tests: define GIT_TEST_PROTOCOL_VERSION
[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 --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
205 You can also set the GIT_TEST_INSTALLED environment variable to
206 the bindir of an existing git installation to test that installation.
207 You still need to have built this git sandbox, from which various
208 test-* support programs, templates, and perl libraries are used.
209 If your installed git is incomplete, it will silently test parts of
210 your built version instead.
211
212 When using GIT_TEST_INSTALLED, you can also set GIT_TEST_EXEC_PATH to
213 override the location of the dashed-form subcommands (what
214 GIT_EXEC_PATH would be used for during normal operation).
215 GIT_TEST_EXEC_PATH defaults to `$GIT_TEST_INSTALLED/git --exec-path`.
216
217
218 Skipping Tests
219 --------------
220
221 In some environments, certain tests have no way of succeeding
222 due to platform limitation, such as lack of 'unzip' program, or
223 filesystem that do not allow arbitrary sequence of non-NUL bytes
224 as pathnames.
225
226 You should be able to say something like
227
228 $ GIT_SKIP_TESTS=t9200.8 sh ./t9200-git-cvsexport-commit.sh
229
230 and even:
231
232 $ GIT_SKIP_TESTS='t[0-4]??? t91?? t9200.8' make
233
234 to omit such tests. The value of the environment variable is a
235 SP separated list of patterns that tells which tests to skip,
236 and either can match the "t[0-9]{4}" part to skip the whole
237 test, or t[0-9]{4} followed by ".$number" to say which
238 particular test to skip.
239
240 For an individual test suite --run could be used to specify that
241 only some tests should be run or that some tests should be
242 excluded from a run.
243
244 The argument for --run is a list of individual test numbers or
245 ranges with an optional negation prefix that define what tests in
246 a test suite to include in the run. A range is two numbers
247 separated with a dash and matches a range of tests with both ends
248 been included. You may omit the first or the second number to
249 mean "from the first test" or "up to the very last test"
250 respectively.
251
252 Optional prefix of '!' means that the test or a range of tests
253 should be excluded from the run.
254
255 If --run starts with an unprefixed number or range the initial
256 set of tests to run is empty. If the first item starts with '!'
257 all the tests are added to the initial set. After initial set is
258 determined every test number or range is added or excluded from
259 the set one by one, from left to right.
260
261 Individual numbers or ranges could be separated either by a space
262 or a comma.
263
264 For example, to run only tests up to a specific test (21), one
265 could do this:
266
267 $ sh ./t9200-git-cvsexport-commit.sh --run='1-21'
268
269 or this:
270
271 $ sh ./t9200-git-cvsexport-commit.sh --run='-21'
272
273 Common case is to run several setup tests (1, 2, 3) and then a
274 specific test (21) that relies on that setup:
275
276 $ sh ./t9200-git-cvsexport-commit.sh --run='1 2 3 21'
277
278 or:
279
280 $ sh ./t9200-git-cvsexport-commit.sh --run=1,2,3,21
281
282 or:
283
284 $ sh ./t9200-git-cvsexport-commit.sh --run='-3 21'
285
286 As noted above, the test set is built by going through the items
287 from left to right, so this:
288
289 $ sh ./t9200-git-cvsexport-commit.sh --run='1-4 !3'
290
291 will run tests 1, 2, and 4. Items that come later have higher
292 precedence. It means that this:
293
294 $ sh ./t9200-git-cvsexport-commit.sh --run='!3 1-4'
295
296 would just run tests from 1 to 4, including 3.
297
298 You may use negation with ranges. The following will run all
299 test in the test suite except from 7 up to 11:
300
301 $ sh ./t9200-git-cvsexport-commit.sh --run='!7-11'
302
303 Some tests in a test suite rely on the previous tests performing
304 certain actions, specifically some tests are designated as
305 "setup" test, so you cannot _arbitrarily_ disable one test and
306 expect the rest to function correctly.
307
308 --run is mostly useful when you want to focus on a specific test
309 and know what setup is needed for it. Or when you want to run
310 everything up to a certain test.
311
312
313 Running tests with special setups
314 ---------------------------------
315
316 The whole test suite could be run to test some special features
317 that cannot be easily covered by a few specific test cases. These
318 could be enabled by running the test suite with correct GIT_TEST_
319 environment set.
