t5304: use helper to report failure of "test foo = bar"
[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
5e3b4fce 85-d::
986aa7f1
JH
86--debug::
87 This may help the person who is developing a new test.
88 It causes the command defined with test_debug to run.
0986de94
PK
89 The "trash" directory (used to store all temporary data
90 during testing) is not deleted even if there are no
91 failed tests so that you can inspect its contents after
92 the test finished.
986aa7f1 93
5e3b4fce 94-i::
986aa7f1
JH
95--immediate::
96 This causes the test to immediately exit upon the first
13cb3bb7
SR
97 failed test. Cleanup commands requested with
98 test_when_finished are not executed if the test failed,
99 in order to keep the state for inspection by the tester
100 to diagnose the bug.
986aa7f1 101
5e3b4fce 102-l::
5e2c08c6
LW
103--long-tests::
104 This causes additional long-running tests to be run (where
105 available), for more exhaustive testing.
106
0445e6f0
IB
107-r::
108--run=<test-selector>::
109 Run only the subset of tests indicated by
110 <test-selector>. See section "Skipping Tests" below for
111 <test-selector> syntax.
112
952af351
TR
113--valgrind=<tool>::
114 Execute all Git binaries under valgrind tool <tool> and exit
115 with status 126 on errors (just like regular tests, this will
116 only stop the test script when running under -i).
986aa7f1 117
3da93652
JS
118 Since it makes no sense to run the tests with --valgrind and
119 not see any output, this option implies --verbose. For
120 convenience, it also implies --tee.
121
952af351
TR
122 <tool> defaults to 'memcheck', just like valgrind itself.
123 Other particularly useful choices include 'helgrind' and
124 'drd', but you may use any tool recognized by your valgrind
125 installation.
126
95d9d5ec
TR
127 As a special case, <tool> can be 'memcheck-fast', which uses
128 memcheck but disables --track-origins. Use this if you are
129 running tests in bulk, to see if there are _any_ memory
130 issues.
131
952af351 132 Note that memcheck is run with the option --leak-check=no,
9aec68d3
CMN
133 as the git process is short-lived and some errors are not
134 interesting. In order to run a single command under the same
135 conditions manually, you should set GIT_VALGRIND to point to
136 the 't/valgrind/' directory and use the commands under
137 't/valgrind/bin/'.
138
5dfc368f
TR
139--valgrind-only=<pattern>::
140 Like --valgrind, but the effect is limited to tests with
141 numbers matching <pattern>. The number matched against is
142 simply the running count of the test within the file.
143
44138559
JS
144--tee::
145 In addition to printing the test output to the terminal,
146 write it to files named 't/test-results/$TEST_NAME.out'.
147 As the names depend on the tests' file names, it is safe to
148 run the tests with this option in parallel.
149
e4597aae
MO
150--with-dashes::
151 By default tests are run without dashed forms of
152 commands (like git-commit) in the PATH (it only uses
153 wrappers from ../bin-wrappers). Use this option to include
154 the build directory (..) in the PATH, which contains all
155 the dashed forms of commands. This option is currently
156 implied by other options like --valgrind and
157 GIT_TEST_INSTALLED.
158
0d4dbcd3
TR
159--root=<directory>::
160 Create "trash" directories used to store all temporary data during
161 testing under <directory>, instead of the t/ directory.
162 Using this option with a RAM-based filesystem (such as tmpfs)
163 can massively speed up the test suite.
164
e160da7f
MO
165You can also set the GIT_TEST_INSTALLED environment variable to
166the bindir of an existing git installation to test that installation.
167You still need to have built this git sandbox, from which various
168test-* support programs, templates, and perl libraries are used.
169If your installed git is incomplete, it will silently test parts of
170your built version instead.
171
172When using GIT_TEST_INSTALLED, you can also set GIT_TEST_EXEC_PATH to
173override the location of the dashed-form subcommands (what
174GIT_EXEC_PATH would be used for during normal operation).
175GIT_TEST_EXEC_PATH defaults to `$GIT_TEST_INSTALLED/git --exec-path`.
