]>
Commit | Line | Data |
---|---|---|
788db145 | 1 | Core Git Tests |
986aa7f1 JH |
2 | ============== |
3 | ||
788db145 | 4 | This directory holds many test scripts for core Git tools. The |
986aa7f1 JH |
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 *** | |
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 | ||
35 | Since the tests all output TAP (see http://testanything.org) they can | |
85b0b34e | 36 | be run with any TAP harness. Here's an example of parallel testing |
5099b99d ÆAB |
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 | |
986aa7f1 | 52 | |
28d836c8 MG |
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 | ||
5099b99d | 59 | You 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 | |
70 | You can pass --verbose (or -v), --debug (or -d), and --immediate | |
4e1be63c | 71 | (or -i) command line argument to the test, or by setting GIT_TEST_OPTS |
78dc0887 MT |
72 | appropriately before running "make". Short options can be bundled, i.e. |
73 | '-d -v' is the same as '-dv'. | |
986aa7f1 | 74 | |
5e3b4fce | 75 | -v:: |
986aa7f1 JH |
76 | --verbose:: |
77 | This makes the test more verbose. Specifically, the | |
78 | command being run and their output if any are also | |
79 | output. | |
80 | ||
ff09af3f TR |
81 | --verbose-only=<pattern>:: |
82 | Like --verbose, but the effect is limited to tests with | |
83 | numbers matching <pattern>. The number matched against is | |
84 | simply the running count of the test within the file. | |
85 | ||
a136f6d8 JK |
86 | -x:: |
87 | Turn on shell tracing (i.e., `set -x`) during the tests | |
94201a2b | 88 | themselves. Implies `--verbose`. |
5fc98e79 SG |
89 | Ignored in test scripts that set the variable 'test_untraceable' |
90 | to a non-empty value, unless it's run with a Bash version | |
91 | supporting BASH_XTRACEFD, i.e. v4.1 or later. | |
a136f6d8 | 92 | |
5e3b4fce | 93 | -d:: |
986aa7f1 JH |
94 | --debug:: |
95 | This may help the person who is developing a new test. | |
96 | It causes the command defined with test_debug to run. | |
0986de94 PK |
97 | The "trash" directory (used to store all temporary data |
98 | during testing) is not deleted even if there are no | |
99 | failed tests so that you can inspect its contents after | |
100 | the test finished. | |
986aa7f1 | 101 | |
5e3b4fce | 102 | -i:: |
986aa7f1 JH |
103 | --immediate:: |
104 | This causes the test to immediately exit upon the first | |
13cb3bb7 SR |
105 | failed test. Cleanup commands requested with |
106 | test_when_finished are not executed if the test failed, | |
107 | in order to keep the state for inspection by the tester | |
108 | to diagnose the bug. | |
986aa7f1 | 109 | |
5e3b4fce | 110 | -l:: |
5e2c08c6 LW |
111 | --long-tests:: |
112 | This causes additional long-running tests to be run (where | |
113 | available), for more exhaustive testing. | |
114 | ||
0445e6f0 IB |
115 | -r:: |
116 | --run=<test-selector>:: | |
117 | Run only the subset of tests indicated by | |
118 | <test-selector>. See section "Skipping Tests" below for | |
119 | <test-selector> syntax. | |
120 | ||
952af351 TR |
121 | --valgrind=<tool>:: |
122 | Execute all Git binaries under valgrind tool <tool> and exit | |
123 | with status 126 on errors (just like regular tests, this will | |
124 | only stop the test script when running under -i). | |
986aa7f1 | 125 | |
3da93652 JS |
126 | Since it makes no sense to run the tests with --valgrind and |
127 | not see any output, this option implies --verbose. For | |
128 | convenience, it also implies --tee. | |
129 | ||
952af351 TR |
130 | <tool> defaults to 'memcheck', just like valgrind itself. |
131 | Other particularly useful choices include 'helgrind' and | |
132 | 'drd', but you may use any tool recognized by your valgrind | |
133 | installation. | |
134 | ||
95d9d5ec TR |
135 | As a special case, <tool> can be 'memcheck-fast', which uses |
136 | memcheck but disables --track-origins. Use this if you are | |
137 | running tests in bulk, to see if there are _any_ memory | |
138 | issues. | |
139 | ||
952af351 | 140 | Note that memcheck is run with the option --leak-check=no, |
9aec68d3 CMN |
141 | as the git process is short-lived and some errors are not |
142 | interesting. In order to run a single command under the same | |
143 | conditions manually, you should set GIT_VALGRIND to point to | |
144 | the 't/valgrind/' directory and use the commands under | |
145 | 't/valgrind/bin/'. | |
146 | ||
5dfc368f TR |
147 | --valgrind-only=<pattern>:: |
148 | Like --valgrind, but the effect is limited to tests with | |
149 | numbers matching <pattern>. The number matched against is | |
150 | simply the running count of the test within the file. | |
151 | ||
44138559 JS |
152 | --tee:: |
153 | In addition to printing the test output to the terminal, | |
154 | write it to files named 't/test-results/$TEST_NAME.out'. | |
155 | As the names depend on the tests' file names, it is safe to | |
156 | run the tests with this option in parallel. | |
157 | ||
a5f52c6d | 158 | -V:: |
452320f1 JK |
159 | --verbose-log:: |
160 | Write verbose output to the same logfile as `--tee`, but do | |
161 | _not_ write it to stdout. Unlike `--tee --verbose`, this option | |
162 | is safe to use when stdout is being consumed by a TAP parser | |
163 | like `prove`. Implies `--tee` and `--verbose`. | |
164 | ||
e4597aae MO |
165 | --with-dashes:: |
166 | By default tests are run without dashed forms of | |
167 | commands (like git-commit) in the PATH (it only uses | |
168 | wrappers from ../bin-wrappers). Use this option to include | |
169 | the build directory (..) in the PATH, which contains all | |
170 | the dashed forms of commands. This option is currently | |
171 | implied by other options like --valgrind and | |
172 | GIT_TEST_INSTALLED. | |
173 | ||
dd167a30 JS |
174 | --no-bin-wrappers:: |
175 | By default, the test suite uses the wrappers in | |
176 | `../bin-wrappers/` to execute `git` and friends. With this option, | |
177 | `../git` and friends are run directly. This is not recommended | |
178 | in general, as the wrappers contain safeguards to ensure that no | |
179 | files from an installed Git are used, but can speed up test runs | |
180 | especially on platforms where running shell scripts is expensive | |
181 | (most notably, Windows). | |
182 | ||
0d4dbcd3 TR |
183 | --root=<directory>:: |
184 | Create "trash" directories used to store all temporary data during | |
185 | testing under <directory>, instead of the t/ directory. | |
