Merge branch 'sg/stress-test'
[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
0d4dbcd3
TR
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
bb79af9d
JK
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
fb7d1e3a
SG
189--stress::
190--stress=<N>::
191 Run the test script repeatedly in multiple parallel jobs until
192 one of them fails. Useful for reproducing rare failures in
193 flaky tests. The number of parallel jobs is, in order of
194 precedence: <N>, or the value of the GIT_TEST_STRESS_LOAD
195 environment variable, or twice the number of available
196 processors (as shown by the 'getconf' utility), or 8.
197 Implies `--verbose -x --immediate` to get the most information
198 about the failure. Note that the verbose output of each test
199 job is saved to 't/test-results/$TEST_NAME.stress-<nr>.out',
200 and only the output of the failed test job is shown on the
201 terminal. The names of the trash directories get a
202 '.stress-<nr>' suffix, and the trash directory of the failed
203 test job is renamed to end with a '.stress-failed' suffix.
204
e160da7f
MO
205You can also set the GIT_TEST_INSTALLED environment variable to
206the bindir of an existing git installation to test that installation.
207You still need to have built this git sandbox, from which various
208test-* support programs, templates, and perl libraries are used.
209If your installed git is incomplete, it will silently test parts of
210your built version instead.
211
212When using GIT_TEST_INSTALLED, you can also set GIT_TEST_EXEC_PATH to
213override the location of the dashed-form subcommands (what
214GIT_EXEC_PATH would be used for during normal operation).
215GIT_TEST_EXEC_PATH defaults to `$GIT_TEST_INSTALLED/git --exec-path`.
216
217
fbd458a3
JN
218Skipping Tests
219--------------
220
221In some environments, certain tests have no way of succeeding
222due to platform limitation, such as lack of 'unzip' program, or
223filesystem that do not allow arbitrary sequence of non-NUL bytes
224as pathnames.
225
226You should be able to say something like
227
228 $ GIT_SKIP_TESTS=t9200.8 sh ./t9200-git-cvsexport-commit.sh
229
230and even:
231
232 $ GIT_SKIP_TESTS='t[0-4]??? t91?? t9200.8' make
233
234to omit such tests. The value of the environment variable is a
235SP separated list of patterns that tells which tests to skip,
236and either can match the "t[0-9]{4}" part to skip the whole
237test, or t[0-9]{4} followed by ".$number" to say which
238particular test to skip.
239
0445e6f0
IB
240For an individual test suite --run could be used to specify that
241only some tests should be run or that some tests should be
242excluded from a run.
243
244The argument for --run is a list of individual test numbers or
245ranges with an optional negation prefix that define what tests in
246a test suite to include in the run. A range is two numbers
247separated with a dash and matches a range of tests with both ends
248been included. You may omit the first or the second number to
249mean "from the first test" or "up to the very last test"
250respectively.
251
252Optional prefix of '!' means that the test or a range of tests
253should be excluded from the run.
254
255If --run starts with an unprefixed number or range the initial
256set of tests to run is empty. If the first item starts with '!'
257all the tests are added to the initial set. After initial set is
258determined every test number or range is added or excluded from
259the set one by one, from left to right.
260
261Individual numbers or ranges could be separated either by a space
262or a comma.
263
264For example, to run only tests up to a specific test (21), one
265could do this:
266
267 $ sh ./t9200-git-cvsexport-commit.sh --run='1-21'
268
269or this:
270
271 $ sh ./t9200-git-cvsexport-commit.sh --run='-21'
272
273Common case is to run several setup tests (1, 2, 3) and then a
274specific test (21) that relies on that setup:
275
276 $ sh ./t9200-git-cvsexport-commit.sh --run='1 2 3 21'
277
278or:
279
280 $ sh ./t9200-git-cvsexport-commit.sh --run=1,2,3,21
281
282or:
283
284 $ sh ./t9200-git-cvsexport-commit.sh --run='-3 21'
285
01e4be6c
KS
286As noted above, the test set is built by going through the items
287from left to right, so this:
0445e6f0
IB
288
289 $ sh ./t9200-git-cvsexport-commit.sh --run='1-4 !3'
290
01e4be6c 291will run tests 1, 2, and 4. Items that come later have higher
2e3a16b2 292precedence. It means that this:
0445e6f0
IB
293
294 $ sh ./t9200-git-cvsexport-commit.sh --run='!3 1-4'
295
296would just run tests from 1 to 4, including 3.
