test-lib: introduce 'test_atexit'
[git/git.git] / t / README
CommitLineData
986aa7f1
JH
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 ***
5099b99d
ÆAB
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
986aa7f1 24 ...
5099b99d
ÆAB
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
85b0b34e 36be run with any TAP harness. Here's an example of parallel testing
5099b99d
ÆAB
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
986aa7f1 52
28d836c8
MG
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
5099b99d 59You can also run each test individually from command line, like this:
986aa7f1 60
5099b99d
ÆAB
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
986aa7f1
JH
69
70You can pass --verbose (or -v), --debug (or -d), and --immediate
4e1be63c
JS
71(or -i) command line argument to the test, or by setting GIT_TEST_OPTS
72appropriately before running "make".
986aa7f1 73
5e3b4fce 74-v::
986aa7f1
JH
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
ff09af3f
TR
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
a136f6d8
JK
85-x::
86 Turn on shell tracing (i.e., `set -x`) during the tests
94201a2b 87 themselves. Implies `--verbose`.
5fc98e79
SG
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.
a136f6d8 91
5e3b4fce 92-d::
986aa7f1
JH
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.
0986de94
PK
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.
986aa7f1 100
5e3b4fce 101-i::
986aa7f1
JH
102--immediate::
103 This causes the test to immediately exit upon the first
13cb3bb7
SR
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.
986aa7f1 108
5e3b4fce 109-l::
5e2c08c6
LW
110--long-tests::
111 This causes additional long-running tests to be run (where
112 available), for more exhaustive testing.
113
0445e6f0
IB
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
952af351
TR
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).
986aa7f1 124
3da93652
JS
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
952af351
TR
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
95d9d5ec
TR
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
952af351 139 Note that memcheck is run with the option --leak-check=no,
9aec68d3
CMN
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
5dfc368f
TR
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
44138559
JS
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
a5f52c6d 157-V::
452320f1
JK
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
e4597aae
MO
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
dd167a30
JS
173--no-bin-wrappers::
174 By default, the test suite uses the wrappers in
175 `../bin-wrappers/` to execute `git` and friends. With this option,
176 `../git` and friends are run directly. This is not recommended
177 in general, as the wrappers contain safeguards to ensure that no
178 files from an installed Git are used, but can speed up test runs
179 especially on platforms where running shell scripts is expensive
180 (most notably, Windows).
181
0d4dbcd3
TR
182--root=<directory>::
183 Create "trash" directories used to store all temporary data during
184 testing under <directory>, instead of the t/ directory.
185 Using this option with a RAM-based filesystem (such as tmpfs)
186 can massively speed up the test suite.
187
bb79af9d
JK
188--chain-lint::
189--no-chain-lint::
190 If --chain-lint is enabled, the test harness will check each
191 test to make sure that it properly "&&-chains" all commands (so
192 that a failure in the middle does not go unnoticed by the final
193 exit code of the test). This check is performed in addition to
194 running the tests themselves. You may also enable or disable
195 this feature by setting the GIT_TEST_CHAIN_LINT environment
196 variable to "1" or "0", respectively.
197
fb7d1e3a
SG
198--stress::
199--stress=<N>::
200 Run the test script repeatedly in multiple parallel jobs until
201 one of them fails. Useful for reproducing rare failures in
202 flaky tests. The number of parallel jobs is, in order of
203 precedence: <N>, or the value of the GIT_TEST_STRESS_LOAD
204 environment variable, or twice the number of available
205 processors (as shown by the 'getconf' utility), or 8.
206 Implies `--verbose -x --immediate` to get the most information
207 about the failure. Note that the verbose output of each test
208 job is saved to 't/test-results/$TEST_NAME.stress-<nr>.out',
209 and only the output of the failed test job is shown on the
210 terminal. The names of the trash directories get a
211 '.stress-<nr>' suffix, and the trash directory of the failed
212 test job is renamed to end with a '.stress-failed' suffix.
213
76e27fbf
SG
214--stress-limit=<N>::
215 When combined with --stress run the test script repeatedly
216 this many times in each of the parallel jobs or until one of
217 them fails, whichever comes first.
218
e160da7f
MO
219You can also set the GIT_TEST_INSTALLED environment variable to
220the bindir of an existing git installation to test that installation.
221You still need to have built this git sandbox, from which various
222test-* support programs, templates, and perl libraries are used.
223If your installed git is incomplete, it will silently test parts of
224your built version instead.
225
226When using GIT_TEST_INSTALLED, you can also set GIT_TEST_EXEC_PATH to
227override the location of the dashed-form subcommands (what
228GIT_EXEC_PATH would be used for during normal operation).
229GIT_TEST_EXEC_PATH defaults to `$GIT_TEST_INSTALLED/git --exec-path`.