176
177
fbd458a3
JN
178Skipping Tests
179--------------
180
181In some environments, certain tests have no way of succeeding
182due to platform limitation, such as lack of 'unzip' program, or
183filesystem that do not allow arbitrary sequence of non-NUL bytes
184as pathnames.
185
186You should be able to say something like
187
188 $ GIT_SKIP_TESTS=t9200.8 sh ./t9200-git-cvsexport-commit.sh
189
190and even:
191
192 $ GIT_SKIP_TESTS='t[0-4]??? t91?? t9200.8' make
193
194to omit such tests. The value of the environment variable is a
195SP separated list of patterns that tells which tests to skip,
196and either can match the "t[0-9]{4}" part to skip the whole
197test, or t[0-9]{4} followed by ".$number" to say which
198particular test to skip.
199
0445e6f0
IB
200For an individual test suite --run could be used to specify that
201only some tests should be run or that some tests should be
202excluded from a run.
203
204The argument for --run is a list of individual test numbers or
205ranges with an optional negation prefix that define what tests in
206a test suite to include in the run. A range is two numbers
207separated with a dash and matches a range of tests with both ends
208been included. You may omit the first or the second number to
209mean "from the first test" or "up to the very last test"
210respectively.
211
212Optional prefix of '!' means that the test or a range of tests
213should be excluded from the run.
214
215If --run starts with an unprefixed number or range the initial
216set of tests to run is empty. If the first item starts with '!'
217all the tests are added to the initial set. After initial set is
218determined every test number or range is added or excluded from
219the set one by one, from left to right.
220
221Individual numbers or ranges could be separated either by a space
222or a comma.
223
224For example, to run only tests up to a specific test (21), one
225could do this:
226
227 $ sh ./t9200-git-cvsexport-commit.sh --run='1-21'
228
229or this:
230
231 $ sh ./t9200-git-cvsexport-commit.sh --run='-21'
232
233Common case is to run several setup tests (1, 2, 3) and then a
234specific test (21) that relies on that setup:
235
236 $ sh ./t9200-git-cvsexport-commit.sh --run='1 2 3 21'
237
238or:
239
240 $ sh ./t9200-git-cvsexport-commit.sh --run=1,2,3,21
241
242or:
243
244 $ sh ./t9200-git-cvsexport-commit.sh --run='-3 21'
245
246As noted above, the test set is built going though items left to
247right, so this:
248
249 $ sh ./t9200-git-cvsexport-commit.sh --run='1-4 !3'
250
251will run tests 1, 2, and 4. Items that comes later have higher
252precendence. It means that this:
253
254 $ sh ./t9200-git-cvsexport-commit.sh --run='!3 1-4'
255
256would just run tests from 1 to 4, including 3.
257
258You may use negation with ranges. The following will run all
259test in the test suite except from 7 up to 11:
260
261 $ sh ./t9200-git-cvsexport-commit.sh --run='!7-11'
262
263Some tests in a test suite rely on the previous tests performing
264certain actions, specifically some tests are designated as
265"setup" test, so you cannot _arbitrarily_ disable one test and
266expect the rest to function correctly.
267
268--run is mostly useful when you want to focus on a specific test
269and know what setup is needed for it. Or when you want to run
270everything up to a certain test.
fbd458a3
JN
271
272
f50c9f76
PB
273Naming Tests
274------------
275
276The test files are named as:
277
278 tNNNN-commandname-details.sh
279
280where N is a decimal digit.
281
282First digit tells the family:
283
284 0 - the absolute basics and global stuff
285 1 - the basic commands concerning database
286 2 - the basic commands concerning the working tree
287 3 - the other basic commands (e.g. ls-files)
288 4 - the diff commands
289 5 - the pull and exporting commands
290 6 - the revision tree commands (even e.g. merge-base)
8f4a9b62 291 7 - the porcelainish commands concerning the working tree
8757749e
JN
292 8 - the porcelainish commands concerning forensics
293 9 - the git tools
f50c9f76
PB
294
295Second digit tells the particular command we are testing.
296
297Third digit (optionally) tells the particular switch or group of switches
298we are testing.
