arg form
598 of the test_* functions (see the "Test harness library" section
599 below), e.g.:
600
601 test_expect_success PERL 'I need Perl' '
602 perl -e "hlagh() if unf_unf()"
603 '
604
605 The advantage of skipping tests like this is that platforms that don't
606 have the PERL and other optional dependencies get an indication of how
607 many tests they're missing.
608
609 If the test code is too hairy for that (i.e. does a lot of setup work
610 outside test assertions) you can also skip all remaining tests by
611 setting skip_all and immediately call test_done:
612
613 if ! test_have_prereq PERL
614 then
615 skip_all='skipping perl interface tests, perl not available'
616 test_done
617 fi
618
619 The string you give to skip_all will be used as an explanation for why
620 the test was skipped.
621
622 End with test_done
623 ------------------
624
625 Your script will be a sequence of tests, using helper functions
626 from the test harness library. At the end of the script, call
627 'test_done'.
628
629
630 Test harness library
631 --------------------
632
633 There are a handful helper functions defined in the test harness
634 library for your script to use.
635
636 - test_expect_success [<prereq>] <message> <script>
637
638 Usually takes two strings as parameters, and evaluates the
639 <script>. If it yields success, test is considered
640 successful. <message> should state what it is testing.
641
642 Example:
643
644 test_expect_success \
645 'git-write-tree should be able to write an empty tree.' \
646 'tree=$(git-write-tree)'
647
648 If you supply three parameters the first will be taken to be a
649 prerequisite; see the test_set_prereq and test_have_prereq
650 documentation below:
651
652 test_expect_success TTY 'git --paginate rev-list uses a pager' \
653 ' ... '
654
655 You can also supply a comma-separated list of prerequisites, in the
656 rare case where your test depends on more than one:
657
658 test_expect_success PERL,PYTHON 'yo dawg' \
659 ' test $(perl -E 'print eval "1 +" . qx[python -c "print 2"]') == "4" '
660
661 - test_expect_failure [<prereq>] <message> <script>
662
663 This is NOT the opposite of test_expect_success, but is used
664 to mark a test that demonstrates a known breakage. Unlike
665 the usual test_expect_success tests, which say "ok" on
666 success and "FAIL" on failure, this will say "FIXED" on
667 success and "still broken" on failure. Failures from these
668 tests won't cause -i (immediate) to stop.
669
670 Like test_expect_success this function can optionally use a three
671 argument invocation with a prerequisite as the first argument.
672
673 - test_debug <script>
674
675 This takes a single argument, <script>, and evaluates it only
676 when the test script is started with --debug command line
677 argument. This is primarily meant for use during the
678 development of a new test script.
679
680 - debug <git-command>
681
682 Run a git command inside a debugger. This is primarily meant for
683 use when debugging a failing test script.
684
685 - test_done
686
687 Your test script must have test_done at the end. Its purpose
688 is to summarize successes and failures in the test script and
689 exit with an appropriate error code.
690
691 - test_tick
692
693 Make commit and tag names consistent by setting the author and
694 committer times to defined state. Subsequent calls will
695 advance the times by a fixed amount.
696
697 - test_commit <message> [<filename> [<contents>]]
698
699 Creates a commit with the given message, committing the given
700 file with the given contents (default for both is to reuse the
701 message string), and adds a tag (again reusing the message
702 string as name). Calls test_tick to make the SHA-1s
703 reproducible.
704
705 - test_merge <message> <commit-or-tag>
706
707 Merges the given rev using the given message. Like test_commit,
708 creates a tag and calls test_tick before committing.
709
710 - test_set_prereq <prereq>
711
712 Set a test prerequisite to be used later with test_have_prereq. The
713 test-lib will set some prerequisites for you, see the
714 "Prerequisites" section below for a full list of these.
715
716 Others you can set yourself and use later with either
717 test_have_prereq directly, or the three argument invocation of
718 test_expect_success and test_expect_failure.
719
720 - test_have_prereq <prereq>
721
722 Check if we have a prerequisite previously set with test_set_prereq.
723 The most common way to use this explicitly (as opposed to the
724 implicit use when an argument is passed to test_expect_*) is to skip
725 all the tests at the start of the test script if we don't have some
726 essential prerequisite:
727
728 if ! test_have_prereq PERL
729 then
730 skip_all='skipping perl interface tests, perl not available'
731 test_done
732 fi
733
734 - test_external [<prereq>] <message> <external> <script>
735
736 Execute a <script> with an <external> interpreter (like perl). This
737 was added for tests like t9700-perl-git.sh which do most of their
738 work in an external test script.
739
740 test_external \
741 'GitwebCache::*FileCache*' \
742 perl "$TEST_DIRECTORY"/t9503/test_cache_interface.pl
743
744 If the test is outputting its own TAP you should set the
745 test_external_has_tap variable somewhere before calling the first
746 test_external* function. See t9700-perl-git.sh for an example.