320
321 GIT_TEST_GETTEXT_POISON=<non-empty?> turns all strings marked for
322 translation into gibberish if non-empty (think "test -n"). Used for
323 spotting those tests that need to be marked with a C_LOCALE_OUTPUT
324 prerequisite when adding more strings for translation. See "Testing
325 marked strings" in po/README for details.
326
327 GIT_TEST_SPLIT_INDEX=<boolean> forces split-index mode on the whole
328 test suite. Accept any boolean values that are accepted by git-config.
329
330 GIT_TEST_PROTOCOL_VERSION=<n>, when set, overrides the
331 'protocol.version' setting to n if it is less than n.
332
333 GIT_TEST_FULL_IN_PACK_ARRAY=<boolean> exercises the uncommon
334 pack-objects code path where there are more than 1024 packs even if
335 the actual number of packs in repository is below this limit. Accept
336 any boolean values that are accepted by git-config.
337
338 GIT_TEST_OE_SIZE=<n> exercises the uncommon pack-objects code path
339 where we do not cache object size in memory and read it from existing
340 packs on demand. This normally only happens when the object size is
341 over 2GB. This variable forces the code path on any object larger than
342 <n> bytes.
343
344 GIT_TEST_OE_DELTA_SIZE=<n> exercises the uncommon pack-objects code
345 path where deltas larger than this limit require extra memory
346 allocation for bookkeeping.
347
348 GIT_TEST_VALIDATE_INDEX_CACHE_ENTRIES=<boolean> checks that cache-tree
349 records are valid when the index is written out or after a merge. This
350 is mostly to catch missing invalidation. Default is true.
351
352 GIT_TEST_COMMIT_GRAPH=<boolean>, when true, forces the commit-graph to
353 be written after every 'git commit' command, and overrides the
354 'core.commitGraph' setting to true.
355
356 GIT_TEST_FSMONITOR=$PWD/t7519/fsmonitor-all exercises the fsmonitor
357 code path for utilizing a file system monitor to speed up detecting
358 new or changed files.
359
360 GIT_TEST_INDEX_VERSION=<n> exercises the index read/write code path
361 for the index version specified. Can be set to any valid version
362 (currently 2, 3, or 4).
363
364 GIT_TEST_PRELOAD_INDEX=<boolean> exercises the preload-index code path
365 by overriding the minimum number of cache entries required per thread.
366
367 GIT_TEST_REBASE_USE_BUILTIN=<boolean>, when false, disables the
368 builtin version of git-rebase. See 'rebase.useBuiltin' in
369 git-config(1).
370
371 GIT_TEST_INDEX_THREADS=<n> enables exercising the multi-threaded loading
372 of the index for the whole test suite by bypassing the default number of
373 cache entries and thread minimums. Setting this to 1 will make the
374 index loading single threaded.
375
376 GIT_TEST_MULTI_PACK_INDEX=<boolean>, when true, forces the multi-pack-
377 index to be written after every 'git repack' command, and overrides the
378 'core.multiPackIndex' setting to true.
379
380 GIT_TEST_SIDEBAND_ALL=<boolean>, when true, overrides the
381 'uploadpack.allowSidebandAll' setting to true, and when false, forces
382 fetch-pack to not request sideband-all (even if the server advertises
383 sideband-all).
384
385 Naming Tests
386 ------------
387
388 The test files are named as:
389
390 tNNNN-commandname-details.sh
391
392 where N is a decimal digit.
393
394 First digit tells the family:
395
396 0 - the absolute basics and global stuff
397 1 - the basic commands concerning database
398 2 - the basic commands concerning the working tree
399 3 - the other basic commands (e.g. ls-files)
400 4 - the diff commands
401 5 - the pull and exporting commands
402 6 - the revision tree commands (even e.g. merge-base)
403 7 - the porcelainish commands concerning the working tree
404 8 - the porcelainish commands concerning forensics
405 9 - the git tools
406
407 Second digit tells the particular command we are testing.
408
409 Third digit (optionally) tells the particular switch or group of switches
410 we are testing.
411
412 If you create files under t/ directory (i.e. here) that is not
413 the top-level test script, never name the file to match the above
414 pattern. The Makefile here considers all such files as the
415 top-level test script and tries to run all of them. Care is
416 especially needed if you are creating a common test library
417 file, similar to test-lib.sh, because such a library file may
418 not be suitable for standalone execution.