297
298You may use negation with ranges. The following will run all
299test in the test suite except from 7 up to 11:
300
301 $ sh ./t9200-git-cvsexport-commit.sh --run='!7-11'
302
303Some tests in a test suite rely on the previous tests performing
304certain actions, specifically some tests are designated as
305"setup" test, so you cannot _arbitrarily_ disable one test and
306expect the rest to function correctly.
307
308--run is mostly useful when you want to focus on a specific test
309and know what setup is needed for it. Or when you want to run
310everything up to a certain test.
fbd458a3
JN
311
312
4c2db938
NTND
313Running tests with special setups
314---------------------------------
315
316The whole test suite could be run to test some special features
317that cannot be easily covered by a few specific test cases. These
318could be enabled by running the test suite with correct GIT_TEST_
319environment set.
320
6cdccfce
ÆAB
321GIT_TEST_GETTEXT_POISON=<non-empty?> turns all strings marked for
322translation into gibberish if non-empty (think "test -n"). Used for
323spotting those tests that need to be marked with a C_LOCALE_OUTPUT
324prerequisite when adding more strings for translation. See "Testing
325marked strings" in po/README for details.
326
4c2db938
NTND
327GIT_TEST_SPLIT_INDEX=<boolean> forces split-index mode on the whole
328test suite. Accept any boolean values that are accepted by git-config.
329
43fa44fa
NTND
330GIT_TEST_FULL_IN_PACK_ARRAY=<boolean> exercises the uncommon
331pack-objects code path where there are more than 1024 packs even if
332the actual number of packs in repository is below this limit. Accept
333any boolean values that are accepted by git-config.
334
ac77d0c3
NTND
335GIT_TEST_OE_SIZE=<n> exercises the uncommon pack-objects code path
336where we do not cache object size in memory and read it from existing
337packs on demand. This normally only happens when the object size is
338over 2GB. This variable forces the code path on any object larger than
339<n> bytes.
340
ac6e12f9 341GIT_TEST_OE_DELTA_SIZE=<n> exercises the uncommon pack-objects code
9ac3f0e5
NTND
342path where deltas larger than this limit require extra memory
343allocation for bookkeeping.
344
5f4436a7
NTND
345GIT_TEST_VALIDATE_INDEX_CACHE_ENTRIES=<boolean> checks that cache-tree
346records are valid when the index is written out or after a merge. This
347is mostly to catch missing invalidation. Default is true.
348
859fdc0c
DS
349GIT_TEST_COMMIT_GRAPH=<boolean>, when true, forces the commit-graph to
350be written after every 'git commit' command, and overrides the
351'core.commitGraph' setting to true.
352
4cb54d0a
BP
353GIT_TEST_FSMONITOR=$PWD/t7519/fsmonitor-all exercises the fsmonitor
354code path for utilizing a file system monitor to speed up detecting
355new or changed files.
356
1f357b04
BP
357GIT_TEST_INDEX_VERSION=<n> exercises the index read/write code path
358for the index version specified. Can be set to any valid version
359(currently 2, 3, or 4).
360
5765d97b
BP
361GIT_TEST_PRELOAD_INDEX=<boolean> exercises the preload-index code path
362by overriding the minimum number of cache entries required per thread.
363
62c23938
ÆAB
364GIT_TEST_REBASE_USE_BUILTIN=<boolean>, when false, disables the
365builtin version of git-rebase. See 'rebase.useBuiltin' in
366git-config(1).
367
c780b9cf
BP
368GIT_TEST_INDEX_THREADS=<n> enables exercising the multi-threaded loading
369of the index for the whole test suite by bypassing the default number of
370cache entries and thread minimums. Setting this to 1 will make the
371index loading single threaded.
372
0465a505
DS
373GIT_TEST_MULTI_PACK_INDEX=<boolean>, when true, forces the multi-pack-
374index to be written after every 'git repack' command, and overrides the
375'core.multiPackIndex' setting to true.
376
f50c9f76
PB
377Naming Tests
378------------
379
380The test files are named as:
381
382 tNNNN-commandname-details.sh
383
384where N is a decimal digit.