186 | Using this option with a RAM-based filesystem (such as tmpfs) | |
187 | can massively speed up the test suite. | |
188 | ||
bb79af9d JK |
189 | --chain-lint:: |
190 | --no-chain-lint:: | |
191 | If --chain-lint is enabled, the test harness will check each | |
192 | test to make sure that it properly "&&-chains" all commands (so | |
193 | that a failure in the middle does not go unnoticed by the final | |
194 | exit code of the test). This check is performed in addition to | |
195 | running the tests themselves. You may also enable or disable | |
196 | this feature by setting the GIT_TEST_CHAIN_LINT environment | |
197 | variable to "1" or "0", respectively. | |
198 | ||
fb7d1e3a | 199 | --stress:: |
fb7d1e3a SG |
200 | Run the test script repeatedly in multiple parallel jobs until |
201 | one of them fails. Useful for reproducing rare failures in | |
202 | flaky tests. The number of parallel jobs is, in order of | |
f5457371 | 203 | precedence: the value of the GIT_TEST_STRESS_LOAD |
fb7d1e3a SG |
204 | environment variable, or twice the number of available |
205 | processors (as shown by the 'getconf' utility), or 8. | |
206 | Implies `--verbose -x --immediate` to get the most information | |
207 | about the failure. Note that the verbose output of each test | |
208 | job is saved to 't/test-results/$TEST_NAME.stress-<nr>.out', | |
209 | and only the output of the failed test job is shown on the | |
210 | terminal. The names of the trash directories get a | |
211 | '.stress-<nr>' suffix, and the trash directory of the failed | |
212 | test job is renamed to end with a '.stress-failed' suffix. | |
213 | ||
f5457371 JS |
214 | --stress-jobs=<N>:: |
215 | Override the number of parallel jobs. Implies `--stress`. | |
216 | ||
76e27fbf SG |
217 | --stress-limit=<N>:: |
218 | When combined with --stress run the test script repeatedly | |
219 | this many times in each of the parallel jobs or until one of | |
de69e6f6 | 220 | them fails, whichever comes first. Implies `--stress`. |
76e27fbf | 221 | |
e160da7f MO |
222 | You can also set the GIT_TEST_INSTALLED environment variable to |
223 | the bindir of an existing git installation to test that installation. | |
224 | You still need to have built this git sandbox, from which various | |
225 | test-* support programs, templates, and perl libraries are used. | |
226 | If your installed git is incomplete, it will silently test parts of | |
227 | your built version instead. | |
228 | ||
229 | When using GIT_TEST_INSTALLED, you can also set GIT_TEST_EXEC_PATH to | |
230 | override the location of the dashed-form subcommands (what | |
231 | GIT_EXEC_PATH would be used for during normal operation). | |
232 | GIT_TEST_EXEC_PATH defaults to `$GIT_TEST_INSTALLED/git --exec-path`. | |
233 | ||
234 | ||
fbd458a3 JN |
235 | Skipping Tests |
236 | -------------- | |
237 | ||
238 | In some environments, certain tests have no way of succeeding | |
239 | due to platform limitation, such as lack of 'unzip' program, or | |
240 | filesystem that do not allow arbitrary sequence of non-NUL bytes | |
241 | as pathnames. | |
242 | ||
243 | You should be able to say something like | |
244 | ||
245 | $ GIT_SKIP_TESTS=t9200.8 sh ./t9200-git-cvsexport-commit.sh | |
246 | ||
247 | and even: | |
248 | ||
249 | $ GIT_SKIP_TESTS='t[0-4]??? t91?? t9200.8' make | |
250 | ||
251 | to omit such tests. The value of the environment variable is a | |
252 | SP separated list of patterns that tells which tests to skip, | |
253 | and either can match the "t[0-9]{4}" part to skip the whole | |
254 | test, or t[0-9]{4} followed by ".$number" to say which | |
255 | particular test to skip. | |
256 | ||
0445e6f0 IB |
257 | For an individual test suite --run could be used to specify that |
258 | only some tests should be run or that some tests should be | |
259 | excluded from a run. | |
260 | ||
f21ac368 EN |
261 | The argument for --run, <test-selector>, is a list of description |
262 | substrings or globs or individual test numbers or ranges with an | |
263 | optional negation prefix (of '!') that define what tests in a test | |
264 | suite to include (or exclude, if negated) in the run. A range is two | |
265 | numbers separated with a dash and matches a range of tests with both | |
266 | ends been included. You may omit the first or the second number to | |
267 | mean "from the first test" or "up to the very last test" respectively. | |
268 | ||
269 | The argument to --run is split on commas into separate strings, | |
270 | numbers, and ranges, and picks all tests that match any of the | |
271 | individual selection criteria. If the substring of the description | |
272 | text that you want to match includes a comma, use the glob character | |
273 | '?' instead. For example --run='rebase,merge?cherry-pick' would match | |
274 | on all tests that match either the glob *rebase* or the glob | |
275 | *merge?cherry-pick*. | |
0445e6f0 IB |
276 | |
277 | If --run starts with an unprefixed number or range the initial | |
278 | set of tests to run is empty. If the first item starts with '!' | |
279 | all the tests are added to the initial set. After initial set is | |
280 | determined every test number or range is added or excluded from | |
281 | the set one by one, from left to right. | |
282 | ||
0445e6f0 IB |
283 | For example, to run only tests up to a specific test (21), one |
284 | could do this: | |
285 | ||
286 | $ sh ./t9200-git-cvsexport-commit.sh --run='1-21' | |
287 | ||
288 | or this: | |
289 | ||
290 | $ sh ./t9200-git-cvsexport-commit.sh --run='-21' | |
291 | ||
292 | Common case is to run several setup tests (1, 2, 3) and then a | |
293 | specific test (21) that relies on that setup: | |
294 | ||
f21ac368 | 295 | $ sh ./t9200-git-cvsexport-commit.sh --run='1,2,3,21' |
0445e6f0 IB |
296 | |
297 | or: | |
298 | ||
299 | $ sh ./t9200-git-cvsexport-commit.sh --run=1,2,3,21 | |
300 | ||
301 | or: | |
302 | ||
f21ac368 | 303 | $ sh ./t9200-git-cvsexport-commit.sh --run='-3,21' |
0445e6f0 | 304 | |
01e4be6c KS |
305 | As noted above, the test set is built by going through the items |
306 | from left to right, so this: | |
0445e6f0 | 307 | |
f21ac368 | 308 | $ sh ./t9200-git-cvsexport-commit.sh --run='1-4,!3' |
0445e6f0 | 309 | |
01e4be6c | 310 | will run tests 1, 2, and 4. Items that come later have higher |
2e3a16b2 | 311 | precedence. It means that this: |
0445e6f0 | 312 | |
f21ac368 | 313 | $ sh ./t9200-git-cvsexport-commit.sh --run='!3,1-4' |