419
420
421 Writing Tests
422 -------------
423
424 The test script is written as a shell script. It should start
425 with the standard "#!/bin/sh", and an
426 assignment to variable 'test_description', like this:
427
428 #!/bin/sh
429
430 test_description='xxx test (option --frotz)
431
432 This test registers the following structure in the cache
433 and tries to run git-ls-files with option --frotz.'
434
435
436 Source 'test-lib.sh'
437 --------------------
438
439 After assigning test_description, the test script should source
440 test-lib.sh like this:
441
442 . ./test-lib.sh
443
444 This test harness library does the following things:
445
446 - If the script is invoked with command line argument --help
447 (or -h), it shows the test_description and exits.
448
449 - Creates an empty test directory with an empty .git/objects database
450 and chdir(2) into it. This directory is 't/trash
451 directory.$test_name_without_dotsh', with t/ subject to change by
452 the --root option documented above, and a '.stress-<N>' suffix
453 appended by the --stress option.
454
455 - Defines standard test helper functions for your scripts to
456 use. These functions are designed to make all scripts behave
457 consistently when command line arguments --verbose (or -v),
458 --debug (or -d), and --immediate (or -i) is given.
459
460 Do's & don'ts
461 -------------
462
463 Here are a few examples of things you probably should and shouldn't do
464 when writing tests.
465
466 Here are the "do's:"
467
468 - Put all code inside test_expect_success and other assertions.
469
470 Even code that isn't a test per se, but merely some setup code
471 should be inside a test assertion.
472
473 - Chain your test assertions
474
475 Write test code like this:
476
477 git merge foo &&
478 git push bar &&
479 test ...
480
481 Instead of:
482
483 git merge hla
484 git push gh
485 test ...
486
487 That way all of the commands in your tests will succeed or fail. If
488 you must ignore the return value of something, consider using a
489 helper function (e.g. use sane_unset instead of unset, in order
490 to avoid unportable return value for unsetting a variable that was
491 already unset), or prepending the command with test_might_fail or
492 test_must_fail.
493
494 - Check the test coverage for your tests. See the "Test coverage"
495 below.
496
497 Don't blindly follow test coverage metrics; if a new function you added
498 doesn't have any coverage, then you're probably doing something wrong,
499 but having 100% coverage doesn't necessarily mean that you tested
500 everything.
501
502 Tests that are likely to smoke out future regressions are better
503 than tests that just inflate the coverage metrics.
504
505 - When a test checks for an absolute path that a git command generated,
506 construct the expected value using $(pwd) rather than $PWD,
507 $TEST_DIRECTORY, or $TRASH_DIRECTORY. It makes a difference on
508 Windows, where the shell (MSYS bash) mangles absolute path names.
509 For details, see the commit message of 4114156ae9.
510
511 - Remember that inside the <script> part, the standard output and
512 standard error streams are discarded, and the test harness only
513 reports "ok" or "not ok" to the end user running the tests. Under
514 --verbose, they are shown to help debug the tests.
515
516 And here are the "don'ts:"
517
518 - Don't exit() within a <script> part.
519
520 The harness will catch this as a programming error of the test.
521 Use test_done instead if you need to stop the tests early (see
522 "Skipping tests" below).
523
524 - Don't use '! git cmd' when you want to make sure the git command
525 exits with failure in a controlled way by calling "die()". Instead,
526 use 'test_must_fail git cmd'. This will signal a failure if git
527 dies in an unexpected way (e.g. segfault).
528
529 On the other hand, don't use test_must_fail for running regular
530 platform commands; just use '! cmd'. We are not in the business
531 of verifying that the world given to us sanely works.
532
533 - Don't feed the output of a git command to a pipe, as in:
534
535 git -C repo ls-files |
536 xargs -n 1 basename |
537 grep foo
538
539 which will discard git's exit code and may mask a crash. In the
540 above example, all exit codes are ignored except grep's.
541
542 Instead, write the output of that command to a temporary
543 file with ">" or assign it to a variable with "x=$(git ...)" rather
544 than pipe it.
545
546 - Don't use command substitution in a way that discards git's exit
547 code. When assigning to a variable, the exit code is not discarded,
548 e.g.:
549
550 x=$(git cat-file -p $sha) &&
551 ...