385
386First digit tells the family:
387
388 0 - the absolute basics and global stuff
389 1 - the basic commands concerning database
390 2 - the basic commands concerning the working tree
391 3 - the other basic commands (e.g. ls-files)
392 4 - the diff commands
393 5 - the pull and exporting commands
394 6 - the revision tree commands (even e.g. merge-base)
8f4a9b62 395 7 - the porcelainish commands concerning the working tree
8757749e
JN
396 8 - the porcelainish commands concerning forensics
397 9 - the git tools
f50c9f76
PB
398
399Second digit tells the particular command we are testing.
400
401Third digit (optionally) tells the particular switch or group of switches
402we are testing.
403
77656600
JH
404If you create files under t/ directory (i.e. here) that is not
405the top-level test script, never name the file to match the above
406pattern. The Makefile here considers all such files as the
63d32945 407top-level test script and tries to run all of them. Care is
77656600
JH
408especially needed if you are creating a common test library
409file, similar to test-lib.sh, because such a library file may
410not be suitable for standalone execution.
411
f50c9f76 412
986aa7f1
JH
413Writing Tests
414-------------
415
416The test script is written as a shell script. It should start
51b7a525 417with the standard "#!/bin/sh", and an
986aa7f1
JH
418assignment to variable 'test_description', like this:
419
420 #!/bin/sh
986aa7f1 421
14cd1ff3 422 test_description='xxx test (option --frotz)
986aa7f1
JH
423
424 This test registers the following structure in the cache
425 and tries to run git-ls-files with option --frotz.'
426
f50c9f76 427
986aa7f1
JH
428Source 'test-lib.sh'
429--------------------
430
431After assigning test_description, the test script should source
432test-lib.sh like this:
433
434 . ./test-lib.sh
435
436This test harness library does the following things:
437
438 - If the script is invoked with command line argument --help
439 (or -h), it shows the test_description and exits.
440
e1ca1c9d
ÆAB
441 - Creates an empty test directory with an empty .git/objects database
442 and chdir(2) into it. This directory is 't/trash
443 directory.$test_name_without_dotsh', with t/ subject to change by
fb7d1e3a
SG
444 the --root option documented above, and a '.stress-<N>' suffix
445 appended by the --stress option.
986aa7f1
JH
446
447 - Defines standard test helper functions for your scripts to
448 use. These functions are designed to make all scripts behave
449 consistently when command line arguments --verbose (or -v),
450 --debug (or -d), and --immediate (or -i) is given.
451
441ee35d
MD
452Do's & don'ts
453-------------
20873f45 454
6fd45295 455Here are a few examples of things you probably should and shouldn't do
20873f45
ÆAB
456when writing tests.
457
441ee35d 458Here are the "do's:"
20873f45 459
6fd45295 460 - Put all code inside test_expect_success and other assertions.
20873f45
ÆAB
461
462 Even code that isn't a test per se, but merely some setup code
6fd45295 463 should be inside a test assertion.
20873f45
ÆAB
464
465 - Chain your test assertions
466
467 Write test code like this:
468
469 git merge foo &&
470 git push bar &&
471 test ...
472
473 Instead of:
474
475 git merge hla
476 git push gh
477 test ...
478
479 That way all of the commands in your tests will succeed or fail. If
00648ba0
EN
480 you must ignore the return value of something, consider using a
481 helper function (e.g. use sane_unset instead of unset, in order
482 to avoid unportable return value for unsetting a variable that was
483 already unset), or prepending the command with test_might_fail or
484 test_must_fail.
20873f45 485
0c357544
ÆAB
486 - Check the test coverage for your tests. See the "Test coverage"
487 below.
488
63d32945
MW
489 Don't blindly follow test coverage metrics; if a new function you added
490 doesn't have any coverage, then you're probably doing something wrong,
e8b55f5c
ÆAB
491 but having 100% coverage doesn't necessarily mean that you tested
492 everything.
493
494 Tests that are likely to smoke out future regressions are better
495 than tests that just inflate the coverage metrics.