0445e6f0 IB |
314 | |
315 | would just run tests from 1 to 4, including 3. | |
316 | ||
317 | You may use negation with ranges. The following will run all | |
318 | test in the test suite except from 7 up to 11: | |
319 | ||
320 | $ sh ./t9200-git-cvsexport-commit.sh --run='!7-11' | |
321 | ||
f21ac368 EN |
322 | Sometimes there may be multiple tests with e.g. "setup" in their name |
323 | that are needed and rather than figuring out the number for all of them | |
324 | we can just use "setup" as a substring/glob to match against the test | |
325 | description: | |
326 | ||
327 | $ sh ./t0050-filesystem.sh --run=setup,9-11 | |
328 | ||
329 | or one could select both the setup tests and the rename ones (assuming all | |
330 | relevant tests had those words in their descriptions): | |
331 | ||
332 | $ sh ./t0050-filesystem.sh --run=setup,rename | |
333 | ||
0445e6f0 IB |
334 | Some tests in a test suite rely on the previous tests performing |
335 | certain actions, specifically some tests are designated as | |
336 | "setup" test, so you cannot _arbitrarily_ disable one test and | |
337 | expect the rest to function correctly. | |
338 | ||
339 | --run is mostly useful when you want to focus on a specific test | |
340 | and know what setup is needed for it. Or when you want to run | |
341 | everything up to a certain test. | |
fbd458a3 JN |
342 | |
343 | ||
4c2db938 NTND |
344 | Running tests with special setups |
345 | --------------------------------- | |
346 | ||
347 | The whole test suite could be run to test some special features | |
348 | that cannot be easily covered by a few specific test cases. These | |
349 | could be enabled by running the test suite with correct GIT_TEST_ | |
350 | environment set. | |
351 | ||
c7400399 | 352 | GIT_TEST_FAIL_PREREQS=<boolean> fails all prerequisites. This is |
dfe1a17d ÆAB |
353 | useful for discovering issues with the tests where say a later test |
354 | implicitly depends on an optional earlier test. | |
355 | ||
356 | There's a "FAIL_PREREQS" prerequisite that can be used to test for | |
357 | whether this mode is active, and e.g. skip some tests that are hard to | |
358 | refactor to deal with it. The "SYMLINKS" prerequisite is currently | |
359 | excluded as so much relies on it, but this might change in the future. | |
360 | ||
4c2db938 NTND |
361 | GIT_TEST_SPLIT_INDEX=<boolean> forces split-index mode on the whole |
362 | test suite. Accept any boolean values that are accepted by git-config. | |
363 | ||
33166f3a JN |
364 | GIT_TEST_PROTOCOL_VERSION=<n>, when set, makes 'protocol.version' |
365 | default to n. | |
8cbeba06 | 366 | |
43fa44fa NTND |
367 | GIT_TEST_FULL_IN_PACK_ARRAY=<boolean> exercises the uncommon |
368 | pack-objects code path where there are more than 1024 packs even if | |
369 | the actual number of packs in repository is below this limit. Accept | |
370 | any boolean values that are accepted by git-config. | |
371 | ||
ac77d0c3 NTND |
372 | GIT_TEST_OE_SIZE=<n> exercises the uncommon pack-objects code path |
373 | where we do not cache object size in memory and read it from existing | |
374 | packs on demand. This normally only happens when the object size is | |
375 | over 2GB. This variable forces the code path on any object larger than | |
376 | <n> bytes. | |
377 | ||
ac6e12f9 | 378 | GIT_TEST_OE_DELTA_SIZE=<n> exercises the uncommon pack-objects code |
9ac3f0e5 NTND |
379 | path where deltas larger than this limit require extra memory |
380 | allocation for bookkeeping. | |
381 | ||
5f4436a7 NTND |
382 | GIT_TEST_VALIDATE_INDEX_CACHE_ENTRIES=<boolean> checks that cache-tree |
383 | records are valid when the index is written out or after a merge. This | |
384 | is mostly to catch missing invalidation. Default is true. | |
385 | ||
859fdc0c DS |
386 | GIT_TEST_COMMIT_GRAPH=<boolean>, when true, forces the commit-graph to |
387 | be written after every 'git commit' command, and overrides the | |
388 | 'core.commitGraph' setting to true. | |
389 | ||
d5b873c8 GS |
390 | GIT_TEST_COMMIT_GRAPH_CHANGED_PATHS=<boolean>, when true, forces |
391 | commit-graph write to compute and write changed path Bloom filters for | |
392 | every 'git commit-graph write', as if the `--changed-paths` option was | |
393 | passed in. | |
394 | ||
4cb54d0a BP |
395 | GIT_TEST_FSMONITOR=$PWD/t7519/fsmonitor-all exercises the fsmonitor |
396 | code path for utilizing a file system monitor to speed up detecting | |
397 | new or changed files. | |
398 | ||
1f357b04 BP |
399 | GIT_TEST_INDEX_VERSION=<n> exercises the index read/write code path |
400 | for the index version specified. Can be set to any valid version | |
401 | (currently 2, 3, or 4). | |
402 | ||
2d657ab9 DS |
403 | GIT_TEST_PACK_SPARSE=<boolean> if disabled will default the pack-objects |
404 | builtin to use the non-sparse object walk. This can still be overridden by | |
405 | the --sparse command-line argument. | |
99dbbfa8 | 406 | |
5765d97b BP |
407 | GIT_TEST_PRELOAD_INDEX=<boolean> exercises the preload-index code path |
408 | by overriding the minimum number of cache entries required per thread. | |
409 | ||
f83dff60 JS |
410 | GIT_TEST_ADD_I_USE_BUILTIN=<boolean>, when true, enables the |
411 | built-in version of git add -i. See 'add.interactive.useBuiltin' in | |
412 | git-config(1). | |
413 | ||
c780b9cf BP |
414 | GIT_TEST_INDEX_THREADS=<n> enables exercising the multi-threaded loading |
415 | of the index for the whole test suite by bypassing the default number of | |
416 | cache entries and thread minimums. Setting this to 1 will make the | |
417 | index loading single threaded. | |
418 | ||
0465a505 DS |
419 | GIT_TEST_MULTI_PACK_INDEX=<boolean>, when true, forces the multi-pack- |
420 | index to be written after every 'git repack' command, and overrides the | |
421 | 'core.multiPackIndex' setting to true. | |
422 | ||
07c3c2aa JT |
423 | GIT_TEST_SIDEBAND_ALL=<boolean>, when true, overrides the |
424 | 'uploadpack.allowSidebandAll' setting to true, and when false, forces | |
425 | fetch-pack to not request sideband-all (even if the server advertises | |
426 | sideband-all). | |
427 | ||
b02e7d5d JS |
428 | GIT_TEST_DISALLOW_ABBREVIATED_OPTIONS=<boolean>, when true (which is |
429 | the default when running tests), errors out when an abbreviated option | |
430 | is used. | |
431 | ||
f3d66ec8 DS |
432 | GIT_TEST_DEFAULT_HASH=<hash-algo> specifies which hash algorithm to |