552
553 is OK because a crash in "git cat-file" will cause the "&&" chain
554 to fail, but:
555
556 test "refs/heads/foo" = "$(git symbolic-ref HEAD)"
557
558 is not OK and a crash in git could go undetected.
559
560 - Don't use perl without spelling it as "$PERL_PATH". This is to help
561 our friends on Windows where the platform Perl often adds CR before
562 the end of line, and they bundle Git with a version of Perl that
563 does not do so, whose path is specified with $PERL_PATH. Note that we
564 provide a "perl" function which uses $PERL_PATH under the hood, so
565 you do not need to worry when simply running perl in the test scripts
566 (but you do, for example, on a shebang line or in a sub script
567 created via "write_script").
568
569 - Don't use sh without spelling it as "$SHELL_PATH", when the script
570 can be misinterpreted by broken platform shell (e.g. Solaris).
571
572 - Don't chdir around in tests. It is not sufficient to chdir to
573 somewhere and then chdir back to the original location later in
574 the test, as any intermediate step can fail and abort the test,
575 causing the next test to start in an unexpected directory. Do so
576 inside a subshell if necessary.
577
578 - Don't save and verify the standard error of compound commands, i.e.
579 group commands, subshells, and shell functions (except test helper
580 functions like 'test_must_fail') like this:
581
582 ( cd dir && git cmd ) 2>error &&
583 test_cmp expect error
584
585 When running the test with '-x' tracing, then the trace of commands
586 executed in the compound command will be included in standard error
587 as well, quite possibly throwing off the subsequent checks examining
588 the output. Instead, save only the relevant git command's standard
589 error:
590
591 ( cd dir && git cmd 2>../error ) &&
592 test_cmp expect error
593
594 - Don't break the TAP output
595
596 The raw output from your test may be interpreted by a TAP harness. TAP
597 harnesses will ignore everything they don't know about, but don't step
598 on their toes in these areas:
599
600 - Don't print lines like "$x..$y" where $x and $y are integers.
601
602 - Don't print lines that begin with "ok" or "not ok".
603
604 TAP harnesses expect a line that begins with either "ok" and "not
605 ok" to signal a test passed or failed (and our harness already
606 produces such lines), so your script shouldn't emit such lines to
607 their output.
608
609 You can glean some further possible issues from the TAP grammar
610 (see https://metacpan.org/pod/TAP::Parser::Grammar#TAP-GRAMMAR)
611 but the best indication is to just run the tests with prove(1),
612 it'll complain if anything is amiss.
613
614
615 Skipping tests
616 --------------
617
618 If you need to skip tests you should do so by using the three-arg form
619 of the test_* functions (see the "Test harness library" section
620 below), e.g.:
621
622 test_expect_success PERL 'I need Perl' '
623 perl -e "hlagh() if unf_unf()"
624 '
625
626 The advantage of skipping tests like this is that platforms that don't
627 have the PERL and other optional dependencies get an indication of how
628 many tests they're missing.
629
630 If the test code is too hairy for that (i.e. does a lot of setup work
631 outside test assertions) you can also skip all remaining tests by
632 setting skip_all and immediately call test_done:
633
634 if ! test_have_prereq PERL
635 then
636 skip_all='skipping perl interface tests, perl not available'
637 test_done
638 fi
639
640 The string you give to skip_all will be used as an explanation for why
641 the test was skipped.
642
643 End with test_done
644 ------------------
645
646 Your script will be a sequence of tests, using helper functions
647 from the test harness library. At the end of the script, call
648 'test_done'.
649
650
651 Test harness library
652 --------------------
653
654 There are a handful helper functions defined in the test harness
655 library for your script to use.
656
657 - test_expect_success [<prereq>] <message> <script>
658
659 Usually takes two strings as parameters, and evaluates the
660 <script>. If it yields success, test is considered
661 successful. <message> should state what it is testing.