496
95b104c8
JS
497 - When a test checks for an absolute path that a git command generated,
498 construct the expected value using $(pwd) rather than $PWD,
499 $TEST_DIRECTORY, or $TRASH_DIRECTORY. It makes a difference on
500 Windows, where the shell (MSYS bash) mangles absolute path names.
501 For details, see the commit message of 4114156ae9.
502
441ee35d
MD
503 - Remember that inside the <script> part, the standard output and
504 standard error streams are discarded, and the test harness only
505 reports "ok" or "not ok" to the end user running the tests. Under
506 --verbose, they are shown to help debug the tests.
507
508And here are the "don'ts:"
20873f45 509
441ee35d 510 - Don't exit() within a <script> part.
20873f45
ÆAB
511
512 The harness will catch this as a programming error of the test.
513 Use test_done instead if you need to stop the tests early (see
514 "Skipping tests" below).
515
441ee35d
MD
516 - Don't use '! git cmd' when you want to make sure the git command
517 exits with failure in a controlled way by calling "die()". Instead,
ad78585e
JH
518 use 'test_must_fail git cmd'. This will signal a failure if git
519 dies in an unexpected way (e.g. segfault).
520
f445500e 521 On the other hand, don't use test_must_fail for running regular
53de7424
JH
522 platform commands; just use '! cmd'. We are not in the business
523 of verifying that the world given to us sanely works.
f445500e 524
a378fee5
MD
525 - Don't feed the output of a git command to a pipe, as in:
526
527 git -C repo ls-files |
528 xargs -n 1 basename |
529 grep foo
530
531 which will discard git's exit code and may mask a crash. In the
532 above example, all exit codes are ignored except grep's.
533
534 Instead, write the output of that command to a temporary
535 file with ">" or assign it to a variable with "x=$(git ...)" rather
536 than pipe it.
537
538 - Don't use command substitution in a way that discards git's exit
539 code. When assigning to a variable, the exit code is not discarded,
540 e.g.:
541
542 x=$(git cat-file -p $sha) &&
543 ...
544
545 is OK because a crash in "git cat-file" will cause the "&&" chain
546 to fail, but:
547
548 test "refs/heads/foo" = "$(git symbolic-ref HEAD)"
549
550 is not OK and a crash in git could go undetected.
551
441ee35d
MD
552 - Don't use perl without spelling it as "$PERL_PATH". This is to help
553 our friends on Windows where the platform Perl often adds CR before
ad78585e 554 the end of line, and they bundle Git with a version of Perl that
a0e0ec9f
JK
555 does not do so, whose path is specified with $PERL_PATH. Note that we
556 provide a "perl" function which uses $PERL_PATH under the hood, so
557 you do not need to worry when simply running perl in the test scripts
558 (but you do, for example, on a shebang line or in a sub script
559 created via "write_script").
ad78585e 560
441ee35d
MD
561 - Don't use sh without spelling it as "$SHELL_PATH", when the script
562 can be misinterpreted by broken platform shell (e.g. Solaris).
ad78585e 563
441ee35d 564 - Don't chdir around in tests. It is not sufficient to chdir to
ad78585e
JH
565 somewhere and then chdir back to the original location later in
566 the test, as any intermediate step can fail and abort the test,
567 causing the next test to start in an unexpected directory. Do so
568 inside a subshell if necessary.
569
441ee35d
MD
570 - Don't save and verify the standard error of compound commands, i.e.
571 group commands, subshells, and shell functions (except test helper
94201a2b
SG
572 functions like 'test_must_fail') like this:
573
574 ( cd dir && git cmd ) 2>error &&
575 test_cmp expect error
576
577 When running the test with '-x' tracing, then the trace of commands
578 executed in the compound command will be included in standard error
579 as well, quite possibly throwing off the subsequent checks examining
580 the output. Instead, save only the relevant git command's standard
581 error:
582
583 ( cd dir && git cmd 2>../error ) &&
584 test_cmp expect error
585
441ee35d 586 - Don't break the TAP output
20873f45 587
6fd45295
JH
588 The raw output from your test may be interpreted by a TAP harness. TAP
589 harnesses will ignore everything they don't know about, but don't step
590 on their toes in these areas:
20873f45
ÆAB
591
592 - Don't print lines like "$x..$y" where $x and $y are integers.
593
594 - Don't print lines that begin with "ok" or "not ok".
595
6fd45295 596 TAP harnesses expect a line that begins with either "ok" and "not
20873f45
ÆAB
597 ok" to signal a test passed or failed (and our harness already
598 produces such lines), so your script shouldn't emit such lines to
599 their output.