747
748 # The external test will outputs its own plan
749 test_external_has_tap=1
750
751 - test_external_without_stderr [<prereq>] <message> <external> <script>
752
753 Like test_external but fail if there's any output on stderr,
754 instead of checking the exit code.
755
756 test_external_without_stderr \
757 'Perl API' \
758 perl "$TEST_DIRECTORY"/t9700/test.pl
759
760 - test_expect_code <exit-code> <command>
761
762 Run a command and ensure that it exits with the given exit code.
763 For example:
764
765 test_expect_success 'Merge with d/f conflicts' '
766 test_expect_code 1 git merge "merge msg" B master
767 '
768
769 - test_must_fail [<options>] <git-command>
770
771 Run a git command and ensure it fails in a controlled way. Use
772 this instead of "! <git-command>". When git-command dies due to a
773 segfault, test_must_fail diagnoses it as an error; "! <git-command>"
774 treats it as just another expected failure, which would let such a
775 bug go unnoticed.
776
777 Accepts the following options:
778
779 ok=<signal-name>[,<...>]:
780 Don't treat an exit caused by the given signal as error.
781 Multiple signals can be specified as a comma separated list.
782 Currently recognized signal names are: sigpipe, success.
783 (Don't use 'success', use 'test_might_fail' instead.)
784
785 - test_might_fail [<options>] <git-command>
786
787 Similar to test_must_fail, but tolerate success, too. Use this
788 instead of "<git-command> || :" to catch failures due to segv.
789
790 Accepts the same options as test_must_fail.
791
792 - test_cmp <expected> <actual>
793
794 Check whether the content of the <actual> file matches the
795 <expected> file. This behaves like "cmp" but produces more
796 helpful output when the test is run with "-v" option.
797
798 - test_cmp_rev <expected> <actual>
799
800 Check whether the <expected> rev points to the same commit as the
801 <actual> rev.
802
803 - test_line_count (= | -lt | -ge | ...) <length> <file>
804
805 Check whether a file has the length it is expected to.
806
807 - test_path_is_file <path> [<diagnosis>]
808 test_path_is_dir <path> [<diagnosis>]
809 test_path_is_missing <path> [<diagnosis>]
810
811 Check if the named path is a file, if the named path is a
812 directory, or if the named path does not exist, respectively,
813 and fail otherwise, showing the <diagnosis> text.
814
815 - test_when_finished <script>
816
817 Prepend <script> to a list of commands to run to clean up
818 at the end of the current test. If some clean-up command
819 fails, the test will not pass.
820
821 Example:
822
823 test_expect_success 'branch pointing to non-commit' '
824 git rev-parse HEAD^{tree} >.git/refs/heads/invalid &&
825 test_when_finished "git update-ref -d refs/heads/invalid" &&
826 ...
827 '
828
829 - test_write_lines <lines>
830
831 Write <lines> on standard output, one line per argument.
832 Useful to prepare multi-line files in a compact form.
833
834 Example:
835
836 test_write_lines a b c d e f g >foo
837
838 Is a more compact equivalent of:
839 cat >foo <<-EOF
840 a
841 b
842 c
843 d
844 e
845 f
846 g
847 EOF
848
849
850 - test_pause
851
852 This command is useful for writing and debugging tests and must be
853 removed before submitting. It halts the execution of the test and
854 spawns a shell in the trash directory. Exit the shell to continue
855 the test. Example:
856
857 test_expect_success 'test' '
858 git do-something >actual &&
859 test_pause &&
860 test_cmp expected actual
861 '
862
863 - test_ln_s_add <path1> <path2>
864
865 This function helps systems whose filesystem does not support symbolic
866 links. Use it to add a symbolic link entry to the index when it is not
867 important that the file system entry is a symbolic link, i.e., instead
868 of the sequence
869
870 ln -s foo bar &&
871 git add bar
872
873 Sometimes it is possible to split a test in a part that does not need
874 the symbolic link in the file system and a part that does; then only
875 the latter part need be protected by a SYMLINKS prerequisite (see below).
876
877 - test_oid_init
878
879 This function loads facts and useful object IDs related to the hash
880 algorithm(s) in use from the files in t/oid-info.
881
882 - test_oid_cache
883
884 This function reads per-hash algorithm information from standard
885 input (usually a heredoc) in the format described in
886 t/oid-info/README. This is useful for test-specific values, such as
887 object IDs, which must vary based on the hash algorithm.
888
889 Certain fixed values, such as hash sizes and common placeholder
890 object IDs, can be loaded with test_oid_init (described above).
891
892 - test_oid <key>
893
894 This function looks up a value for the hash algorithm in use, based
895 on the key given. The value must have been loaded using
896 test_oid_init or test_oid_cache. Providing an unknown key is an
897 error.
898
899 Prerequisites
900 -------------
901
902 These are the prerequisites that the test library predefines with
903 test_have_prereq.