662
663 Example:
664
665 test_expect_success \
666 'git-write-tree should be able to write an empty tree.' \
667 'tree=$(git-write-tree)'
668
669 If you supply three parameters the first will be taken to be a
670 prerequisite; see the test_set_prereq and test_have_prereq
671 documentation below:
672
673 test_expect_success TTY 'git --paginate rev-list uses a pager' \
674 ' ... '
675
676 You can also supply a comma-separated list of prerequisites, in the
677 rare case where your test depends on more than one:
678
679 test_expect_success PERL,PYTHON 'yo dawg' \
680 ' test $(perl -E 'print eval "1 +" . qx[python -c "print 2"]') == "4" '
681
682 - test_expect_failure [<prereq>] <message> <script>
683
684 This is NOT the opposite of test_expect_success, but is used
685 to mark a test that demonstrates a known breakage. Unlike
686 the usual test_expect_success tests, which say "ok" on
687 success and "FAIL" on failure, this will say "FIXED" on
688 success and "still broken" on failure. Failures from these
689 tests won't cause -i (immediate) to stop.
690
691 Like test_expect_success this function can optionally use a three
692 argument invocation with a prerequisite as the first argument.
693
694 - test_debug <script>
695
696 This takes a single argument, <script>, and evaluates it only
697 when the test script is started with --debug command line
698 argument. This is primarily meant for use during the
699 development of a new test script.
700
701 - debug <git-command>
702
703 Run a git command inside a debugger. This is primarily meant for
704 use when debugging a failing test script.
705
706 - test_done
707
708 Your test script must have test_done at the end. Its purpose
709 is to summarize successes and failures in the test script and
710 exit with an appropriate error code.
711
712 - test_tick
713
714 Make commit and tag names consistent by setting the author and
715 committer times to defined state. Subsequent calls will
716 advance the times by a fixed amount.
717
718 - test_commit <message> [<filename> [<contents>]]
719
720 Creates a commit with the given message, committing the given
721 file with the given contents (default for both is to reuse the
722 message string), and adds a tag (again reusing the message
723 string as name). Calls test_tick to make the SHA-1s
724 reproducible.
725
726 - test_merge <message> <commit-or-tag>
727
728 Merges the given rev using the given message. Like test_commit,
729 creates a tag and calls test_tick before committing.
730
731 - test_set_prereq <prereq>
732
733 Set a test prerequisite to be used later with test_have_prereq. The
734 test-lib will set some prerequisites for you, see the
735 "Prerequisites" section below for a full list of these.
736
737 Others you can set yourself and use later with either
738 test_have_prereq directly, or the three argument invocation of
739 test_expect_success and test_expect_failure.
740
741 - test_have_prereq <prereq>
742
743 Check if we have a prerequisite previously set with test_set_prereq.
744 The most common way to use this explicitly (as opposed to the
745 implicit use when an argument is passed to test_expect_*) is to skip
746 all the tests at the start of the test script if we don't have some
747 essential prerequisite:
748
749 if ! test_have_prereq PERL
750 then
751 skip_all='skipping perl interface tests, perl not available'
752 test_done
753 fi
754
755 - test_external [<prereq>] <message> <external> <script>
756
757 Execute a <script> with an <external> interpreter (like perl). This
758 was added for tests like t9700-perl-git.sh which do most of their
759 work in an external test script.
760
761 test_external \
762 'GitwebCache::*FileCache*' \
763 perl "$TEST_DIRECTORY"/t9503/test_cache_interface.pl
764
765 If the test is outputting its own TAP you should set the
766 test_external_has_tap variable somewhere before calling the first
767 test_external* function. See t9700-perl-git.sh for an example.
768
769 # The external test will outputs its own plan
770 test_external_has_tap=1
771
772 - test_external_without_stderr [<prereq>] <message> <external> <script>
773
774 Like test_external but fail if there's any output on stderr,
775 instead of checking the exit code.
776
777 test_external_without_stderr \
778 'Perl API' \
779 perl "$TEST_DIRECTORY"/t9700/test.pl
780
781 - test_expect_code <exit-code> <command>
782
783 Run a command and ensure that it exits with the given exit code.
784 For example:
785
786 test_expect_success 'Merge with d/f conflicts' '
787 test_expect_code 1 git merge "merge msg" B master
788 '
789
790 - test_must_fail [<options>] <git-command>
791
792 Run a git command and ensure it fails in a controlled way. Use
793 this instead of "! <git-command>". When git-command dies due to a
794 segfault, test_must_fail diagnoses it as an error; "! <git-command>"
795 treats it as just another expected failure, which would let such a
796 bug go unnoticed.
797
798 Accepts the following options:
799
800 ok=<signal-name>[,<...>]:
801 Don't treat an exit caused by the given signal as error.
802 Multiple signals can be specified as a comma separated list.
803 Currently recognized signal names are: sigpipe, success.
804 (Don't use 'success', use 'test_might_fail' instead.)