600
601 You can glean some further possible issues from the TAP grammar
c1d44cee 602 (see https://metacpan.org/pod/TAP::Parser::Grammar#TAP-GRAMMAR)
20873f45
ÆAB
603 but the best indication is to just run the tests with prove(1),
604 it'll complain if anything is amiss.
605
20873f45 606
b5500d16
ÆAB
607Skipping tests
608--------------
609
681186ae 610If you need to skip tests you should do so by using the three-arg form
99d9050d
ÆAB
611of the test_* functions (see the "Test harness library" section
612below), e.g.:
613
ad78585e 614 test_expect_success PERL 'I need Perl' '
a0e0ec9f 615 perl -e "hlagh() if unf_unf()"
ad78585e 616 '
99d9050d
ÆAB
617
618The advantage of skipping tests like this is that platforms that don't
619have the PERL and other optional dependencies get an indication of how
620many tests they're missing.
621
622If the test code is too hairy for that (i.e. does a lot of setup work
623outside test assertions) you can also skip all remaining tests by
624setting skip_all and immediately call test_done:
b5500d16
ÆAB
625
626 if ! test_have_prereq PERL
627 then
628 skip_all='skipping perl interface tests, perl not available'
629 test_done
630 fi
14cd1ff3 631
99d9050d
ÆAB
632The string you give to skip_all will be used as an explanation for why
633the test was skipped.
634
986aa7f1
JH
635End with test_done
636------------------
637
638Your script will be a sequence of tests, using helper functions
639from the test harness library. At the end of the script, call
640'test_done'.
641
642
643Test harness library
644--------------------
645
646There are a handful helper functions defined in the test harness
647library for your script to use.
648
9a897893 649 - test_expect_success [<prereq>] <message> <script>
986aa7f1 650
72942a61 651 Usually takes two strings as parameters, and evaluates the
986aa7f1
JH
652 <script>. If it yields success, test is considered
653 successful. <message> should state what it is testing.
654
655 Example:
656
657 test_expect_success \
658 'git-write-tree should be able to write an empty tree.' \
659 'tree=$(git-write-tree)'
660
9a897893 661 If you supply three parameters the first will be taken to be a
72942a61 662 prerequisite; see the test_set_prereq and test_have_prereq
9a897893
ÆAB
663 documentation below:
664
665 test_expect_success TTY 'git --paginate rev-list uses a pager' \
666 ' ... '
667
93a57246
ÆAB
668 You can also supply a comma-separated list of prerequisites, in the
669 rare case where your test depends on more than one:
670
671 test_expect_success PERL,PYTHON 'yo dawg' \
672 ' test $(perl -E 'print eval "1 +" . qx[python -c "print 2"]') == "4" '
673
9a897893 674 - test_expect_failure [<prereq>] <message> <script>
986aa7f1 675
41ac414e
JH
676 This is NOT the opposite of test_expect_success, but is used
677 to mark a test that demonstrates a known breakage. Unlike
678 the usual test_expect_success tests, which say "ok" on
679 success and "FAIL" on failure, this will say "FIXED" on
680 success and "still broken" on failure. Failures from these
681 tests won't cause -i (immediate) to stop.
986aa7f1 682
9a897893
ÆAB
683 Like test_expect_success this function can optionally use a three
684 argument invocation with a prerequisite as the first argument.
685
986aa7f1
JH
686 - test_debug <script>
687
688 This takes a single argument, <script>, and evaluates it only
689 when the test script is started with --debug command line
690 argument. This is primarily meant for use during the
691 development of a new test script.
692
6a94088c
JS
693 - debug <git-command>
694
695 Run a git command inside a debugger. This is primarily meant for
696 use when debugging a failing test script.
697
986aa7f1
JH
698 - test_done
699
700 Your test script must have test_done at the end. Its purpose
701 is to summarize successes and failures in the test script and
702 exit with an appropriate error code.
703
00884968
JS
704 - test_tick
705
706 Make commit and tag names consistent by setting the author and
63d32945 707 committer times to defined state. Subsequent calls will
00884968
JS
708 advance the times by a fixed amount.