230
231
fbd458a3
JN
232Skipping Tests
233--------------
234
235In some environments, certain tests have no way of succeeding
236due to platform limitation, such as lack of 'unzip' program, or
237filesystem that do not allow arbitrary sequence of non-NUL bytes
238as pathnames.
239
240You should be able to say something like
241
242 $ GIT_SKIP_TESTS=t9200.8 sh ./t9200-git-cvsexport-commit.sh
243
244and even:
245
246 $ GIT_SKIP_TESTS='t[0-4]??? t91?? t9200.8' make
247
248to omit such tests. The value of the environment variable is a
249SP separated list of patterns that tells which tests to skip,
250and either can match the "t[0-9]{4}" part to skip the whole
251test, or t[0-9]{4} followed by ".$number" to say which
252particular test to skip.
253
0445e6f0
IB
254For an individual test suite --run could be used to specify that
255only some tests should be run or that some tests should be
256excluded from a run.
257
258The argument for --run is a list of individual test numbers or
259ranges with an optional negation prefix that define what tests in
260a test suite to include in the run. A range is two numbers
261separated with a dash and matches a range of tests with both ends
262been included. You may omit the first or the second number to
263mean "from the first test" or "up to the very last test"
264respectively.
265
266Optional prefix of '!' means that the test or a range of tests
267should be excluded from the run.
268
269If --run starts with an unprefixed number or range the initial
270set of tests to run is empty. If the first item starts with '!'
271all the tests are added to the initial set. After initial set is
272determined every test number or range is added or excluded from
273the set one by one, from left to right.
274
275Individual numbers or ranges could be separated either by a space
276or a comma.
277
278For example, to run only tests up to a specific test (21), one
279could do this:
280
281 $ sh ./t9200-git-cvsexport-commit.sh --run='1-21'
282
283or this:
284
285 $ sh ./t9200-git-cvsexport-commit.sh --run='-21'
286
287Common case is to run several setup tests (1, 2, 3) and then a
288specific test (21) that relies on that setup:
289
290 $ sh ./t9200-git-cvsexport-commit.sh --run='1 2 3 21'
291
292or:
293
294 $ sh ./t9200-git-cvsexport-commit.sh --run=1,2,3,21
295
296or:
297
298 $ sh ./t9200-git-cvsexport-commit.sh --run='-3 21'
299
01e4be6c
KS
300As noted above, the test set is built by going through the items
301from left to right, so this:
0445e6f0
IB
302
303 $ sh ./t9200-git-cvsexport-commit.sh --run='1-4 !3'
304
01e4be6c 305will run tests 1, 2, and 4. Items that come later have higher
2e3a16b2 306precedence. It means that this:
0445e6f0
IB
307
308 $ sh ./t9200-git-cvsexport-commit.sh --run='!3 1-4'
309
310would just run tests from 1 to 4, including 3.
311
312You may use negation with ranges. The following will run all
313test in the test suite except from 7 up to 11:
314
315 $ sh ./t9200-git-cvsexport-commit.sh --run='!7-11'
316
317Some tests in a test suite rely on the previous tests performing
318certain actions, specifically some tests are designated as
319"setup" test, so you cannot _arbitrarily_ disable one test and
320expect the rest to function correctly.
321
322--run is mostly useful when you want to focus on a specific test
323and know what setup is needed for it. Or when you want to run
324everything up to a certain test.
fbd458a3
JN
325
326
4c2db938
NTND
327Running tests with special setups
328---------------------------------
329
330The whole test suite could be run to test some special features
331that cannot be easily covered by a few specific test cases. These
332could be enabled by running the test suite with correct GIT_TEST_
333environment set.
334
6cdccfce
ÆAB
335GIT_TEST_GETTEXT_POISON=<non-empty?> turns all strings marked for
336translation into gibberish if non-empty (think "test -n"). Used for
337spotting those tests that need to be marked with a C_LOCALE_OUTPUT
338prerequisite when adding more strings for translation. See "Testing
339marked strings" in po/README for details.
340
4c2db938
NTND
341GIT_TEST_SPLIT_INDEX=<boolean> forces split-index mode on the whole
342test suite. Accept any boolean values that are accepted by git-config.
343
43fa44fa
NTND
344GIT_TEST_FULL_IN_PACK_ARRAY=<boolean> exercises the uncommon
345pack-objects code path where there are more than 1024 packs even if
346the actual number of packs in repository is below this limit. Accept
347any boolean values that are accepted by git-config.
348
ac77d0c3
NTND
349GIT_TEST_OE_SIZE=<n> exercises the uncommon pack-objects code path
350where we do not cache object size in memory and read it from existing
351packs on demand. This normally only happens when the object size is
352over 2GB. This variable forces the code path on any object larger than
353<n> bytes.
354
ac6e12f9 355GIT_TEST_OE_DELTA_SIZE=<n> exercises the uncommon pack-objects code
9ac3f0e5
NTND
356path where deltas larger than this limit require extra memory
357allocation for bookkeeping.