433 | use in the test scripts. Recognized values for <hash-algo> are "sha1" | |
434 | and "sha256". | |
435 | ||
e8c58f89 TB |
436 | GIT_TEST_WRITE_REV_INDEX=<boolean>, when true enables the |
437 | 'pack.writeReverseIndex' setting. | |
438 | ||
f50c9f76 PB |
439 | Naming Tests |
440 | ------------ | |
441 | ||
442 | The test files are named as: | |
443 | ||
444 | tNNNN-commandname-details.sh | |
445 | ||
446 | where N is a decimal digit. | |
447 | ||
448 | First digit tells the family: | |
449 | ||
450 | 0 - the absolute basics and global stuff | |
451 | 1 - the basic commands concerning database | |
452 | 2 - the basic commands concerning the working tree | |
453 | 3 - the other basic commands (e.g. ls-files) | |
454 | 4 - the diff commands | |
455 | 5 - the pull and exporting commands | |
456 | 6 - the revision tree commands (even e.g. merge-base) | |
8f4a9b62 | 457 | 7 - the porcelainish commands concerning the working tree |
8757749e JN |
458 | 8 - the porcelainish commands concerning forensics |
459 | 9 - the git tools | |
f50c9f76 PB |
460 | |
461 | Second digit tells the particular command we are testing. | |
462 | ||
463 | Third digit (optionally) tells the particular switch or group of switches | |
464 | we are testing. | |
465 | ||
77656600 JH |
466 | If you create files under t/ directory (i.e. here) that is not |
467 | the top-level test script, never name the file to match the above | |
468 | pattern. The Makefile here considers all such files as the | |
63d32945 | 469 | top-level test script and tries to run all of them. Care is |
77656600 JH |
470 | especially needed if you are creating a common test library |
471 | file, similar to test-lib.sh, because such a library file may | |
472 | not be suitable for standalone execution. | |
473 | ||
f50c9f76 | 474 | |
986aa7f1 JH |
475 | Writing Tests |
476 | ------------- | |
477 | ||
478 | The test script is written as a shell script. It should start | |
51b7a525 | 479 | with the standard "#!/bin/sh", and an |
986aa7f1 JH |
480 | assignment to variable 'test_description', like this: |
481 | ||
482 | #!/bin/sh | |
986aa7f1 | 483 | |
14cd1ff3 | 484 | test_description='xxx test (option --frotz) |
986aa7f1 JH |
485 | |
486 | This test registers the following structure in the cache | |
487 | and tries to run git-ls-files with option --frotz.' | |
488 | ||
f50c9f76 | 489 | |
986aa7f1 JH |
490 | Source 'test-lib.sh' |
491 | -------------------- | |
492 | ||
493 | After assigning test_description, the test script should source | |
494 | test-lib.sh like this: | |
495 | ||
496 | . ./test-lib.sh | |
497 | ||
498 | This test harness library does the following things: | |
499 | ||
500 | - If the script is invoked with command line argument --help | |
501 | (or -h), it shows the test_description and exits. | |
502 | ||
e1ca1c9d ÆAB |
503 | - Creates an empty test directory with an empty .git/objects database |
504 | and chdir(2) into it. This directory is 't/trash | |
505 | directory.$test_name_without_dotsh', with t/ subject to change by | |
fb7d1e3a SG |
506 | the --root option documented above, and a '.stress-<N>' suffix |
507 | appended by the --stress option. | |
986aa7f1 JH |
508 | |
509 | - Defines standard test helper functions for your scripts to | |
510 | use. These functions are designed to make all scripts behave | |
511 | consistently when command line arguments --verbose (or -v), | |
512 | --debug (or -d), and --immediate (or -i) is given. | |
513 | ||
441ee35d MD |
514 | Do's & don'ts |
515 | ------------- | |
20873f45 | 516 | |
6fd45295 | 517 | Here are a few examples of things you probably should and shouldn't do |
20873f45 ÆAB |
518 | when writing tests. |
519 | ||
441ee35d | 520 | Here are the "do's:" |
20873f45 | 521 | |
6fd45295 | 522 | - Put all code inside test_expect_success and other assertions. |
20873f45 ÆAB |
523 | |
524 | Even code that isn't a test per se, but merely some setup code | |
6fd45295 | 525 | should be inside a test assertion. |
20873f45 ÆAB |
526 | |
527 | - Chain your test assertions | |
528 | ||
529 | Write test code like this: | |
530 | ||
531 | git merge foo && | |
532 | git push bar && | |
533 | test ... | |
534 | ||
535 | Instead of: | |
536 | ||
537 | git merge hla | |
538 | git push gh | |
539 | test ... | |
540 | ||
541 | That way all of the commands in your tests will succeed or fail. If | |
00648ba0 EN |
542 | you must ignore the return value of something, consider using a |
543 | helper function (e.g. use sane_unset instead of unset, in order | |
544 | to avoid unportable return value for unsetting a variable that was | |
545 | already unset), or prepending the command with test_might_fail or | |
546 | test_must_fail. | |
20873f45 | 547 | |
0c357544 ÆAB |
548 | - Check the test coverage for your tests. See the "Test coverage" |
549 | below. | |
550 | ||
63d32945 MW |
551 | Don't blindly follow test coverage metrics; if a new function you added |
552 | doesn't have any coverage, then you're probably doing something wrong, | |
e8b55f5c ÆAB |
553 | but having 100% coverage doesn't necessarily mean that you tested |
554 | everything. | |
555 | ||
556 | Tests that are likely to smoke out future regressions are better | |
557 | than tests that just inflate the coverage metrics. | |
558 | ||
95b104c8 JS |
559 | - When a test checks for an absolute path that a git command generated, |
560 | construct the expected value using $(pwd) rather than $PWD, | |
561 | $TEST_DIRECTORY, or $TRASH_DIRECTORY. It makes a difference on | |
562 | Windows, where the shell (MSYS bash) mangles absolute path names. | |
563 | For details, see the commit message of 4114156ae9. | |
564 | ||
441ee35d MD |
565 | - Remember that inside the <script> part, the standard output and |
566 | standard error streams are discarded, and the test harness only | |
567 | reports "ok" or "not ok" to the end user running the tests. Under | |
568 | --verbose, they are shown to help debug the tests. | |
569 | ||
7cc112dc JH |
570 | - Be careful when you loop |
571 | ||
572 | You may need to verify multiple things in a loop, but the | |
573 | following does not work correctly: | |
574 | ||
575 | test_expect_success 'test three things' ' | |
576 | for i in one two three | |
577 | do | |
578 | test_something "$i" | |
579 | done && | |
580 | test_something_else | |
581 | ' | |
582 | ||
583 | Because the status of the loop itself is the exit status of the | |
584 | test_something in the last round, the loop does not fail when | |