299
77656600
JH
300If you create files under t/ directory (i.e. here) that is not
301the top-level test script, never name the file to match the above
302pattern. The Makefile here considers all such files as the
63d32945 303top-level test script and tries to run all of them. Care is
77656600
JH
304especially needed if you are creating a common test library
305file, similar to test-lib.sh, because such a library file may
306not be suitable for standalone execution.
307
f50c9f76 308
986aa7f1
JH
309Writing Tests
310-------------
311
312The test script is written as a shell script. It should start
313with the standard "#!/bin/sh" with copyright notices, and an
314assignment to variable 'test_description', like this:
315
316 #!/bin/sh
317 #
318 # Copyright (c) 2005 Junio C Hamano
319 #
320
14cd1ff3 321 test_description='xxx test (option --frotz)
986aa7f1
JH
322
323 This test registers the following structure in the cache
324 and tries to run git-ls-files with option --frotz.'
325
f50c9f76 326
986aa7f1
JH
327Source 'test-lib.sh'
328--------------------
329
330After assigning test_description, the test script should source
331test-lib.sh like this:
332
333 . ./test-lib.sh
334
335This test harness library does the following things:
336
337 - If the script is invoked with command line argument --help
338 (or -h), it shows the test_description and exits.
339
e1ca1c9d
ÆAB
340 - Creates an empty test directory with an empty .git/objects database
341 and chdir(2) into it. This directory is 't/trash
342 directory.$test_name_without_dotsh', with t/ subject to change by
343 the --root option documented above.
986aa7f1
JH
344
345 - Defines standard test helper functions for your scripts to
346 use. These functions are designed to make all scripts behave
347 consistently when command line arguments --verbose (or -v),
348 --debug (or -d), and --immediate (or -i) is given.
349
20873f45
ÆAB
350Do's, don'ts & things to keep in mind
351-------------------------------------
352
6fd45295 353Here are a few examples of things you probably should and shouldn't do
20873f45
ÆAB
354when writing tests.
355
356Do:
357
6fd45295 358 - Put all code inside test_expect_success and other assertions.
20873f45
ÆAB
359
360 Even code that isn't a test per se, but merely some setup code
6fd45295 361 should be inside a test assertion.
20873f45
ÆAB
362
363 - Chain your test assertions
364
365 Write test code like this:
366
367 git merge foo &&
368 git push bar &&
369 test ...
370
371 Instead of:
372
373 git merge hla
374 git push gh
375 test ...
376
377 That way all of the commands in your tests will succeed or fail. If
00648ba0
EN
378 you must ignore the return value of something, consider using a
379 helper function (e.g. use sane_unset instead of unset, in order
380 to avoid unportable return value for unsetting a variable that was
381 already unset), or prepending the command with test_might_fail or
382 test_must_fail.
20873f45 383
0c357544
ÆAB
384 - Check the test coverage for your tests. See the "Test coverage"
385 below.
386
63d32945
MW
387 Don't blindly follow test coverage metrics; if a new function you added
388 doesn't have any coverage, then you're probably doing something wrong,
e8b55f5c
ÆAB
389 but having 100% coverage doesn't necessarily mean that you tested
390 everything.
391
392 Tests that are likely to smoke out future regressions are better
393 than tests that just inflate the coverage metrics.
394
95b104c8
JS
395 - When a test checks for an absolute path that a git command generated,
396 construct the expected value using $(pwd) rather than $PWD,
397 $TEST_DIRECTORY, or $TRASH_DIRECTORY. It makes a difference on
398 Windows, where the shell (MSYS bash) mangles absolute path names.
399 For details, see the commit message of 4114156ae9.
400
20873f45
ÆAB
401Don't:
402
403 - exit() within a <script> part.
404
405 The harness will catch this as a programming error of the test.
406 Use test_done instead if you need to stop the tests early (see
407 "Skipping tests" below).
408
ad78585e
JH
409 - use '! git cmd' when you want to make sure the git command exits
410 with failure in a controlled way by calling "die()". Instead,
411 use 'test_must_fail git cmd'. This will signal a failure if git
412 dies in an unexpected way (e.g. segfault).