358
5f4436a7
NTND
359GIT_TEST_VALIDATE_INDEX_CACHE_ENTRIES=<boolean> checks that cache-tree
360records are valid when the index is written out or after a merge. This
361is mostly to catch missing invalidation. Default is true.
362
859fdc0c
DS
363GIT_TEST_COMMIT_GRAPH=<boolean>, when true, forces the commit-graph to
364be written after every 'git commit' command, and overrides the
365'core.commitGraph' setting to true.
366
4cb54d0a
BP
367GIT_TEST_FSMONITOR=$PWD/t7519/fsmonitor-all exercises the fsmonitor
368code path for utilizing a file system monitor to speed up detecting
369new or changed files.
370
1f357b04
BP
371GIT_TEST_INDEX_VERSION=<n> exercises the index read/write code path
372for the index version specified. Can be set to any valid version
373(currently 2, 3, or 4).
374
99dbbfa8
DS
375GIT_TEST_PACK_SPARSE=<boolean> if enabled will default the pack-objects
376builtin to use the sparse object walk. This can still be overridden by
377the --no-sparse command-line argument.
378
5765d97b
BP
379GIT_TEST_PRELOAD_INDEX=<boolean> exercises the preload-index code path
380by overriding the minimum number of cache entries required per thread.
381
62c23938
ÆAB
382GIT_TEST_REBASE_USE_BUILTIN=<boolean>, when false, disables the
383builtin version of git-rebase. See 'rebase.useBuiltin' in
384git-config(1).
385
c780b9cf
BP
386GIT_TEST_INDEX_THREADS=<n> enables exercising the multi-threaded loading
387of the index for the whole test suite by bypassing the default number of
388cache entries and thread minimums. Setting this to 1 will make the
389index loading single threaded.
390
0465a505
DS
391GIT_TEST_MULTI_PACK_INDEX=<boolean>, when true, forces the multi-pack-
392index to be written after every 'git repack' command, and overrides the
393'core.multiPackIndex' setting to true.
394
07c3c2aa
JT
395GIT_TEST_SIDEBAND_ALL=<boolean>, when true, overrides the
396'uploadpack.allowSidebandAll' setting to true, and when false, forces
397fetch-pack to not request sideband-all (even if the server advertises
398sideband-all).
399
f50c9f76
PB
400Naming Tests
401------------
402
403The test files are named as:
404
405 tNNNN-commandname-details.sh
406
407where N is a decimal digit.
408
409First digit tells the family:
410
411 0 - the absolute basics and global stuff
412 1 - the basic commands concerning database
413 2 - the basic commands concerning the working tree
414 3 - the other basic commands (e.g. ls-files)
415 4 - the diff commands
416 5 - the pull and exporting commands
417 6 - the revision tree commands (even e.g. merge-base)
8f4a9b62 418 7 - the porcelainish commands concerning the working tree
8757749e
JN
419 8 - the porcelainish commands concerning forensics
420 9 - the git tools
f50c9f76
PB
421
422Second digit tells the particular command we are testing.
423
424Third digit (optionally) tells the particular switch or group of switches
425we are testing.
426
77656600
JH
427If you create files under t/ directory (i.e. here) that is not
428the top-level test script, never name the file to match the above
429pattern. The Makefile here considers all such files as the
63d32945 430top-level test script and tries to run all of them. Care is
77656600
JH
431especially needed if you are creating a common test library
432file, similar to test-lib.sh, because such a library file may
433not be suitable for standalone execution.
434
f50c9f76 435
986aa7f1
JH
436Writing Tests
437-------------
438
439The test script is written as a shell script. It should start
51b7a525 440with the standard "#!/bin/sh", and an
986aa7f1
JH
441assignment to variable 'test_description', like this:
442
443 #!/bin/sh
986aa7f1 444
14cd1ff3 445 test_description='xxx test (option --frotz)
986aa7f1
JH
446
447 This test registers the following structure in the cache
448 and tries to run git-ls-files with option --frotz.'
449
f50c9f76 450
986aa7f1
JH
451Source 'test-lib.sh'
452--------------------
453
454After assigning test_description, the test script should source
455test-lib.sh like this:
456
457 . ./test-lib.sh
458
459This test harness library does the following things:
460
461 - If the script is invoked with command line argument --help
462 (or -h), it shows the test_description and exits.
463
e1ca1c9d
ÆAB
464 - Creates an empty test directory with an empty .git/objects database
465 and chdir(2) into it. This directory is 't/trash
466 directory.$test_name_without_dotsh', with t/ subject to change by
fb7d1e3a
SG
467 the --root option documented above, and a '.stress-<N>' suffix
468 appended by the --stress option.
986aa7f1
JH
469
470 - Defines standard test helper functions for your scripts to
471 use. These functions are designed to make all scripts behave
472 consistently when command line arguments --verbose (or -v),
473 --debug (or -d), and --immediate (or -i) is given.