585 | "test_something" for "one" or "two" fails. This is not what you | |
586 | want. | |
587 | ||
588 | Instead, you can break out of the loop immediately when you see a | |
589 | failure. Because all test_expect_* snippets are executed inside | |
590 | a function, "return 1" can be used to fail the test immediately | |
591 | upon a failure: | |
592 | ||
593 | test_expect_success 'test three things' ' | |
594 | for i in one two three | |
595 | do | |
596 | test_something "$i" || return 1 | |
597 | done && | |
598 | test_something_else | |
599 | ' | |
600 | ||
601 | Note that we still &&-chain the loop to propagate failures from | |
602 | earlier commands. | |
603 | ||
604 | ||
441ee35d | 605 | And here are the "don'ts:" |
20873f45 | 606 | |
441ee35d | 607 | - Don't exit() within a <script> part. |
20873f45 ÆAB |
608 | |
609 | The harness will catch this as a programming error of the test. | |
610 | Use test_done instead if you need to stop the tests early (see | |
611 | "Skipping tests" below). | |
612 | ||
441ee35d MD |
613 | - Don't use '! git cmd' when you want to make sure the git command |
614 | exits with failure in a controlled way by calling "die()". Instead, | |
ad78585e JH |
615 | use 'test_must_fail git cmd'. This will signal a failure if git |
616 | dies in an unexpected way (e.g. segfault). | |
617 | ||
f445500e | 618 | On the other hand, don't use test_must_fail for running regular |
53de7424 JH |
619 | platform commands; just use '! cmd'. We are not in the business |
620 | of verifying that the world given to us sanely works. | |
f445500e | 621 | |
a378fee5 MD |
622 | - Don't feed the output of a git command to a pipe, as in: |
623 | ||
624 | git -C repo ls-files | | |
625 | xargs -n 1 basename | | |
626 | grep foo | |
627 | ||
628 | which will discard git's exit code and may mask a crash. In the | |
629 | above example, all exit codes are ignored except grep's. | |
630 | ||
631 | Instead, write the output of that command to a temporary | |
632 | file with ">" or assign it to a variable with "x=$(git ...)" rather | |
633 | than pipe it. | |
634 | ||
635 | - Don't use command substitution in a way that discards git's exit | |
636 | code. When assigning to a variable, the exit code is not discarded, | |
637 | e.g.: | |
638 | ||
639 | x=$(git cat-file -p $sha) && | |
640 | ... | |
641 | ||
642 | is OK because a crash in "git cat-file" will cause the "&&" chain | |
643 | to fail, but: | |
644 | ||
645 | test "refs/heads/foo" = "$(git symbolic-ref HEAD)" | |
646 | ||
647 | is not OK and a crash in git could go undetected. | |
648 | ||
441ee35d MD |
649 | - Don't use perl without spelling it as "$PERL_PATH". This is to help |
650 | our friends on Windows where the platform Perl often adds CR before | |
ad78585e | 651 | the end of line, and they bundle Git with a version of Perl that |
a0e0ec9f JK |
652 | does not do so, whose path is specified with $PERL_PATH. Note that we |
653 | provide a "perl" function which uses $PERL_PATH under the hood, so | |
654 | you do not need to worry when simply running perl in the test scripts | |
655 | (but you do, for example, on a shebang line or in a sub script | |
656 | created via "write_script"). | |
ad78585e | 657 | |
441ee35d MD |
658 | - Don't use sh without spelling it as "$SHELL_PATH", when the script |
659 | can be misinterpreted by broken platform shell (e.g. Solaris). | |
ad78585e | 660 | |
441ee35d | 661 | - Don't chdir around in tests. It is not sufficient to chdir to |
ad78585e JH |
662 | somewhere and then chdir back to the original location later in |
663 | the test, as any intermediate step can fail and abort the test, | |
664 | causing the next test to start in an unexpected directory. Do so | |
665 | inside a subshell if necessary. | |
666 | ||
441ee35d MD |
667 | - Don't save and verify the standard error of compound commands, i.e. |
668 | group commands, subshells, and shell functions (except test helper | |
94201a2b SG |
669 | functions like 'test_must_fail') like this: |
670 | ||
671 | ( cd dir && git cmd ) 2>error && | |
672 | test_cmp expect error | |
673 | ||
674 | When running the test with '-x' tracing, then the trace of commands | |
675 | executed in the compound command will be included in standard error | |
676 | as well, quite possibly throwing off the subsequent checks examining | |
677 | the output. Instead, save only the relevant git command's standard | |
678 | error: | |
679 | ||
680 | ( cd dir && git cmd 2>../error ) && | |
681 | test_cmp expect error | |
682 | ||
441ee35d | 683 | - Don't break the TAP output |
20873f45 | 684 | |
6fd45295 JH |
685 | The raw output from your test may be interpreted by a TAP harness. TAP |
686 | harnesses will ignore everything they don't know about, but don't step | |
687 | on their toes in these areas: | |
20873f45 ÆAB |
688 | |
689 | - Don't print lines like "$x..$y" where $x and $y are integers. | |
690 | ||
691 | - Don't print lines that begin with "ok" or "not ok". | |
692 | ||
6fd45295 | 693 | TAP harnesses expect a line that begins with either "ok" and "not |
20873f45 ÆAB |
694 | ok" to signal a test passed or failed (and our harness already |
695 | produces such lines), so your script shouldn't emit such lines to | |
696 | their output. | |
697 | ||
698 | You can glean some further possible issues from the TAP grammar | |
c1d44cee | 699 | (see https://metacpan.org/pod/TAP::Parser::Grammar#TAP-GRAMMAR) |
20873f45 ÆAB |
700 | but the best indication is to just run the tests with prove(1), |
701 | it'll complain if anything is amiss. | |
702 | ||
20873f45 | 703 | |
b5500d16 ÆAB |
704 | Skipping tests |
705 | -------------- | |
706 | ||
681186ae | 707 | If you need to skip tests you should do so by using the three-arg form |
99d9050d ÆAB |
708 | of the test_* functions (see the "Test harness library" section |
709 | below), e.g.: | |
710 | ||
ad78585e | 711 | test_expect_success PERL 'I need Perl' ' |
a0e0ec9f | 712 | perl -e "hlagh() if unf_unf()" |
ad78585e | 713 | ' |
99d9050d ÆAB |
714 | |
715 | The advantage of skipping tests like this is that platforms that don't | |
716 | have the PERL and other optional dependencies get an indication of how | |
717 | many tests they're missing. | |
718 | ||
719 | If the test code is too hairy for that (i.e. does a lot of setup work | |
720 | outside test assertions) you can also skip all remaining tests by | |
721 | setting skip_all and immediately call test_done: | |
b5500d16 ÆAB |
722 | |
723 | if ! test_have_prereq PERL | |