709
710 - test_commit <message> [<filename> [<contents>]]
711
712 Creates a commit with the given message, committing the given
713 file with the given contents (default for both is to reuse the
714 message string), and adds a tag (again reusing the message
715 string as name). Calls test_tick to make the SHA-1s
716 reproducible.
717
718 - test_merge <message> <commit-or-tag>
719
720 Merges the given rev using the given message. Like test_commit,
721 creates a tag and calls test_tick before committing.
986aa7f1 722
72942a61 723 - test_set_prereq <prereq>
9a897893
ÆAB
724
725 Set a test prerequisite to be used later with test_have_prereq. The
be53deef
ÆAB
726 test-lib will set some prerequisites for you, see the
727 "Prerequisites" section below for a full list of these.
728
729 Others you can set yourself and use later with either
730 test_have_prereq directly, or the three argument invocation of
731 test_expect_success and test_expect_failure.
9a897893 732
72942a61 733 - test_have_prereq <prereq>
9a897893 734
4473060b
ÆAB
735 Check if we have a prerequisite previously set with test_set_prereq.
736 The most common way to use this explicitly (as opposed to the
737 implicit use when an argument is passed to test_expect_*) is to skip
738 all the tests at the start of the test script if we don't have some
739 essential prerequisite:
9a897893
ÆAB
740
741 if ! test_have_prereq PERL
742 then
743 skip_all='skipping perl interface tests, perl not available'
744 test_done
745 fi
746
2fac6a4b
ÆAB
747 - test_external [<prereq>] <message> <external> <script>
748
749 Execute a <script> with an <external> interpreter (like perl). This
750 was added for tests like t9700-perl-git.sh which do most of their
751 work in an external test script.
752
753 test_external \
754 'GitwebCache::*FileCache*' \
a0e0ec9f 755 perl "$TEST_DIRECTORY"/t9503/test_cache_interface.pl
2fac6a4b
ÆAB
756
757 If the test is outputting its own TAP you should set the
758 test_external_has_tap variable somewhere before calling the first
759 test_external* function. See t9700-perl-git.sh for an example.
760
761 # The external test will outputs its own plan
762 test_external_has_tap=1
763
764 - test_external_without_stderr [<prereq>] <message> <external> <script>
765
766 Like test_external but fail if there's any output on stderr,
767 instead of checking the exit code.
768
769 test_external_without_stderr \
770 'Perl API' \
a0e0ec9f 771 perl "$TEST_DIRECTORY"/t9700/test.pl
2fac6a4b 772
892e6f7e
ÆAB
773 - test_expect_code <exit-code> <command>
774
775 Run a command and ensure that it exits with the given exit code.
776 For example:
777
778 test_expect_success 'Merge with d/f conflicts' '
779 test_expect_code 1 git merge "merge msg" B master
780 '
781
12e31a6b 782 - test_must_fail [<options>] <git-command>
c9667456
JN
783
784 Run a git command and ensure it fails in a controlled way. Use
971ecbd1
BC
785 this instead of "! <git-command>". When git-command dies due to a
786 segfault, test_must_fail diagnoses it as an error; "! <git-command>"
787 treats it as just another expected failure, which would let such a
788 bug go unnoticed.
c9667456 789
12e31a6b
SG
790 Accepts the following options:
791
792 ok=<signal-name>[,<...>]:
793 Don't treat an exit caused by the given signal as error.
794 Multiple signals can be specified as a comma separated list.
795 Currently recognized signal names are: sigpipe, success.
796 (Don't use 'success', use 'test_might_fail' instead.)
797
798 - test_might_fail [<options>] <git-command>
c9667456
JN
799
800 Similar to test_must_fail, but tolerate success, too. Use this
801 instead of "<git-command> || :" to catch failures due to segv.
802
12e31a6b
SG
803 Accepts the same options as test_must_fail.
804
c9667456
JN
805 - test_cmp <expected> <actual>
806
807 Check whether the content of the <actual> file matches the
808 <expected> file. This behaves like "cmp" but produces more
809 helpful output when the test is run with "-v" option.
810
5a052626
TG
811 - test_cmp_rev <expected> <actual>
812
813 Check whether the <expected> rev points to the same commit as the
814 <actual> rev.