474
441ee35d
MD
475Do's & don'ts
476-------------
20873f45 477
6fd45295 478Here are a few examples of things you probably should and shouldn't do
20873f45
ÆAB
479when writing tests.
480
441ee35d 481Here are the "do's:"
20873f45 482
6fd45295 483 - Put all code inside test_expect_success and other assertions.
20873f45
ÆAB
484
485 Even code that isn't a test per se, but merely some setup code
6fd45295 486 should be inside a test assertion.
20873f45
ÆAB
487
488 - Chain your test assertions
489
490 Write test code like this:
491
492 git merge foo &&
493 git push bar &&
494 test ...
495
496 Instead of:
497
498 git merge hla
499 git push gh
500 test ...
501
502 That way all of the commands in your tests will succeed or fail. If
00648ba0
EN
503 you must ignore the return value of something, consider using a
504 helper function (e.g. use sane_unset instead of unset, in order
505 to avoid unportable return value for unsetting a variable that was
506 already unset), or prepending the command with test_might_fail or
507 test_must_fail.
20873f45 508
0c357544
ÆAB
509 - Check the test coverage for your tests. See the "Test coverage"
510 below.
511
63d32945
MW
512 Don't blindly follow test coverage metrics; if a new function you added
513 doesn't have any coverage, then you're probably doing something wrong,
e8b55f5c
ÆAB
514 but having 100% coverage doesn't necessarily mean that you tested
515 everything.
516
517 Tests that are likely to smoke out future regressions are better
518 than tests that just inflate the coverage metrics.
519
95b104c8
JS
520 - When a test checks for an absolute path that a git command generated,
521 construct the expected value using $(pwd) rather than $PWD,
522 $TEST_DIRECTORY, or $TRASH_DIRECTORY. It makes a difference on
523 Windows, where the shell (MSYS bash) mangles absolute path names.
524 For details, see the commit message of 4114156ae9.
525
441ee35d
MD
526 - Remember that inside the <script> part, the standard output and
527 standard error streams are discarded, and the test harness only
528 reports "ok" or "not ok" to the end user running the tests. Under
529 --verbose, they are shown to help debug the tests.
530
531And here are the "don'ts:"
20873f45 532
441ee35d 533 - Don't exit() within a <script> part.
20873f45
ÆAB
534
535 The harness will catch this as a programming error of the test.
536 Use test_done instead if you need to stop the tests early (see
537 "Skipping tests" below).
538
441ee35d
MD
539 - Don't use '! git cmd' when you want to make sure the git command
540 exits with failure in a controlled way by calling "die()". Instead,
ad78585e
JH
541 use 'test_must_fail git cmd'. This will signal a failure if git
542 dies in an unexpected way (e.g. segfault).
543
f445500e 544 On the other hand, don't use test_must_fail for running regular
53de7424
JH
545 platform commands; just use '! cmd'. We are not in the business
546 of verifying that the world given to us sanely works.
f445500e 547
a378fee5
MD
548 - Don't feed the output of a git command to a pipe, as in:
549
550 git -C repo ls-files |
551 xargs -n 1 basename |
552 grep foo
553
554 which will discard git's exit code and may mask a crash. In the
555 above example, all exit codes are ignored except grep's.
556
557 Instead, write the output of that command to a temporary
558 file with ">" or assign it to a variable with "x=$(git ...)" rather
559 than pipe it.
560
561 - Don't use command substitution in a way that discards git's exit
562 code. When assigning to a variable, the exit code is not discarded,
563 e.g.:
564
565 x=$(git cat-file -p $sha) &&
566 ...
567
568 is OK because a crash in "git cat-file" will cause the "&&" chain
569 to fail, but:
570
571 test "refs/heads/foo" = "$(git symbolic-ref HEAD)"
572
573 is not OK and a crash in git could go undetected.
574
441ee35d
MD
575 - Don't use perl without spelling it as "$PERL_PATH". This is to help
576 our friends on Windows where the platform Perl often adds CR before
ad78585e 577 the end of line, and they bundle Git with a version of Perl that
a0e0ec9f
JK
578 does not do so, whose path is specified with $PERL_PATH. Note that we
579 provide a "perl" function which uses $PERL_PATH under the hood, so
580 you do not need to worry when simply running perl in the test scripts
581 (but you do, for example, on a shebang line or in a sub script
582 created via "write_script").
ad78585e 583
441ee35d
MD
584 - Don't use sh without spelling it as "$SHELL_PATH", when the script
585 can be misinterpreted by broken platform shell (e.g. Solaris).
ad78585e 586
441ee35d 587 - Don't chdir around in tests. It is not sufficient to chdir to
ad78585e
JH
588 somewhere and then chdir back to the original location later in
589 the test, as any intermediate step can fail and abort the test,
590 causing the next test to start in an unexpected directory. Do so
591 inside a subshell if necessary.
