Repairing wp-env’s Testing Database

Sometimes when I interrupt PHP unit tests in a wp-env environment I end up getting the database in a bad state so that the next time I try to run tests I get:

wp_die() called
Message: One or more database tables are unavailable. The database may need to be <a href="maint/repair.php?referrer=is_blog_installed">repaired</a>.
Title: WordPress &rsaquo; Error
Args:
        response: 500
        code: wp_die
        exit: 1
        back_link: 
        link_url: 
        link_text: 
        text_direction: ltr
        charset: UTF-8
        additional_errors: array (
)
✖ Command failed with exit code 1Code language: plaintext (plaintext)

I then have to re-remember what I do to fix this every time, so I’m blogging it so I won’t have to try to remember again.

First I try to repair the DB with WP-CLI by doing the following (very concise) command:

npm run wp-env run tests-cli wp db repairCode language: Shell Session (shell)

I then get the output like the following:

> wp-env
> wp-env run tests-cli wp db repair

ℹ Starting 'wp db repair' on the tests-cli container. 

tests-wordpress.wp_users
note     : The storage engine for the table doesn't support repair
Success: Database repaired.
✔ Ran `wp db repair` in 'tests-cli'. (in 0s 754ms)Code language: plaintext (plaintext)

But tests then still fail. But what did work for me is to reset the tests DB as follows:

npm run wp-env run tests-cli wp db resetCode language: Shell Session (shell)

Then tests start working again.

I guess there’s a way to achieve this by adding the WP_ALLOW_REPAIR constant in the test environment’s wp-config.php but for some reason the test environment doesn’t seem to be using the config located at:

$(npm run --silent wp-env install-path 2>/dev/null)/tests-WordPress/wp-config.phpCode language: Shell Session (shell)

I hope this is helpful to you (and me in the future)!