904
905 See the prereq argument to the test_* functions in the "Test harness
906 library" section above and the "test_have_prereq" function for how to
907 use these, and "test_set_prereq" for how to define your own.
908
909 - PYTHON
910
911 Git wasn't compiled with NO_PYTHON=YesPlease. Wrap any tests that
912 need Python with this.
913
914 - PERL
915
916 Git wasn't compiled with NO_PERL=YesPlease.
917
918 Even without the PERL prerequisite, tests can assume there is a
919 usable perl interpreter at $PERL_PATH, though it need not be
920 particularly modern.
921
922 - POSIXPERM
923
924 The filesystem supports POSIX style permission bits.
925
926 - BSLASHPSPEC
927
928 Backslashes in pathspec are not directory separators. This is not
929 set on Windows. See 6fd1106a for details.
930
931 - EXECKEEPSPID
932
933 The process retains the same pid across exec(2). See fb9a2bea for
934 details.
935
936 - PIPE
937
938 The filesystem we're on supports creation of FIFOs (named pipes)
939 via mkfifo(1).
940
941 - SYMLINKS
942
943 The filesystem we're on supports symbolic links. E.g. a FAT
944 filesystem doesn't support these. See 704a3143 for details.
945
946 - SANITY
947
948 Test is not run by root user, and an attempt to write to an
949 unwritable file is expected to fail correctly.
950
951 - PCRE
952
953 Git was compiled with support for PCRE. Wrap any tests
954 that use git-grep --perl-regexp or git-grep -P in these.
955
956 - LIBPCRE1
957
958 Git was compiled with PCRE v1 support via
959 USE_LIBPCRE1=YesPlease. Wrap any PCRE using tests that for some
960 reason need v1 of the PCRE library instead of v2 in these.
961
962 - LIBPCRE2
963
964 Git was compiled with PCRE v2 support via
965 USE_LIBPCRE2=YesPlease. Wrap any PCRE using tests that for some
966 reason need v2 of the PCRE library instead of v1 in these.
967
968 - CASE_INSENSITIVE_FS
969
970 Test is run on a case insensitive file system.
971
972 - UTF8_NFD_TO_NFC
973
974 Test is run on a filesystem which converts decomposed utf-8 (nfd)
975 to precomposed utf-8 (nfc).
976
977 - PTHREADS
978
979 Git wasn't compiled with NO_PTHREADS=YesPlease.
980
981 Tips for Writing Tests
982 ----------------------
983
984 As with any programming projects, existing programs are the best
985 source of the information. However, do _not_ emulate
986 t0000-basic.sh when writing your tests. The test is special in
987 that it tries to validate the very core of GIT. For example, it
988 knows that there will be 256 subdirectories under .git/objects/,
989 and it knows that the object ID of an empty tree is a certain
990 40-byte string. This is deliberately done so in t0000-basic.sh
991 because the things the very basic core test tries to achieve is
992 to serve as a basis for people who are changing the GIT internal
993 drastically. For these people, after making certain changes,
994 not seeing failures from the basic test _is_ a failure. And
995 such drastic changes to the core GIT that even changes these
996 otherwise supposedly stable object IDs should be accompanied by
997 an update to t0000-basic.sh.
998
999 However, other tests that simply rely on basic parts of the core
1000 GIT working properly should not have that level of intimate
1001 knowledge of the core GIT internals. If all the test scripts
1002 hardcoded the object IDs like t0000-basic.sh does, that defeats
1003 the purpose of t0000-basic.sh, which is to isolate that level of
1004 validation in one place. Your test also ends up needing
1005 updating when such a change to the internal happens, so do _not_
1006 do it and leave the low level of validation to t0000-basic.sh.
1007
1008 Test coverage
1009 -------------
1010
1011 You can use the coverage tests to find code paths that are not being
1012 used or properly exercised yet.
1013
1014 To do that, run the coverage target at the top-level (not in the t/
1015 directory):
1016
1017 make coverage
1018
1019 That'll compile Git with GCC's coverage arguments, and generate a test
1020 report with gcov after the tests finish. Running the coverage tests
1021 can take a while, since running the tests in parallel is incompatible
1022 with GCC's coverage mode.
1023
1024 After the tests have run you can generate a list of untested
1025 functions:
1026
1027 make coverage-untested-functions
1028
1029 You can also generate a detailed per-file HTML report using the
1030 Devel::Cover module. To install it do:
1031
1032 # On Debian or Ubuntu:
1033 sudo aptitude install libdevel-cover-perl
1034
1035 # From the CPAN with cpanminus
1036 curl -L http://cpanmin.us | perl - --sudo --self-upgrade
1037 cpanm --sudo Devel::Cover
1038
1039 Then, at the top-level:
1040
1041 make cover_db_html
1042
1043 That'll generate a detailed cover report in the "cover_db_html"
1044 directory, which you can then copy to a webserver, or inspect locally
1045 in a browser.