592
441ee35d
MD
593 - Don't save and verify the standard error of compound commands, i.e.
594 group commands, subshells, and shell functions (except test helper
94201a2b
SG
595 functions like 'test_must_fail') like this:
596
597 ( cd dir && git cmd ) 2>error &&
598 test_cmp expect error
599
600 When running the test with '-x' tracing, then the trace of commands
601 executed in the compound command will be included in standard error
602 as well, quite possibly throwing off the subsequent checks examining
603 the output. Instead, save only the relevant git command's standard
604 error:
605
606 ( cd dir && git cmd 2>../error ) &&
607 test_cmp expect error
608
441ee35d 609 - Don't break the TAP output
20873f45 610
6fd45295
JH
611 The raw output from your test may be interpreted by a TAP harness. TAP
612 harnesses will ignore everything they don't know about, but don't step
613 on their toes in these areas:
20873f45
ÆAB
614
615 - Don't print lines like "$x..$y" where $x and $y are integers.
616
617 - Don't print lines that begin with "ok" or "not ok".
618
6fd45295 619 TAP harnesses expect a line that begins with either "ok" and "not
20873f45
ÆAB
620 ok" to signal a test passed or failed (and our harness already
621 produces such lines), so your script shouldn't emit such lines to
622 their output.
623
624 You can glean some further possible issues from the TAP grammar
c1d44cee 625 (see https://metacpan.org/pod/TAP::Parser::Grammar#TAP-GRAMMAR)
20873f45
ÆAB
626 but the best indication is to just run the tests with prove(1),
627 it'll complain if anything is amiss.
628
20873f45 629
b5500d16
ÆAB
630Skipping tests
631--------------
632
681186ae 633If you need to skip tests you should do so by using the three-arg form
99d9050d
ÆAB
634of the test_* functions (see the "Test harness library" section
635below), e.g.:
636
ad78585e 637 test_expect_success PERL 'I need Perl' '
a0e0ec9f 638 perl -e "hlagh() if unf_unf()"
ad78585e 639 '
99d9050d
ÆAB
640
641The advantage of skipping tests like this is that platforms that don't
642have the PERL and other optional dependencies get an indication of how
643many tests they're missing.
644
645If the test code is too hairy for that (i.e. does a lot of setup work
646outside test assertions) you can also skip all remaining tests by
647setting skip_all and immediately call test_done:
b5500d16
ÆAB
648
649 if ! test_have_prereq PERL
650 then
651 skip_all='skipping perl interface tests, perl not available'
652 test_done
653 fi
14cd1ff3 654
99d9050d
ÆAB
655The string you give to skip_all will be used as an explanation for why
656the test was skipped.
657
986aa7f1
JH
658End with test_done
659------------------
660
661Your script will be a sequence of tests, using helper functions
662from the test harness library. At the end of the script, call
663'test_done'.
664
665
666Test harness library
667--------------------
668
669There are a handful helper functions defined in the test harness
670library for your script to use.
671
9a897893 672 - test_expect_success [<prereq>] <message> <script>
986aa7f1 673
72942a61 674 Usually takes two strings as parameters, and evaluates the
986aa7f1
JH
675 <script>. If it yields success, test is considered
676 successful. <message> should state what it is testing.
677
678 Example:
679
680 test_expect_success \
681 'git-write-tree should be able to write an empty tree.' \
682 'tree=$(git-write-tree)'
683
9a897893 684 If you supply three parameters the first will be taken to be a
72942a61 685 prerequisite; see the test_set_prereq and test_have_prereq
9a897893
ÆAB
686 documentation below:
687
688 test_expect_success TTY 'git --paginate rev-list uses a pager' \
689 ' ... '
690
93a57246
ÆAB
691 You can also supply a comma-separated list of prerequisites, in the
692 rare case where your test depends on more than one:
693
694 test_expect_success PERL,PYTHON 'yo dawg' \
695 ' test $(perl -E 'print eval "1 +" . qx[python -c "print 2"]') == "4" '
696
9a897893 697 - test_expect_failure [<prereq>] <message> <script>
986aa7f1 698
41ac414e
JH
699 This is NOT the opposite of test_expect_success, but is used
700 to mark a test that demonstrates a known breakage. Unlike
701 the usual test_expect_success tests, which say "ok" on
702 success and "FAIL" on failure, this will say "FIXED" on
703 success and "still broken" on failure. Failures from these
704 tests won't cause -i (immediate) to stop.
986aa7f1 705
9a897893
ÆAB
706 Like test_expect_success this function can optionally use a three
707 argument invocation with a prerequisite as the first argument.
708
986aa7f1
JH
709 - test_debug <script>
710
711 This takes a single argument, <script>, and evaluates it only
712 when the test script is started with --debug command line
713 argument. This is primarily meant for use during the
714 development of a new test script.