413
f445500e
JH
414 On the other hand, don't use test_must_fail for running regular
415 platform commands; just use '! cmd'.
416
ad78585e
JH
417 - use perl without spelling it as "$PERL_PATH". This is to help our
418 friends on Windows where the platform Perl often adds CR before
419 the end of line, and they bundle Git with a version of Perl that
a0e0ec9f
JK
420 does not do so, whose path is specified with $PERL_PATH. Note that we
421 provide a "perl" function which uses $PERL_PATH under the hood, so
422 you do not need to worry when simply running perl in the test scripts
423 (but you do, for example, on a shebang line or in a sub script
424 created via "write_script").
ad78585e
JH
425
426 - use sh without spelling it as "$SHELL_PATH", when the script can
427 be misinterpreted by broken platform shell (e.g. Solaris).
428
429 - chdir around in tests. It is not sufficient to chdir to
430 somewhere and then chdir back to the original location later in
431 the test, as any intermediate step can fail and abort the test,
432 causing the next test to start in an unexpected directory. Do so
433 inside a subshell if necessary.
434
20873f45
ÆAB
435 - Break the TAP output
436
6fd45295
JH
437 The raw output from your test may be interpreted by a TAP harness. TAP
438 harnesses will ignore everything they don't know about, but don't step
439 on their toes in these areas:
20873f45
ÆAB
440
441 - Don't print lines like "$x..$y" where $x and $y are integers.
442
443 - Don't print lines that begin with "ok" or "not ok".
444
6fd45295 445 TAP harnesses expect a line that begins with either "ok" and "not
20873f45
ÆAB
446 ok" to signal a test passed or failed (and our harness already
447 produces such lines), so your script shouldn't emit such lines to
448 their output.
449
450 You can glean some further possible issues from the TAP grammar
451 (see http://search.cpan.org/perldoc?TAP::Parser::Grammar#TAP_Grammar)
452 but the best indication is to just run the tests with prove(1),
453 it'll complain if anything is amiss.
454
455Keep in mind:
456
6fd45295 457 - Inside <script> part, the standard output and standard error
20873f45
ÆAB
458 streams are discarded, and the test harness only reports "ok" or
459 "not ok" to the end user running the tests. Under --verbose, they
460 are shown to help debugging the tests.
461
462
b5500d16
ÆAB
463Skipping tests
464--------------
465
681186ae 466If you need to skip tests you should do so by using the three-arg form
99d9050d
ÆAB
467of the test_* functions (see the "Test harness library" section
468below), e.g.:
469
ad78585e 470 test_expect_success PERL 'I need Perl' '
a0e0ec9f 471 perl -e "hlagh() if unf_unf()"
ad78585e 472 '
99d9050d
ÆAB
473
474The advantage of skipping tests like this is that platforms that don't
475have the PERL and other optional dependencies get an indication of how
476many tests they're missing.
477
478If the test code is too hairy for that (i.e. does a lot of setup work
479outside test assertions) you can also skip all remaining tests by
480setting skip_all and immediately call test_done:
b5500d16
ÆAB
481
482 if ! test_have_prereq PERL
483 then
484 skip_all='skipping perl interface tests, perl not available'
485 test_done
486 fi
14cd1ff3 487
99d9050d
ÆAB
488The string you give to skip_all will be used as an explanation for why
489the test was skipped.
490
986aa7f1
JH
491End with test_done
492------------------
493
494Your script will be a sequence of tests, using helper functions
495from the test harness library. At the end of the script, call
496'test_done'.
497
498
499Test harness library
500--------------------
501
502There are a handful helper functions defined in the test harness
503library for your script to use.
504
9a897893 505 - test_expect_success [<prereq>] <message> <script>
986aa7f1 506
72942a61 507 Usually takes two strings as parameters, and evaluates the
986aa7f1
JH
508 <script>. If it yields success, test is considered
509 successful. <message> should state what it is testing.