805
806 - test_might_fail [<options>] <git-command>
807
808 Similar to test_must_fail, but tolerate success, too. Use this
809 instead of "<git-command> || :" to catch failures due to segv.
810
811 Accepts the same options as test_must_fail.
812
813 - test_cmp <expected> <actual>
814
815 Check whether the content of the <actual> file matches the
816 <expected> file. This behaves like "cmp" but produces more
817 helpful output when the test is run with "-v" option.
818
819 - test_cmp_rev <expected> <actual>
820
821 Check whether the <expected> rev points to the same commit as the
822 <actual> rev.
823
824 - test_line_count (= | -lt | -ge | ...) <length> <file>
825
826 Check whether a file has the length it is expected to.
827
828 - test_path_is_file <path> [<diagnosis>]
829 test_path_is_dir <path> [<diagnosis>]
830 test_path_is_missing <path> [<diagnosis>]
831
832 Check if the named path is a file, if the named path is a
833 directory, or if the named path does not exist, respectively,
834 and fail otherwise, showing the <diagnosis> text.
835
836 - test_when_finished <script>
837
838 Prepend <script> to a list of commands to run to clean up
839 at the end of the current test. If some clean-up command
840 fails, the test will not pass.
841
842 Example:
843
844 test_expect_success 'branch pointing to non-commit' '
845 git rev-parse HEAD^{tree} >.git/refs/heads/invalid &&
846 test_when_finished "git update-ref -d refs/heads/invalid" &&
847 ...
848 '
849
850 - test_write_lines <lines>
851
852 Write <lines> on standard output, one line per argument.
853 Useful to prepare multi-line files in a compact form.
854
855 Example:
856
857 test_write_lines a b c d e f g >foo
858
859 Is a more compact equivalent of:
860 cat >foo <<-EOF
861 a
862 b
863 c
864 d
865 e
866 f
867 g
868 EOF
869
870
871 - test_pause
872
873 This command is useful for writing and debugging tests and must be
874 removed before submitting. It halts the execution of the test and
875 spawns a shell in the trash directory. Exit the shell to continue
876 the test. Example:
877
878 test_expect_success 'test' '
879 git do-something >actual &&
880 test_pause &&
881 test_cmp expected actual
882 '
883
884 - test_ln_s_add <path1> <path2>
885
886 This function helps systems whose filesystem does not support symbolic
887 links. Use it to add a symbolic link entry to the index when it is not
888 important that the file system entry is a symbolic link, i.e., instead
889 of the sequence
890
891 ln -s foo bar &&
892 git add bar
893
894 Sometimes it is possible to split a test in a part that does not need
895 the symbolic link in the file system and a part that does; then only
896 the latter part need be protected by a SYMLINKS prerequisite (see below).
897
898 - test_oid_init
899
900 This function loads facts and useful object IDs related to the hash
901 algorithm(s) in use from the files in t/oid-info.
902
903 - test_oid_cache
904
905 This function reads per-hash algorithm information from standard
906 input (usually a heredoc) in the format described in
907 t/oid-info/README. This is useful for test-specific values, such as
908 object IDs, which must vary based on the hash algorithm.
909
910 Certain fixed values, such as hash sizes and common placeholder
911 object IDs, can be loaded with test_oid_init (described above).
912
913 - test_oid <key>
914
915 This function looks up a value for the hash algorithm in use, based
916 on the key given. The value must have been loaded using
917 test_oid_init or test_oid_cache. Providing an unknown key is an
918 error.
919
920 Prerequisites
921 -------------
922
923 These are the prerequisites that the test library predefines with
924 test_have_prereq.
925
926 See the prereq argument to the test_* functions in the "Test harness
927 library" section above and the "test_have_prereq" function for how to
928 use these, and "test_set_prereq" for how to define your own.
929
930 - PYTHON
931
932 Git wasn't compiled with NO_PYTHON=YesPlease. Wrap any tests that
933 need Python with this.
934
935 - PERL
936
937 Git wasn't compiled with NO_PERL=YesPlease.
938
939 Even without the PERL prerequisite, tests can assume there is a
940 usable perl interpreter at $PERL_PATH, though it need not be
941 particularly modern.