715
6a94088c
JS
716 - debug <git-command>
717
718 Run a git command inside a debugger. This is primarily meant for
719 use when debugging a failing test script.
720
986aa7f1
JH
721 - test_done
722
723 Your test script must have test_done at the end. Its purpose
724 is to summarize successes and failures in the test script and
725 exit with an appropriate error code.
726
00884968
JS
727 - test_tick
728
729 Make commit and tag names consistent by setting the author and
63d32945 730 committer times to defined state. Subsequent calls will
00884968
JS
731 advance the times by a fixed amount.
732
733 - test_commit <message> [<filename> [<contents>]]
734
735 Creates a commit with the given message, committing the given
736 file with the given contents (default for both is to reuse the
737 message string), and adds a tag (again reusing the message
738 string as name). Calls test_tick to make the SHA-1s
739 reproducible.
740
741 - test_merge <message> <commit-or-tag>
742
743 Merges the given rev using the given message. Like test_commit,
744 creates a tag and calls test_tick before committing.
986aa7f1 745
72942a61 746 - test_set_prereq <prereq>
9a897893
ÆAB
747
748 Set a test prerequisite to be used later with test_have_prereq. The
be53deef
ÆAB
749 test-lib will set some prerequisites for you, see the
750 "Prerequisites" section below for a full list of these.
751
752 Others you can set yourself and use later with either
753 test_have_prereq directly, or the three argument invocation of
754 test_expect_success and test_expect_failure.
9a897893 755
72942a61 756 - test_have_prereq <prereq>
9a897893 757
4473060b
ÆAB
758 Check if we have a prerequisite previously set with test_set_prereq.
759 The most common way to use this explicitly (as opposed to the
760 implicit use when an argument is passed to test_expect_*) is to skip
761 all the tests at the start of the test script if we don't have some
762 essential prerequisite:
9a897893
ÆAB
763
764 if ! test_have_prereq PERL
765 then
766 skip_all='skipping perl interface tests, perl not available'
767 test_done
768 fi
769
2fac6a4b
ÆAB
770 - test_external [<prereq>] <message> <external> <script>
771
772 Execute a <script> with an <external> interpreter (like perl). This
773 was added for tests like t9700-perl-git.sh which do most of their
774 work in an external test script.
775
776 test_external \
777 'GitwebCache::*FileCache*' \
a0e0ec9f 778 perl "$TEST_DIRECTORY"/t9503/test_cache_interface.pl
2fac6a4b
ÆAB
779
780 If the test is outputting its own TAP you should set the
781 test_external_has_tap variable somewhere before calling the first
782 test_external* function. See t9700-perl-git.sh for an example.
783
784 # The external test will outputs its own plan
785 test_external_has_tap=1
786
787 - test_external_without_stderr [<prereq>] <message> <external> <script>
788
789 Like test_external but fail if there's any output on stderr,
790 instead of checking the exit code.
791
792 test_external_without_stderr \
793 'Perl API' \
a0e0ec9f 794 perl "$TEST_DIRECTORY"/t9700/test.pl
2fac6a4b 795
892e6f7e
ÆAB
796 - test_expect_code <exit-code> <command>
797
798 Run a command and ensure that it exits with the given exit code.
799 For example:
800
801 test_expect_success 'Merge with d/f conflicts' '
802 test_expect_code 1 git merge "merge msg" B master
803 '
804
12e31a6b 805 - test_must_fail [<options>] <git-command>
c9667456
JN
806
807 Run a git command and ensure it fails in a controlled way. Use
971ecbd1
BC
808 this instead of "! <git-command>". When git-command dies due to a
809 segfault, test_must_fail diagnoses it as an error; "! <git-command>"
810 treats it as just another expected failure, which would let such a
811 bug go unnoticed.
c9667456 812
12e31a6b
SG
813 Accepts the following options:
814
815 ok=<signal-name>[,<...>]:
816 Don't treat an exit caused by the given signal as error.
817 Multiple signals can be specified as a comma separated list.
818 Currently recognized signal names are: sigpipe, success.
819 (Don't use 'success', use 'test_might_fail' instead.)
820
821 - test_might_fail [<options>] <git-command>
c9667456
JN
822
823 Similar to test_must_fail, but tolerate success, too. Use this
824 instead of "<git-command> || :" to catch failures due to segv.
825
12e31a6b
SG
826 Accepts the same options as test_must_fail.
827
c9667456
JN
828 - test_cmp <expected> <actual>
829
830 Check whether the content of the <actual> file matches the
831 <expected> file. This behaves like "cmp" but produces more
832 helpful output when the test is run with "-v" option.
833
5a052626
TG
834 - test_cmp_rev <expected> <actual>
835
836 Check whether the <expected> rev points to the same commit as the
837 <actual> rev.