510
511 Example:
512
513 test_expect_success \
514 'git-write-tree should be able to write an empty tree.' \
515 'tree=$(git-write-tree)'
516
9a897893 517 If you supply three parameters the first will be taken to be a
72942a61 518 prerequisite; see the test_set_prereq and test_have_prereq
9a897893
ÆAB
519 documentation below:
520
521 test_expect_success TTY 'git --paginate rev-list uses a pager' \
522 ' ... '
523
93a57246
ÆAB
524 You can also supply a comma-separated list of prerequisites, in the
525 rare case where your test depends on more than one:
526
527 test_expect_success PERL,PYTHON 'yo dawg' \
528 ' test $(perl -E 'print eval "1 +" . qx[python -c "print 2"]') == "4" '
529
9a897893 530 - test_expect_failure [<prereq>] <message> <script>
986aa7f1 531
41ac414e
JH
532 This is NOT the opposite of test_expect_success, but is used
533 to mark a test that demonstrates a known breakage. Unlike
534 the usual test_expect_success tests, which say "ok" on
535 success and "FAIL" on failure, this will say "FIXED" on
536 success and "still broken" on failure. Failures from these
537 tests won't cause -i (immediate) to stop.
986aa7f1 538
9a897893
ÆAB
539 Like test_expect_success this function can optionally use a three
540 argument invocation with a prerequisite as the first argument.
541
986aa7f1
JH
542 - test_debug <script>
543
544 This takes a single argument, <script>, and evaluates it only
545 when the test script is started with --debug command line
546 argument. This is primarily meant for use during the
547 development of a new test script.
548
549 - test_done
550
551 Your test script must have test_done at the end. Its purpose
552 is to summarize successes and failures in the test script and
553 exit with an appropriate error code.
554
00884968
JS
555 - test_tick
556
557 Make commit and tag names consistent by setting the author and
63d32945 558 committer times to defined state. Subsequent calls will
00884968
JS
559 advance the times by a fixed amount.
560
561 - test_commit <message> [<filename> [<contents>]]
562
563 Creates a commit with the given message, committing the given
564 file with the given contents (default for both is to reuse the
565 message string), and adds a tag (again reusing the message
566 string as name). Calls test_tick to make the SHA-1s
567 reproducible.
568
569 - test_merge <message> <commit-or-tag>
570
571 Merges the given rev using the given message. Like test_commit,
572 creates a tag and calls test_tick before committing.
986aa7f1 573
72942a61 574 - test_set_prereq <prereq>
9a897893
ÆAB
575
576 Set a test prerequisite to be used later with test_have_prereq. The
be53deef
ÆAB
577 test-lib will set some prerequisites for you, see the
578 "Prerequisites" section below for a full list of these.
579
580 Others you can set yourself and use later with either
581 test_have_prereq directly, or the three argument invocation of
582 test_expect_success and test_expect_failure.
9a897893 583
72942a61 584 - test_have_prereq <prereq>
9a897893
ÆAB
585
586 Check if we have a prerequisite previously set with
587 test_set_prereq. The most common use of this directly is to skip
588 all the tests if we don't have some essential prerequisite:
589
590 if ! test_have_prereq PERL
591 then
592 skip_all='skipping perl interface tests, perl not available'
593 test_done
594 fi
595
2fac6a4b
ÆAB
596 - test_external [<prereq>] <message> <external> <script>
597
598 Execute a <script> with an <external> interpreter (like perl). This
599 was added for tests like t9700-perl-git.sh which do most of their
600 work in an external test script.
601
602 test_external \
603 'GitwebCache::*FileCache*' \
a0e0ec9f 604 perl "$TEST_DIRECTORY"/t9503/test_cache_interface.pl
2fac6a4b
ÆAB
605
606 If the test is outputting its own TAP you should set the
607 test_external_has_tap variable somewhere before calling the first
608 test_external* function. See t9700-perl-git.sh for an example.
609
610 # The external test will outputs its own plan
611 test_external_has_tap=1
612
613 - test_external_without_stderr [<prereq>] <message> <external> <script>
614
615 Like test_external but fail if there's any output on stderr,
616 instead of checking the exit code.