724 | then | |
725 | skip_all='skipping perl interface tests, perl not available' | |
726 | test_done | |
727 | fi | |
14cd1ff3 | 728 | |
99d9050d ÆAB |
729 | The string you give to skip_all will be used as an explanation for why |
730 | the test was skipped. | |
731 | ||
986aa7f1 JH |
732 | End with test_done |
733 | ------------------ | |
734 | ||
735 | Your script will be a sequence of tests, using helper functions | |
736 | from the test harness library. At the end of the script, call | |
737 | 'test_done'. | |
738 | ||
739 | ||
740 | Test harness library | |
741 | -------------------- | |
742 | ||
743 | There are a handful helper functions defined in the test harness | |
744 | library for your script to use. | |
745 | ||
9a897893 | 746 | - test_expect_success [<prereq>] <message> <script> |
986aa7f1 | 747 | |
72942a61 | 748 | Usually takes two strings as parameters, and evaluates the |
986aa7f1 JH |
749 | <script>. If it yields success, test is considered |
750 | successful. <message> should state what it is testing. | |
751 | ||
752 | Example: | |
753 | ||
754 | test_expect_success \ | |
755 | 'git-write-tree should be able to write an empty tree.' \ | |
756 | 'tree=$(git-write-tree)' | |
757 | ||
9a897893 | 758 | If you supply three parameters the first will be taken to be a |
72942a61 | 759 | prerequisite; see the test_set_prereq and test_have_prereq |
9a897893 ÆAB |
760 | documentation below: |
761 | ||
762 | test_expect_success TTY 'git --paginate rev-list uses a pager' \ | |
763 | ' ... ' | |
764 | ||
93a57246 ÆAB |
765 | You can also supply a comma-separated list of prerequisites, in the |
766 | rare case where your test depends on more than one: | |
767 | ||
768 | test_expect_success PERL,PYTHON 'yo dawg' \ | |
769 | ' test $(perl -E 'print eval "1 +" . qx[python -c "print 2"]') == "4" ' | |
770 | ||
9a897893 | 771 | - test_expect_failure [<prereq>] <message> <script> |
986aa7f1 | 772 | |
41ac414e JH |
773 | This is NOT the opposite of test_expect_success, but is used |
774 | to mark a test that demonstrates a known breakage. Unlike | |
775 | the usual test_expect_success tests, which say "ok" on | |
776 | success and "FAIL" on failure, this will say "FIXED" on | |
777 | success and "still broken" on failure. Failures from these | |
778 | tests won't cause -i (immediate) to stop. | |
986aa7f1 | 779 | |
9a897893 ÆAB |
780 | Like test_expect_success this function can optionally use a three |
781 | argument invocation with a prerequisite as the first argument. | |
782 | ||
986aa7f1 JH |
783 | - test_debug <script> |
784 | ||
785 | This takes a single argument, <script>, and evaluates it only | |
786 | when the test script is started with --debug command line | |
787 | argument. This is primarily meant for use during the | |
788 | development of a new test script. | |
789 | ||
6a94088c JS |
790 | - debug <git-command> |
791 | ||
792 | Run a git command inside a debugger. This is primarily meant for | |
793 | use when debugging a failing test script. | |
794 | ||
986aa7f1 JH |
795 | - test_done |
796 | ||
797 | Your test script must have test_done at the end. Its purpose | |
798 | is to summarize successes and failures in the test script and | |
799 | exit with an appropriate error code. | |
800 | ||
00884968 JS |
801 | - test_tick |
802 | ||
803 | Make commit and tag names consistent by setting the author and | |
63d32945 | 804 | committer times to defined state. Subsequent calls will |
00884968 JS |
805 | advance the times by a fixed amount. |
806 | ||
807 | - test_commit <message> [<filename> [<contents>]] | |
808 | ||
809 | Creates a commit with the given message, committing the given | |
810 | file with the given contents (default for both is to reuse the | |
811 | message string), and adds a tag (again reusing the message | |
812 | string as name). Calls test_tick to make the SHA-1s | |
813 | reproducible. | |
814 | ||
815 | - test_merge <message> <commit-or-tag> | |
816 | ||
817 | Merges the given rev using the given message. Like test_commit, | |
818 | creates a tag and calls test_tick before committing. | |
986aa7f1 | 819 | |
72942a61 | 820 | - test_set_prereq <prereq> |
9a897893 ÆAB |
821 | |
822 | Set a test prerequisite to be used later with test_have_prereq. The | |
be53deef ÆAB |
823 | test-lib will set some prerequisites for you, see the |
824 | "Prerequisites" section below for a full list of these. | |
825 | ||
826 | Others you can set yourself and use later with either | |
827 | test_have_prereq directly, or the three argument invocation of | |
828 | test_expect_success and test_expect_failure. | |
9a897893 | 829 | |
72942a61 | 830 | - test_have_prereq <prereq> |
9a897893 | 831 | |
4473060b ÆAB |
832 | Check if we have a prerequisite previously set with test_set_prereq. |
833 | The most common way to use this explicitly (as opposed to the | |
834 | implicit use when an argument is passed to test_expect_*) is to skip | |
835 | all the tests at the start of the test script if we don't have some | |
836 | essential prerequisite: | |
9a897893 ÆAB |
837 | |
838 | if ! test_have_prereq PERL | |
839 | then | |
840 | skip_all='skipping perl interface tests, perl not available' | |
841 | test_done | |
842 | fi | |
843 | ||
2fac6a4b ÆAB |
844 | - test_external [<prereq>] <message> <external> <script> |
845 | ||
846 | Execute a <script> with an <external> interpreter (like perl). This | |
847 | was added for tests like t9700-perl-git.sh which do most of their | |
848 | work in an external test script. | |
849 | ||
850 | test_external \ | |
851 | 'GitwebCache::*FileCache*' \ | |
a0e0ec9f | 852 | perl "$TEST_DIRECTORY"/t9503/test_cache_interface.pl |
2fac6a4b ÆAB |
853 | |
854 | If the test is outputting its own TAP you should set the | |
855 | test_external_has_tap variable somewhere before calling the first | |
856 | test_external* function. See t9700-perl-git.sh for an example. | |
857 | ||
858 | # The external test will outputs its own plan | |
859 | test_external_has_tap=1 | |
860 | ||
861 | - test_external_without_stderr [<prereq>] <message> <external> <script> | |
862 | ||
863 | Like test_external but fail if there's any output on stderr, | |
864 | instead of checking the exit code. | |
865 | ||
866 | test_external_without_stderr \ | |
867 | 'Perl API' \ | |
a0e0ec9f | 868 | perl "$TEST_DIRECTORY"/t9700/test.pl |
2fac6a4b | 869 | |
892e6f7e ÆAB |
870 | - test_expect_code <exit-code> <command> |
871 | ||