942
943 - POSIXPERM
944
945 The filesystem supports POSIX style permission bits.
946
947 - BSLASHPSPEC
948
949 Backslashes in pathspec are not directory separators. This is not
950 set on Windows. See 6fd1106a for details.
951
952 - EXECKEEPSPID
953
954 The process retains the same pid across exec(2). See fb9a2bea for
955 details.
956
957 - PIPE
958
959 The filesystem we're on supports creation of FIFOs (named pipes)
960 via mkfifo(1).
961
962 - SYMLINKS
963
964 The filesystem we're on supports symbolic links. E.g. a FAT
965 filesystem doesn't support these. See 704a3143 for details.
966
967 - SANITY
968
969 Test is not run by root user, and an attempt to write to an
970 unwritable file is expected to fail correctly.
971
972 - PCRE
973
974 Git was compiled with support for PCRE. Wrap any tests
975 that use git-grep --perl-regexp or git-grep -P in these.
976
977 - LIBPCRE1
978
979 Git was compiled with PCRE v1 support via
980 USE_LIBPCRE1=YesPlease. Wrap any PCRE using tests that for some
981 reason need v1 of the PCRE library instead of v2 in these.
982
983 - LIBPCRE2
984
985 Git was compiled with PCRE v2 support via
986 USE_LIBPCRE2=YesPlease. Wrap any PCRE using tests that for some
987 reason need v2 of the PCRE library instead of v1 in these.
988
989 - CASE_INSENSITIVE_FS
990
991 Test is run on a case insensitive file system.
992
993 - UTF8_NFD_TO_NFC
994
995 Test is run on a filesystem which converts decomposed utf-8 (nfd)
996 to precomposed utf-8 (nfc).
997
998 - PTHREADS
999
1000 Git wasn't compiled with NO_PTHREADS=YesPlease.
1001
1002 Tips for Writing Tests
1003 ----------------------
1004
1005 As with any programming projects, existing programs are the best
1006 source of the information. However, do _not_ emulate
1007 t0000-basic.sh when writing your tests. The test is special in
1008 that it tries to validate the very core of GIT. For example, it
1009 knows that there will be 256 subdirectories under .git/objects/,
1010 and it knows that the object ID of an empty tree is a certain
1011 40-byte string. This is deliberately done so in t0000-basic.sh
1012 because the things the very basic core test tries to achieve is
1013 to serve as a basis for people who are changing the GIT internal
1014 drastically. For these people, after making certain changes,
1015 not seeing failures from the basic test _is_ a failure. And
1016 such drastic changes to the core GIT that even changes these
1017 otherwise supposedly stable object IDs should be accompanied by
1018 an update to t0000-basic.sh.
1019
1020 However, other tests that simply rely on basic parts of the core
1021 GIT working properly should not have that level of intimate
1022 knowledge of the core GIT internals. If all the test scripts
1023 hardcoded the object IDs like t0000-basic.sh does, that defeats
1024 the purpose of t0000-basic.sh, which is to isolate that level of
1025 validation in one place. Your test also ends up needing
1026 updating when such a change to the internal happens, so do _not_
1027 do it and leave the low level of validation to t0000-basic.sh.
1028
1029 Test coverage
1030 -------------
1031
1032 You can use the coverage tests to find code paths that are not being
1033 used or properly exercised yet.
1034
1035 To do that, run the coverage target at the top-level (not in the t/
1036 directory):
1037
1038 make coverage
1039
1040 That'll compile Git with GCC's coverage arguments, and generate a test
1041 report with gcov after the tests finish. Running the coverage tests
1042 can take a while, since running the tests in parallel is incompatible
1043 with GCC's coverage mode.
1044
1045 After the tests have run you can generate a list of untested
1046 functions:
1047
1048 make coverage-untested-functions
1049
1050 You can also generate a detailed per-file HTML report using the
1051 Devel::Cover module. To install it do:
1052
1053 # On Debian or Ubuntu:
1054 sudo aptitude install libdevel-cover-perl
1055
1056 # From the CPAN with cpanminus
1057 curl -L http://cpanmin.us | perl - --sudo --self-upgrade
1058 cpanm --sudo Devel::Cover
1059
1060 Then, at the top-level:
1061
1062 make cover_db_html
1063
1064 That'll generate a detailed cover report in the "cover_db_html"
1065 directory, which you can then copy to a webserver, or inspect locally
1066 in a browser.