617
618 test_external_without_stderr \
619 'Perl API' \
a0e0ec9f 620 perl "$TEST_DIRECTORY"/t9700/test.pl
2fac6a4b 621
892e6f7e
ÆAB
622 - test_expect_code <exit-code> <command>
623
624 Run a command and ensure that it exits with the given exit code.
625 For example:
626
627 test_expect_success 'Merge with d/f conflicts' '
628 test_expect_code 1 git merge "merge msg" B master
629 '
630
c9667456
JN
631 - test_must_fail <git-command>
632
633 Run a git command and ensure it fails in a controlled way. Use
971ecbd1
BC
634 this instead of "! <git-command>". When git-command dies due to a
635 segfault, test_must_fail diagnoses it as an error; "! <git-command>"
636 treats it as just another expected failure, which would let such a
637 bug go unnoticed.
c9667456
JN
638
639 - test_might_fail <git-command>
640
641 Similar to test_must_fail, but tolerate success, too. Use this
642 instead of "<git-command> || :" to catch failures due to segv.
643
644 - test_cmp <expected> <actual>
645
646 Check whether the content of the <actual> file matches the
647 <expected> file. This behaves like "cmp" but produces more
648 helpful output when the test is run with "-v" option.
649
fb3340a6
JN
650 - test_line_count (= | -lt | -ge | ...) <length> <file>
651
652 Check whether a file has the length it is expected to.
653
72942a61
ML
654 - test_path_is_file <path> [<diagnosis>]
655 test_path_is_dir <path> [<diagnosis>]
2caf20c5
MM
656 test_path_is_missing <path> [<diagnosis>]
657
72942a61
ML
658 Check if the named path is a file, if the named path is a
659 directory, or if the named path does not exist, respectively,
660 and fail otherwise, showing the <diagnosis> text.
2caf20c5 661
c9667456
JN
662 - test_when_finished <script>
663
664 Prepend <script> to a list of commands to run to clean up
665 at the end of the current test. If some clean-up command
666 fails, the test will not pass.
667
668 Example:
669
670 test_expect_success 'branch pointing to non-commit' '
671 git rev-parse HEAD^{tree} >.git/refs/heads/invalid &&
672 test_when_finished "git update-ref -d refs/heads/invalid" &&
673 ...
674 '
675
bb98b01e 676 - test_write_lines <lines>
ac9afcc3 677
bb98b01e 678 Write <lines> on standard output, one line per argument.
ac9afcc3
MT
679 Useful to prepare multi-line files in a compact form.
680
681 Example:
682
bb98b01e 683 test_write_lines a b c d e f g >foo
ac9afcc3
MT
684
685 Is a more compact equivalent of:
686 cat >foo <<-EOF
687 a
688 b
689 c
690 d
691 e
692 f
693 g
694 EOF
695
696
c4d2539a
JL
697 - test_pause
698
699 This command is useful for writing and debugging tests and must be
700 removed before submitting. It halts the execution of the test and
701 spawns a shell in the trash directory. Exit the shell to continue
702 the test. Example:
703
704 test_expect_success 'test' '
705 git do-something >actual &&
706 test_pause &&
707 test_cmp expected actual
708 '
709
9ce415d9
JS
710 - test_ln_s_add <path1> <path2>
711
712 This function helps systems whose filesystem does not support symbolic
713 links. Use it to add a symbolic link entry to the index when it is not
714 important that the file system entry is a symbolic link, i.e., instead
715 of the sequence
716
717 ln -s foo bar &&
718 git add bar
719
720 Sometimes it is possible to split a test in a part that does not need
721 the symbolic link in the file system and a part that does; then only
722 the latter part need be protected by a SYMLINKS prerequisite (see below).
723
be53deef
ÆAB
724Prerequisites
725-------------
726
727These are the prerequisites that the test library predefines with
728test_have_prereq.
729
730See the prereq argument to the test_* functions in the "Test harness
731library" section above and the "test_have_prereq" function for how to
732use these, and "test_set_prereq" for how to define your own.
733
f8fc0ee3 734 - PYTHON
be53deef 735
f8fc0ee3
JN
736 Git wasn't compiled with NO_PYTHON=YesPlease. Wrap any tests that
737 need Python with this.