872 | Run a command and ensure that it exits with the given exit code. | |
873 | For example: | |
874 | ||
875 | test_expect_success 'Merge with d/f conflicts' ' | |
876 | test_expect_code 1 git merge "merge msg" B master | |
877 | ' | |
878 | ||
12e31a6b | 879 | - test_must_fail [<options>] <git-command> |
c9667456 JN |
880 | |
881 | Run a git command and ensure it fails in a controlled way. Use | |
971ecbd1 BC |
882 | this instead of "! <git-command>". When git-command dies due to a |
883 | segfault, test_must_fail diagnoses it as an error; "! <git-command>" | |
884 | treats it as just another expected failure, which would let such a | |
885 | bug go unnoticed. | |
c9667456 | 886 | |
12e31a6b SG |
887 | Accepts the following options: |
888 | ||
889 | ok=<signal-name>[,<...>]: | |
890 | Don't treat an exit caused by the given signal as error. | |
891 | Multiple signals can be specified as a comma separated list. | |
892 | Currently recognized signal names are: sigpipe, success. | |
893 | (Don't use 'success', use 'test_might_fail' instead.) | |
894 | ||
895 | - test_might_fail [<options>] <git-command> | |
c9667456 JN |
896 | |
897 | Similar to test_must_fail, but tolerate success, too. Use this | |
898 | instead of "<git-command> || :" to catch failures due to segv. | |
899 | ||
12e31a6b SG |
900 | Accepts the same options as test_must_fail. |
901 | ||
c9667456 JN |
902 | - test_cmp <expected> <actual> |
903 | ||
904 | Check whether the content of the <actual> file matches the | |
905 | <expected> file. This behaves like "cmp" but produces more | |
906 | helpful output when the test is run with "-v" option. | |
907 | ||
5a052626 TG |
908 | - test_cmp_rev <expected> <actual> |
909 | ||
910 | Check whether the <expected> rev points to the same commit as the | |
911 | <actual> rev. | |
912 | ||
fb3340a6 JN |
913 | - test_line_count (= | -lt | -ge | ...) <length> <file> |
914 | ||
915 | Check whether a file has the length it is expected to. | |
916 | ||
72942a61 ML |
917 | - test_path_is_file <path> [<diagnosis>] |
918 | test_path_is_dir <path> [<diagnosis>] | |
2caf20c5 MM |
919 | test_path_is_missing <path> [<diagnosis>] |
920 | ||
72942a61 ML |
921 | Check if the named path is a file, if the named path is a |
922 | directory, or if the named path does not exist, respectively, | |
923 | and fail otherwise, showing the <diagnosis> text. | |
2caf20c5 | 924 | |
c9667456 JN |
925 | - test_when_finished <script> |
926 | ||
927 | Prepend <script> to a list of commands to run to clean up | |
928 | at the end of the current test. If some clean-up command | |
929 | fails, the test will not pass. | |
930 | ||
931 | Example: | |
932 | ||
933 | test_expect_success 'branch pointing to non-commit' ' | |
934 | git rev-parse HEAD^{tree} >.git/refs/heads/invalid && | |
935 | test_when_finished "git update-ref -d refs/heads/invalid" && | |
936 | ... | |
937 | ' | |
938 | ||
900721e1 JS |
939 | - test_atexit <script> |
940 | ||
941 | Prepend <script> to a list of commands to run unconditionally to | |
942 | clean up before the test script exits, e.g. to stop a daemon: | |
943 | ||
944 | test_expect_success 'test git daemon' ' | |
945 | git daemon & | |
946 | daemon_pid=$! && | |
947 | test_atexit 'kill $daemon_pid' && | |
948 | hello world | |
949 | ' | |
950 | ||
951 | The commands will be executed before the trash directory is removed, | |
952 | i.e. the atexit commands will still be able to access any pidfiles or | |
953 | socket files. | |
954 | ||
955 | Note that these commands will be run even when a test script run | |
956 | with '--immediate' fails. Be careful with your atexit commands to | |
957 | minimize any changes to the failed state. | |
958 | ||
bb98b01e | 959 | - test_write_lines <lines> |
ac9afcc3 | 960 | |
bb98b01e | 961 | Write <lines> on standard output, one line per argument. |
ac9afcc3 MT |
962 | Useful to prepare multi-line files in a compact form. |
963 | ||
964 | Example: | |
965 | ||
bb98b01e | 966 | test_write_lines a b c d e f g >foo |
ac9afcc3 MT |
967 | |
968 | Is a more compact equivalent of: | |
969 | cat >foo <<-EOF | |
970 | a | |
971 | b | |
972 | c | |
973 | d | |
974 | e | |
975 | f | |
976 | g | |
977 | EOF | |
978 | ||
979 | ||
c4d2539a JL |
980 | - test_pause |
981 | ||
982 | This command is useful for writing and debugging tests and must be | |
983 | removed before submitting. It halts the execution of the test and | |
984 | spawns a shell in the trash directory. Exit the shell to continue | |
985 | the test. Example: | |
986 | ||
987 | test_expect_success 'test' ' | |
988 | git do-something >actual && | |
989 | test_pause && | |
990 | test_cmp expected actual | |
991 | ' | |
992 | ||
9ce415d9 JS |
993 | - test_ln_s_add <path1> <path2> |
994 | ||
995 | This function helps systems whose filesystem does not support symbolic | |
996 | links. Use it to add a symbolic link entry to the index when it is not | |
997 | important that the file system entry is a symbolic link, i.e., instead | |
998 | of the sequence | |
999 | ||
1000 | ln -s foo bar && | |
1001 | git add bar | |
1002 | ||
1003 | Sometimes it is possible to split a test in a part that does not need | |
1004 | the symbolic link in the file system and a part that does; then only | |
1005 | the latter part need be protected by a SYMLINKS prerequisite (see below). | |
1006 | ||
2c02b110 | 1007 | - test_oid_init |
1008 | ||
1009 | This function loads facts and useful object IDs related to the hash | |
1010 | algorithm(s) in use from the files in t/oid-info. | |
1011 | ||
1012 | - test_oid_cache | |
1013 | ||
1014 | This function reads per-hash algorithm information from standard | |
1015 | input (usually a heredoc) in the format described in | |
1016 | t/oid-info/README. This is useful for test-specific values, such as | |
1017 | object IDs, which must vary based on the hash algorithm. | |
1018 | ||
1019 | Certain fixed values, such as hash sizes and common placeholder | |
1020 | object IDs, can be loaded with test_oid_init (described above). | |
1021 | ||
1022 | - test_oid <key> | |
1023 | ||
1024 | This function looks up a value for the hash algorithm in use, based | |
1025 | on the key given. The value must have been loaded using | |
1026 | test_oid_init or test_oid_cache. Providing an unknown key is an | |
1027 | error. | |
1028 | ||
11f470ae JH |
1029 | - yes [<string>] |
1030 | ||
1031 | This is often seen in modern UNIX but some platforms lack it, so | |
1032 | the test harness overrides the platform implementation with a | |