838
fb3340a6
JN
839 - test_line_count (= | -lt | -ge | ...) <length> <file>
840
841 Check whether a file has the length it is expected to.
842
72942a61
ML
843 - test_path_is_file <path> [<diagnosis>]
844 test_path_is_dir <path> [<diagnosis>]
2caf20c5
MM
845 test_path_is_missing <path> [<diagnosis>]
846
72942a61
ML
847 Check if the named path is a file, if the named path is a
848 directory, or if the named path does not exist, respectively,
849 and fail otherwise, showing the <diagnosis> text.
2caf20c5 850
c9667456
JN
851 - test_when_finished <script>
852
853 Prepend <script> to a list of commands to run to clean up
854 at the end of the current test. If some clean-up command
855 fails, the test will not pass.
856
857 Example:
858
859 test_expect_success 'branch pointing to non-commit' '
860 git rev-parse HEAD^{tree} >.git/refs/heads/invalid &&
861 test_when_finished "git update-ref -d refs/heads/invalid" &&
862 ...
863 '
864
900721e1
JS
865 - test_atexit <script>
866
867 Prepend <script> to a list of commands to run unconditionally to
868 clean up before the test script exits, e.g. to stop a daemon:
869
870 test_expect_success 'test git daemon' '
871 git daemon &
872 daemon_pid=$! &&
873 test_atexit 'kill $daemon_pid' &&
874 hello world
875 '
876
877 The commands will be executed before the trash directory is removed,
878 i.e. the atexit commands will still be able to access any pidfiles or
879 socket files.
880
881 Note that these commands will be run even when a test script run
882 with '--immediate' fails. Be careful with your atexit commands to
883 minimize any changes to the failed state.
884
bb98b01e 885 - test_write_lines <lines>
ac9afcc3 886
bb98b01e 887 Write <lines> on standard output, one line per argument.
ac9afcc3
MT
888 Useful to prepare multi-line files in a compact form.
889
890 Example:
891
bb98b01e 892 test_write_lines a b c d e f g >foo
ac9afcc3
MT
893
894 Is a more compact equivalent of:
895 cat >foo <<-EOF
896 a
897 b
898 c
899 d
900 e
901 f
902 g
903 EOF
904
905
c4d2539a
JL
906 - test_pause
907
908 This command is useful for writing and debugging tests and must be
909 removed before submitting. It halts the execution of the test and
910 spawns a shell in the trash directory. Exit the shell to continue
911 the test. Example:
912
913 test_expect_success 'test' '
914 git do-something >actual &&
915 test_pause &&
916 test_cmp expected actual
917 '
918
9ce415d9
JS
919 - test_ln_s_add <path1> <path2>
920
921 This function helps systems whose filesystem does not support symbolic
922 links. Use it to add a symbolic link entry to the index when it is not
923 important that the file system entry is a symbolic link, i.e., instead
924 of the sequence
925
926 ln -s foo bar &&
927 git add bar
928
929 Sometimes it is possible to split a test in a part that does not need
930 the symbolic link in the file system and a part that does; then only
931 the latter part need be protected by a SYMLINKS prerequisite (see below).
932
2c02b110 933 - test_oid_init
934
935 This function loads facts and useful object IDs related to the hash
936 algorithm(s) in use from the files in t/oid-info.
937
938 - test_oid_cache
939
940 This function reads per-hash algorithm information from standard
941 input (usually a heredoc) in the format described in
942 t/oid-info/README. This is useful for test-specific values, such as
943 object IDs, which must vary based on the hash algorithm.
944
945 Certain fixed values, such as hash sizes and common placeholder
946 object IDs, can be loaded with test_oid_init (described above).
947
948 - test_oid <key>
949
950 This function looks up a value for the hash algorithm in use, based
951 on the key given. The value must have been loaded using
952 test_oid_init or test_oid_cache. Providing an unknown key is an
953 error.
954
11f470ae
JH
955 - yes [<string>]
956
957 This is often seen in modern UNIX but some platforms lack it, so
958 the test harness overrides the platform implementation with a
959 more limited one. Use this only when feeding a handful lines of
960 output to the downstream---unlike the real version, it generates
961 only up to 99 lines.
962
963
be53deef
ÆAB
964Prerequisites
965-------------
966
967These are the prerequisites that the test library predefines with
968test_have_prereq.
969
970See the prereq argument to the test_* functions in the "Test harness
971library" section above and the "test_have_prereq" function for how to
972use these, and "test_set_prereq" for how to define your own.
973
f8fc0ee3 974 - PYTHON
be53deef 975
f8fc0ee3
JN
976 Git wasn't compiled with NO_PYTHON=YesPlease. Wrap any tests that
977 need Python with this.