738
739 - PERL
740
741 Git wasn't compiled with NO_PERL=YesPlease.
742
743 Even without the PERL prerequisite, tests can assume there is a
744 usable perl interpreter at $PERL_PATH, though it need not be
745 particularly modern.
be53deef
ÆAB
746
747 - POSIXPERM
748
749 The filesystem supports POSIX style permission bits.
750
751 - BSLASHPSPEC
752
753 Backslashes in pathspec are not directory separators. This is not
754 set on Windows. See 6fd1106a for details.
755
756 - EXECKEEPSPID
757
758 The process retains the same pid across exec(2). See fb9a2bea for
759 details.
760
20073274
AS
761 - PIPE
762
763 The filesystem we're on supports creation of FIFOs (named pipes)
764 via mkfifo(1).
765
be53deef
ÆAB
766 - SYMLINKS
767
768 The filesystem we're on supports symbolic links. E.g. a FAT
769 filesystem doesn't support these. See 704a3143 for details.
2fac6a4b 770
c91cfd19
ÆAB
771 - SANITY
772
773 Test is not run by root user, and an attempt to write to an
774 unwritable file is expected to fail correctly.
2fac6a4b 775
8f852ce6
MK
776 - LIBPCRE
777
778 Git was compiled with USE_LIBPCRE=YesPlease. Wrap any tests
779 that use git-grep --perl-regexp or git-grep -P in these.
780
ac39aa61
MG
781 - CASE_INSENSITIVE_FS
782
783 Test is run on a case insensitive file system.
784
5b0b5dd4
MG
785 - UTF8_NFD_TO_NFC
786
787 Test is run on a filesystem which converts decomposed utf-8 (nfd)
788 to precomposed utf-8 (nfc).
789
986aa7f1
JH
790Tips for Writing Tests
791----------------------
792
793As with any programming projects, existing programs are the best
794source of the information. However, do _not_ emulate
795t0000-basic.sh when writing your tests. The test is special in
796that it tries to validate the very core of GIT. For example, it
797knows that there will be 256 subdirectories under .git/objects/,
798and it knows that the object ID of an empty tree is a certain
79940-byte string. This is deliberately done so in t0000-basic.sh
800because the things the very basic core test tries to achieve is
801to serve as a basis for people who are changing the GIT internal
802drastically. For these people, after making certain changes,
803not seeing failures from the basic test _is_ a failure. And
804such drastic changes to the core GIT that even changes these
805otherwise supposedly stable object IDs should be accompanied by
806an update to t0000-basic.sh.
807
808However, other tests that simply rely on basic parts of the core
809GIT working properly should not have that level of intimate
810knowledge of the core GIT internals. If all the test scripts
811hardcoded the object IDs like t0000-basic.sh does, that defeats
812the purpose of t0000-basic.sh, which is to isolate that level of
813validation in one place. Your test also ends up needing
814updating when such a change to the internal happens, so do _not_
815do it and leave the low level of validation to t0000-basic.sh.
d15e9ebc 816
0c357544
ÆAB
817Test coverage
818-------------
819
820You can use the coverage tests to find code paths that are not being
821used or properly exercised yet.
822
823To do that, run the coverage target at the top-level (not in the t/
824directory):
825
826 make coverage
827
828That'll compile Git with GCC's coverage arguments, and generate a test
829report with gcov after the tests finish. Running the coverage tests
830can take a while, since running the tests in parallel is incompatible
831with GCC's coverage mode.
832
833After the tests have run you can generate a list of untested
834functions:
835
836 make coverage-untested-functions
837
838You can also generate a detailed per-file HTML report using the
839Devel::Cover module. To install it do:
840
841 # On Debian or Ubuntu:
842 sudo aptitude install libdevel-cover-perl
843
844 # From the CPAN with cpanminus
845 curl -L http://cpanmin.us | perl - --sudo --self-upgrade
846 cpanm --sudo Devel::Cover
847
848Then, at the top-level:
849
850 make cover_db_html
851
852That'll generate a detailed cover report in the "cover_db_html"
853directory, which you can then copy to a webserver, or inspect locally
854in a browser.