git/t/lib-httpd.sh, branch v2.2.0

Merge branch 'ep/avoid-test-a-o'

2014-06-25T19:23:56Z

Update tests and scripts to avoid "test ... -a ...", which is often more error-prone than "test ... && test ...". Squashed misconversion fix-up into git-submodule.sh updates. * ep/avoid-test-a-o: git-submodule.sh: avoid "echo" path-like values git-submodule.sh: avoid "test -a/-o " t/test-lib-functions.sh: avoid "test -a/-o " t/t9814-git-p4-rename.sh: avoid "test -a/-o " t/t5538-push-shallow.sh: avoid "test -a/-o " t/t5403-post-checkout-hook.sh: avoid "test -a/-o " t/t5000-tar-tree.sh: avoid "test -a/-o " t/t4102-apply-rename.sh: avoid "test -a/-o " t/t0026-eol-config.sh: avoid "test -a/-o " t/t0025-crlf-auto.sh: avoid "test -a/-o " t/lib-httpd.sh: avoid "test -a/-o " git-rebase--interactive.sh: avoid "test -a/-o " git-mergetool.sh: avoid "test -a/-o " git-bisect.sh: avoid "test -a/-o " contrib/examples/git-resolve.sh: avoid "test -a/-o " contrib/examples/git-repack.sh: avoid "test -a/-o " contrib/examples/git-merge.sh: avoid "test -a/-o " contrib/examples/git-commit.sh: avoid "test -a/-o " contrib/examples/git-clone.sh: avoid "test -a/-o " check_bindir: avoid "test -a/-o "

Merge branch 'jk/http-errors'

2014-06-16T19:18:36Z

Propagate the error messages from the webserver better to the client coming over the HTTP transport. * jk/http-errors: http: default text charset to iso-8859-1 remote-curl: reencode http error messages strbuf: add strbuf_reencode helper http: optionally extract charset parameter from content-type http: extract type/subtype portion of content-type t5550: test display of remote http error messages t/lib-httpd: use write_script to copy CGI scripts test-lib: preserve GIT_CURL_VERBOSE from the environment

Merge branch 'jk/do-not-run-httpd-tests-as-root'

2014-06-16T17:06:05Z

* jk/do-not-run-httpd-tests-as-root: t/lib-httpd: require SANITY prereq

t/lib-httpd.sh: avoid "test -a/-o "

2014-06-09T22:53:41Z

The construct is error-prone; "test" being built-in in most modern shells, the reason to avoid "test && test " spawning one extra process by using a single "test -a " no longer exists. Signed-off-by: Elia Pinto Reviewed-by: Matthieu Moy Signed-off-by: Junio C Hamano

t5550: test display of remote http error messages

2014-05-23T19:43:51Z

Since commit 426e70d (remote-curl: show server content on http errors, 2013-04-05), we relay any text/plain error messages from the remote server to the user. However, we never tested it. Signed-off-by: Jeff King Signed-off-by: Junio C Hamano

t/lib-httpd: use write_script to copy CGI scripts

2014-05-23T19:41:50Z

Using write_script will set our shebang line appropriately with $SHELL_PATH. The script that is there now is quite simple and likely to succeed even with a non-POSIX /bin/sh, but it does not hurt to be defensive. Signed-off-by: Jeff King Signed-off-by: Junio C Hamano

t/lib-httpd: require SANITY prereq

2014-05-12T17:19:23Z

Our test httpd setup will not generally run as root, because Apache will want to setuid, and we do not set up the "User" config directive. On some systems, like current Debian unstable, Apache fails to start, and we skip the tests: $ sudo ./t5539-fetch-http-shallow.sh --debug 1..0 # SKIP web server setup failed $ cat trash*t5539*/httpd/error.log [...] (22)Invalid argument: AH00024: Couldn't set permissions on the rewrite-map mutex; check User and Group directives AH00016: Configuration Failed However, on other systems (reportedly Ubuntu 11.04), Apache seems to start, and then bails during our tests with: getpwuid: couldn't determine user name from uid 4294967295, you probably need to modify the User directive Child 12037 returned a Fatal error... Apache is exiting! This may be related to the pre-fork/threading model in use (note that the second one complains of the child dying). However, it's not even worth investigating; in either case we just want to skip the tests, and we already recommend against running the test suite as root. Let's just explicitly check this condition and skip the tests rather than expecting Apache to do the right thing. Signed-off-by: Jeff King Signed-off-by: Junio C Hamano

Merge branch 'jk/run-network-tests-by-default'

2014-03-05T23:06:45Z

Teach "make test" to run networking tests when possible by default. * jk/run-network-tests-by-default: tests: turn on network daemon tests by default

tests: turn on network daemon tests by default

2014-02-14T16:13:51Z

We do not run the httpd nor git-daemon tests by default, as they are rather heavyweight and require network access (albeit over localhost). However, it would be nice if more pepole ran them, for two reasons: 1. We would get more test coverage on more systems. 2. The point of the test suite is to find regressions. It is very easy to change some of the underlying code and break the httpd code without realizing you are even affecting it. Running the httpd tests helps find these problems sooner (ideally before the patches even hit the list). We still want to leave an "out", though, for people who really do not want to run them. For that reason, the GIT_TEST_HTTPD and GIT_TEST_GIT_DAEMON variables are now tri-state booleans (true/false/auto), so you can say GIT_TEST_HTTPD=false to turn the tests back off. To support those who want a stable single way to disable these tests across versions of Git before and after this change, an empty string explicitly set to these variables is also taken as "false", so the behaviour changes only for those who: a. did not express any preference by leaving these variables unset. They did not test these features before, but now they do; or b. did express that they want to test these features by setting GIT_TEST_FEATURE=false (or any equivalent other ways to tell "false" to Git, e.g. "0"), which has been a valid but funny way to say that they do want to test the feature only because we used to interpret any non-empty string to mean "yes please test". They no longer test that feature. In addition, we are forgiving of common setup failures (e.g., you do not have apache installed, or have an old version) when the tri-state is "auto" (or unset), but report an error when it is "true". This makes "auto" a sane default, as we should not cause failures on setups where the tests cannot run. But it allows people who use "true" to catch regressions in their system (e.g., they uninstalled apache, but were expecting their automated test runs to test git-httpd, and would want to be notified). Signed-off-by: Jeff King Signed-off-by: Junio C Hamano

tests: auto-set LIB_HTTPD_PORT from test name

2014-02-10T18:20:45Z

We set the default apache port for each of the httpd tests to the 4-digit test number of the test script. We want these to remain unique so that the tests do not conflict with each other when run in parallel. Instead of doing it manually in each test script, let's just set it from the test name at run time. This is simpler, and is one less thing to be updated when test scripts are renamed (e.g., when being re-rolled or when conflicting after being merged with another topic). Incidentally, this fixes a case where t5537 and t5538 used the same port number (5537), and could conflict with each other when run in parallel. Signed-off-by: Jeff King Signed-off-by: Junio C Hamano