978
979 - PERL
980
981 Git wasn't compiled with NO_PERL=YesPlease.
982
983 Even without the PERL prerequisite, tests can assume there is a
984 usable perl interpreter at $PERL_PATH, though it need not be
985 particularly modern.
be53deef
ÆAB
986
987 - POSIXPERM
988
989 The filesystem supports POSIX style permission bits.
990
991 - BSLASHPSPEC
992
993 Backslashes in pathspec are not directory separators. This is not
994 set on Windows. See 6fd1106a for details.
995
996 - EXECKEEPSPID
997
998 The process retains the same pid across exec(2). See fb9a2bea for
999 details.
1000
20073274
AS
1001 - PIPE
1002
1003 The filesystem we're on supports creation of FIFOs (named pipes)
1004 via mkfifo(1).
1005
be53deef
ÆAB
1006 - SYMLINKS
1007
1008 The filesystem we're on supports symbolic links. E.g. a FAT
1009 filesystem doesn't support these. See 704a3143 for details.
2fac6a4b 1010
c91cfd19
ÆAB
1011 - SANITY
1012
1013 Test is not run by root user, and an attempt to write to an
1014 unwritable file is expected to fail correctly.
2fac6a4b 1015
3eb585c1 1016 - PCRE
8f852ce6 1017
3eb585c1 1018 Git was compiled with support for PCRE. Wrap any tests
8f852ce6
MK
1019 that use git-grep --perl-regexp or git-grep -P in these.
1020
ce9a2570
ÆAB
1021 - LIBPCRE1
1022
1023 Git was compiled with PCRE v1 support via
1024 USE_LIBPCRE1=YesPlease. Wrap any PCRE using tests that for some
1025 reason need v1 of the PCRE library instead of v2 in these.
1026
1027 - LIBPCRE2
1028
1029 Git was compiled with PCRE v2 support via
1030 USE_LIBPCRE2=YesPlease. Wrap any PCRE using tests that for some
1031 reason need v2 of the PCRE library instead of v1 in these.
1032
ac39aa61
MG
1033 - CASE_INSENSITIVE_FS
1034
1035 Test is run on a case insensitive file system.
1036
5b0b5dd4
MG
1037 - UTF8_NFD_TO_NFC
1038
1039 Test is run on a filesystem which converts decomposed utf-8 (nfd)
1040 to precomposed utf-8 (nfc).
1041
68c7d276
ÆAB
1042 - PTHREADS
1043
1044 Git wasn't compiled with NO_PTHREADS=YesPlease.
1045
986aa7f1
JH
1046Tips for Writing Tests
1047----------------------
1048
1049As with any programming projects, existing programs are the best
1050source of the information. However, do _not_ emulate
1051t0000-basic.sh when writing your tests. The test is special in
1052that it tries to validate the very core of GIT. For example, it
1053knows that there will be 256 subdirectories under .git/objects/,
1054and it knows that the object ID of an empty tree is a certain
105540-byte string. This is deliberately done so in t0000-basic.sh
1056because the things the very basic core test tries to achieve is
1057to serve as a basis for people who are changing the GIT internal
1058drastically. For these people, after making certain changes,
1059not seeing failures from the basic test _is_ a failure. And
1060such drastic changes to the core GIT that even changes these
1061otherwise supposedly stable object IDs should be accompanied by
1062an update to t0000-basic.sh.
1063
1064However, other tests that simply rely on basic parts of the core
1065GIT working properly should not have that level of intimate
1066knowledge of the core GIT internals. If all the test scripts
1067hardcoded the object IDs like t0000-basic.sh does, that defeats
1068the purpose of t0000-basic.sh, which is to isolate that level of
1069validation in one place. Your test also ends up needing
1070updating when such a change to the internal happens, so do _not_
1071do it and leave the low level of validation to t0000-basic.sh.
d15e9ebc 1072
0c357544
ÆAB
1073Test coverage
1074-------------
1075
1076You can use the coverage tests to find code paths that are not being
1077used or properly exercised yet.
1078
1079To do that, run the coverage target at the top-level (not in the t/
1080directory):
1081
1082 make coverage
1083
1084That'll compile Git with GCC's coverage arguments, and generate a test
1085report with gcov after the tests finish. Running the coverage tests
1086can take a while, since running the tests in parallel is incompatible
1087with GCC's coverage mode.
1088
1089After the tests have run you can generate a list of untested
1090functions:
1091
1092 make coverage-untested-functions
1093
1094You can also generate a detailed per-file HTML report using the
1095Devel::Cover module. To install it do:
1096
1097 # On Debian or Ubuntu:
1098 sudo aptitude install libdevel-cover-perl
1099
1100 # From the CPAN with cpanminus
1101 curl -L http://cpanmin.us | perl - --sudo --self-upgrade
1102 cpanm --sudo Devel::Cover
1103
1104Then, at the top-level:
1105
1106 make cover_db_html
1107
1108That'll generate a detailed cover report in the "cover_db_html"
1109directory, which you can then copy to a webserver, or inspect locally
1110in a browser.