815
fb3340a6
JN
816 - test_line_count (= | -lt | -ge | ...) <length> <file>
817
818 Check whether a file has the length it is expected to.
819
72942a61
ML
820 - test_path_is_file <path> [<diagnosis>]
821 test_path_is_dir <path> [<diagnosis>]
2caf20c5
MM
822 test_path_is_missing <path> [<diagnosis>]
823
72942a61
ML
824 Check if the named path is a file, if the named path is a
825 directory, or if the named path does not exist, respectively,
826 and fail otherwise, showing the <diagnosis> text.
2caf20c5 827
c9667456
JN
828 - test_when_finished <script>
829
830 Prepend <script> to a list of commands to run to clean up
831 at the end of the current test. If some clean-up command
832 fails, the test will not pass.
833
834 Example:
835
836 test_expect_success 'branch pointing to non-commit' '
837 git rev-parse HEAD^{tree} >.git/refs/heads/invalid &&
838 test_when_finished "git update-ref -d refs/heads/invalid" &&
839 ...
840 '
841
bb98b01e 842 - test_write_lines <lines>
ac9afcc3 843
bb98b01e 844 Write <lines> on standard output, one line per argument.
ac9afcc3
MT
845 Useful to prepare multi-line files in a compact form.
846
847 Example:
848
bb98b01e 849 test_write_lines a b c d e f g >foo
ac9afcc3
MT
850
851 Is a more compact equivalent of:
852 cat >foo <<-EOF
853 a
854 b
855 c
856 d
857 e
858 f
859 g
860 EOF
861
862
c4d2539a
JL
863 - test_pause
864
865 This command is useful for writing and debugging tests and must be
866 removed before submitting. It halts the execution of the test and
867 spawns a shell in the trash directory. Exit the shell to continue
868 the test. Example:
869
870 test_expect_success 'test' '
871 git do-something >actual &&
872 test_pause &&
873 test_cmp expected actual
874 '
875
9ce415d9
JS
876 - test_ln_s_add <path1> <path2>
877
878 This function helps systems whose filesystem does not support symbolic
879 links. Use it to add a symbolic link entry to the index when it is not
880 important that the file system entry is a symbolic link, i.e., instead
881 of the sequence
882
883 ln -s foo bar &&
884 git add bar
885
886 Sometimes it is possible to split a test in a part that does not need
887 the symbolic link in the file system and a part that does; then only
888 the latter part need be protected by a SYMLINKS prerequisite (see below).
889
2c02b110 890 - test_oid_init
891
892 This function loads facts and useful object IDs related to the hash
893 algorithm(s) in use from the files in t/oid-info.
894
895 - test_oid_cache
896
897 This function reads per-hash algorithm information from standard
898 input (usually a heredoc) in the format described in
899 t/oid-info/README. This is useful for test-specific values, such as
900 object IDs, which must vary based on the hash algorithm.
901
902 Certain fixed values, such as hash sizes and common placeholder
903 object IDs, can be loaded with test_oid_init (described above).
904
905 - test_oid <key>
906
907 This function looks up a value for the hash algorithm in use, based
908 on the key given. The value must have been loaded using
909 test_oid_init or test_oid_cache. Providing an unknown key is an
910 error.
911
be53deef
ÆAB
912Prerequisites
913-------------
914
915These are the prerequisites that the test library predefines with
916test_have_prereq.
917
918See the prereq argument to the test_* functions in the "Test harness
919library" section above and the "test_have_prereq" function for how to
920use these, and "test_set_prereq" for how to define your own.
921
f8fc0ee3 922 - PYTHON
be53deef 923
f8fc0ee3
JN
924 Git wasn't compiled with NO_PYTHON=YesPlease. Wrap any tests that
925 need Python with this.
926
927 - PERL
928
929 Git wasn't compiled with NO_PERL=YesPlease.
930
931 Even without the PERL prerequisite, tests can assume there is a
932 usable perl interpreter at $PERL_PATH, though it need not be
933 particularly modern.
be53deef
ÆAB
934
935 - POSIXPERM
936
937 The filesystem supports POSIX style permission bits.
938
939 - BSLASHPSPEC
940
941 Backslashes in pathspec are not directory separators. This is not
942 set on Windows. See 6fd1106a for details.
943
944 - EXECKEEPSPID
945
946 The process retains the same pid across exec(2). See fb9a2bea for
947 details.
