Check for the presence of the MySQL binaries before calling them #389
Conversation
There was a problem hiding this comment.
Hi @gwolf
I haven't yet tested this (I don't have any environments without these on to test). I will run the automated tests again once the changes are made, but might also see if anyone else is able to test the negative conditions. Most changes are about using FALSE instead of null, and making the bee_message strings translatable. Also a function docblock required.
Thank you
Use FALSE instead of null. Write a proper docblock for the introduced function.
|
Thanks! |
Aren\'t we a bit over-pedantic? 😉
|
All tests passed! \o/ |
There was a problem hiding this comment.
Sorry @gwolf - a bit more to do. After thinking about the name of the helper function, I've decided it would be better to go into includes/filesystem.inc. There's also a few tweaks with the function doc bloc and variable names.
Change variable names to full words instead of abbr Rename introduced function to a more project-wide descriptive one Move introduced function to the filesystem includes
|
Thanks @gwolf - this is looking good. I tested the negative condition by temporarily changing the defined executable in the Thank you for your patience in getting this over the line. |
Fixes #388
Patch to verify for the presence of
mysqlandmysqldump(andgunzip, as it was in the same command construction) before actually calling it.Thanks!