1033 | more limited one. Use this only when feeding a handful lines of | |
1034 | output to the downstream---unlike the real version, it generates | |
1035 | only up to 99 lines. | |
1036 | ||
43a2afee SG |
1037 | - test_bool_env <env-variable-name> <default-value> |
1038 | ||
1039 | Given the name of an environment variable with a bool value, | |
1040 | normalize its value to a 0 (true) or 1 (false or empty string) | |
1041 | return code. Return with code corresponding to the given default | |
1042 | value if the variable is unset. | |
1043 | Abort the test script if either the value of the variable or the | |
1044 | default are not valid bool values. | |
1045 | ||
11f470ae | 1046 | |
be53deef ÆAB |
1047 | Prerequisites |
1048 | ------------- | |
1049 | ||
1050 | These are the prerequisites that the test library predefines with | |
1051 | test_have_prereq. | |
1052 | ||
1053 | See the prereq argument to the test_* functions in the "Test harness | |
1054 | library" section above and the "test_have_prereq" function for how to | |
1055 | use these, and "test_set_prereq" for how to define your own. | |
1056 | ||
f8fc0ee3 | 1057 | - PYTHON |
be53deef | 1058 | |
f8fc0ee3 JN |
1059 | Git wasn't compiled with NO_PYTHON=YesPlease. Wrap any tests that |
1060 | need Python with this. | |
1061 | ||
1062 | - PERL | |
1063 | ||
1064 | Git wasn't compiled with NO_PERL=YesPlease. | |
1065 | ||
1066 | Even without the PERL prerequisite, tests can assume there is a | |
1067 | usable perl interpreter at $PERL_PATH, though it need not be | |
1068 | particularly modern. | |
be53deef ÆAB |
1069 | |
1070 | - POSIXPERM | |
1071 | ||
1072 | The filesystem supports POSIX style permission bits. | |
1073 | ||
1074 | - BSLASHPSPEC | |
1075 | ||
1076 | Backslashes in pathspec are not directory separators. This is not | |
1077 | set on Windows. See 6fd1106a for details. | |
1078 | ||
1079 | - EXECKEEPSPID | |
1080 | ||
1081 | The process retains the same pid across exec(2). See fb9a2bea for | |
1082 | details. | |
1083 | ||
20073274 AS |
1084 | - PIPE |
1085 | ||
1086 | The filesystem we're on supports creation of FIFOs (named pipes) | |
1087 | via mkfifo(1). | |
1088 | ||
be53deef ÆAB |
1089 | - SYMLINKS |
1090 | ||
1091 | The filesystem we're on supports symbolic links. E.g. a FAT | |
1092 | filesystem doesn't support these. See 704a3143 for details. | |
2fac6a4b | 1093 | |
c91cfd19 ÆAB |
1094 | - SANITY |
1095 | ||
1096 | Test is not run by root user, and an attempt to write to an | |
1097 | unwritable file is expected to fail correctly. | |
2fac6a4b | 1098 | |
3eb585c1 | 1099 | - PCRE |
8f852ce6 | 1100 | |
3eb585c1 | 1101 | Git was compiled with support for PCRE. Wrap any tests |
8f852ce6 MK |
1102 | that use git-grep --perl-regexp or git-grep -P in these. |
1103 | ||
ac39aa61 MG |
1104 | - CASE_INSENSITIVE_FS |
1105 | ||
1106 | Test is run on a case insensitive file system. | |
1107 | ||
5b0b5dd4 MG |
1108 | - UTF8_NFD_TO_NFC |
1109 | ||
1110 | Test is run on a filesystem which converts decomposed utf-8 (nfd) | |
1111 | to precomposed utf-8 (nfc). | |
1112 | ||
68c7d276 ÆAB |
1113 | - PTHREADS |
1114 | ||
1115 | Git wasn't compiled with NO_PTHREADS=YesPlease. | |
1116 | ||
986aa7f1 JH |
1117 | Tips for Writing Tests |
1118 | ---------------------- | |
1119 | ||
1120 | As with any programming projects, existing programs are the best | |
1121 | source of the information. However, do _not_ emulate | |
1122 | t0000-basic.sh when writing your tests. The test is special in | |
788db145 | 1123 | that it tries to validate the very core of Git. For example, it |
986aa7f1 JH |
1124 | knows that there will be 256 subdirectories under .git/objects/, |
1125 | and it knows that the object ID of an empty tree is a certain | |
1126 | 40-byte string. This is deliberately done so in t0000-basic.sh | |
1127 | because the things the very basic core test tries to achieve is | |
788db145 | 1128 | to serve as a basis for people who are changing the Git internals |
986aa7f1 JH |
1129 | drastically. For these people, after making certain changes, |
1130 | not seeing failures from the basic test _is_ a failure. And | |
788db145 | 1131 | such drastic changes to the core Git that even changes these |
986aa7f1 JH |
1132 | otherwise supposedly stable object IDs should be accompanied by |
1133 | an update to t0000-basic.sh. | |
1134 | ||
1135 | However, other tests that simply rely on basic parts of the core | |
788db145 DL |
1136 | Git working properly should not have that level of intimate |
1137 | knowledge of the core Git internals. If all the test scripts | |
986aa7f1 JH |
1138 | hardcoded the object IDs like t0000-basic.sh does, that defeats |
1139 | the purpose of t0000-basic.sh, which is to isolate that level of | |
1140 | validation in one place. Your test also ends up needing | |
1141 | updating when such a change to the internal happens, so do _not_ | |
1142 | do it and leave the low level of validation to t0000-basic.sh. | |
d15e9ebc | 1143 | |
0c357544 ÆAB |
1144 | Test coverage |
1145 | ------------- | |
1146 | ||
1147 | You can use the coverage tests to find code paths that are not being | |
1148 | used or properly exercised yet. | |
1149 | ||
1150 | To do that, run the coverage target at the top-level (not in the t/ | |
1151 | directory): | |
1152 | ||
1153 | make coverage | |
1154 | ||
1155 | That'll compile Git with GCC's coverage arguments, and generate a test | |
1156 | report with gcov after the tests finish. Running the coverage tests | |
1157 | can take a while, since running the tests in parallel is incompatible | |
1158 | with GCC's coverage mode. | |
1159 | ||
1160 | After the tests have run you can generate a list of untested | |
1161 | functions: | |
1162 | ||
1163 | make coverage-untested-functions | |
1164 | ||
1165 | You can also generate a detailed per-file HTML report using the | |
1166 | Devel::Cover module. To install it do: | |
1167 | ||
1168 | # On Debian or Ubuntu: | |
1169 | sudo aptitude install libdevel-cover-perl | |
1170 | ||
1171 | # From the CPAN with cpanminus | |
1172 | curl -L http://cpanmin.us | perl - --sudo --self-upgrade | |
1173 | cpanm --sudo Devel::Cover | |
1174 | ||
1175 | Then, at the top-level: | |
1176 | ||
1177 | make cover_db_html | |
1178 | ||
1179 | That'll generate a detailed cover report in the "cover_db_html" | |
1180 | directory, which you can then copy to a webserver, or inspect locally | |
1181 | in a browser. |