948
20073274
AS
949 - PIPE
950
951 The filesystem we're on supports creation of FIFOs (named pipes)
952 via mkfifo(1).
953
be53deef
ÆAB
954 - SYMLINKS
955
956 The filesystem we're on supports symbolic links. E.g. a FAT
957 filesystem doesn't support these. See 704a3143 for details.
2fac6a4b 958
c91cfd19
ÆAB
959 - SANITY
960
961 Test is not run by root user, and an attempt to write to an
962 unwritable file is expected to fail correctly.
2fac6a4b 963
3eb585c1 964 - PCRE
8f852ce6 965
3eb585c1 966 Git was compiled with support for PCRE. Wrap any tests
8f852ce6
MK
967 that use git-grep --perl-regexp or git-grep -P in these.
968
ce9a2570
ÆAB
969 - LIBPCRE1
970
971 Git was compiled with PCRE v1 support via
972 USE_LIBPCRE1=YesPlease. Wrap any PCRE using tests that for some
973 reason need v1 of the PCRE library instead of v2 in these.
974
975 - LIBPCRE2
976
977 Git was compiled with PCRE v2 support via
978 USE_LIBPCRE2=YesPlease. Wrap any PCRE using tests that for some
979 reason need v2 of the PCRE library instead of v1 in these.
980
ac39aa61
MG
981 - CASE_INSENSITIVE_FS
982
983 Test is run on a case insensitive file system.
984
5b0b5dd4
MG
985 - UTF8_NFD_TO_NFC
986
987 Test is run on a filesystem which converts decomposed utf-8 (nfd)
988 to precomposed utf-8 (nfc).
989
68c7d276
ÆAB
990 - PTHREADS
991
992 Git wasn't compiled with NO_PTHREADS=YesPlease.
993
986aa7f1
JH
994Tips for Writing Tests
995----------------------
996
997As with any programming projects, existing programs are the best
998source of the information. However, do _not_ emulate
999t0000-basic.sh when writing your tests. The test is special in
1000that it tries to validate the very core of GIT. For example, it
1001knows that there will be 256 subdirectories under .git/objects/,
1002and it knows that the object ID of an empty tree is a certain
100340-byte string. This is deliberately done so in t0000-basic.sh
1004because the things the very basic core test tries to achieve is
1005to serve as a basis for people who are changing the GIT internal
1006drastically. For these people, after making certain changes,
1007not seeing failures from the basic test _is_ a failure. And
1008such drastic changes to the core GIT that even changes these
1009otherwise supposedly stable object IDs should be accompanied by
1010an update to t0000-basic.sh.
1011
1012However, other tests that simply rely on basic parts of the core
1013GIT working properly should not have that level of intimate
1014knowledge of the core GIT internals. If all the test scripts
1015hardcoded the object IDs like t0000-basic.sh does, that defeats
1016the purpose of t0000-basic.sh, which is to isolate that level of
1017validation in one place. Your test also ends up needing
1018updating when such a change to the internal happens, so do _not_
1019do it and leave the low level of validation to t0000-basic.sh.
d15e9ebc 1020
0c357544
ÆAB
1021Test coverage
1022-------------
1023
1024You can use the coverage tests to find code paths that are not being
1025used or properly exercised yet.
1026
1027To do that, run the coverage target at the top-level (not in the t/
1028directory):
1029
1030 make coverage
1031
1032That'll compile Git with GCC's coverage arguments, and generate a test
1033report with gcov after the tests finish. Running the coverage tests
1034can take a while, since running the tests in parallel is incompatible
1035with GCC's coverage mode.
1036
1037After the tests have run you can generate a list of untested
1038functions:
1039
1040 make coverage-untested-functions
1041
1042You can also generate a detailed per-file HTML report using the
1043Devel::Cover module. To install it do:
1044
1045 # On Debian or Ubuntu:
1046 sudo aptitude install libdevel-cover-perl
1047
1048 # From the CPAN with cpanminus
1049 curl -L http://cpanmin.us | perl - --sudo --self-upgrade
1050 cpanm --sudo Devel::Cover
1051
1052Then, at the top-level:
1053
1054 make cover_db_html
1055
1056That'll generate a detailed cover report in the "cover_db_html"
1057directory, which you can then copy to a webserver, or inspect locally
1058in a browser.