Feature test inflections #2187
Feature test inflections #2187
Conversation
…terminus into feature/3.x-unit-tests
perl -pi -e 's/->setExpectedException\((\w+::class), (.+)\);$/->expectException(\1);\n \$this->expectExceptionMessage($2);/;'
…terminus into feature/3.x-unit-tests
| "curl -LSs https://box-project.github.io/box2/installer.php | php", | ||
| "mkdir -p tools", | ||
| "mv -f box.phar tools/box" | ||
| ], |
There was a problem hiding this comment.
I put this stuff here while trying to get the unit tests to run.. but I don't know that it's actually needed for them
| To run the functional tests: | ||
|
|
||
| 1. If you need the basics (php, git, direnv, etc): `brew bundle install` | ||
| 2. For code coverage: `pecl install pcov` |
There was a problem hiding this comment.
for what it's worth, this was required for them to run, at least when I first got them to run and I can't see why it would have changed.
|
|
||
| 1. If you need the basics (php, git, direnv, etc): `brew bundle install` | ||
| 2. For code coverage: `pecl install pcov` | ||
| 3. Prepare a new test site, e.g. make a new multidev in ci-terminus-composer |
There was a problem hiding this comment.
The tests currently have the environment hard-coded in most cases, e.g. LocalCommandsTest.php line 64, and line 81 as well. In the tests/Functional directory, when I run ack 'live' | grep 'live' | wc -l gives returns 29, and skimming the results, those matches look like the environment name in every case.
So even if you create a multidev, most of the time it's going to run against the hard-coded environment.
Obviously this needs to be fixed.
README.md
Outdated
| 2. For code coverage: `pecl install pcov` | ||
| 3. Prepare a new test site, e.g. make a new multidev in ci-terminus-composer | ||
| 4. Copy .env.dist to .env and make an .envrc to include it with `dotenv`, or copy .env.dist to .envrc and add `export` at the head of each line. Customize values as needed. Use `direnv allow` to enable automatic loading of environment variables needed for tests. | ||
| 5. Run `composer test:short` to run the short functional tests, or `composer test:long` to run everything. |
There was a problem hiding this comment.
composer test:long runs only the long-running tests, i.e. it won't run the short tests. This was so we could run all tests by running both commands separately. Until I changed test:functional to run both short and long tests, late last week, it ran all the tests, including the un-implemented ones which still fail.
There was a problem hiding this comment.
pushed update for this.
See how the changes made for the unit tests fare under the functional tests.