{"id":20399,"date":"2018-02-01T14:34:40","date_gmt":"2018-02-01T19:34:40","guid":{"rendered":"https:\/\/mjtsai.com\/blog\/?p=20399"},"modified":"2018-02-05T14:53:18","modified_gmt":"2018-02-05T19:53:18","slug":"things-i-wish-id-known-about-bash","status":"publish","type":"post","link":"https:\/\/mjtsai.com\/blog\/2018\/02\/01\/things-i-wish-id-known-about-bash\/","title":{"rendered":"Things I Wish I’d Known About Bash"},"content":{"rendered":"

zwischenzugs<\/a> (via Hacker News<\/a>):<\/p>\n

The difference between [<\/code> and [[<\/code> was another thing I never really understood. [<\/code> is the original form for tests, and then [[<\/code> was introduced, which is more flexible and intuitive. In the first if<\/code> block above, the if statement barfs because the $(grep not_there \/dev\/null)<\/code> is evaluated to nothing, resulting in this comparison:<\/p>\n

[ = '' ]<\/code><\/p>\n

which makes no sense. The double bracket form handles this for you.<\/p>\n

This is why you occasionally see comparisons like this in bash scripts:<\/p>\n

if [ x$(grep not_there \/dev\/null) = 'x' ]<\/code><\/p>\n

so that if the command returns nothing it still runs. There’s no need for it, but that’s why it exists.<\/p>\n

[…]<\/p>\n

Bash has configurable options which can be set on the fly. I use two of these all the time:<\/p>\n

set -e<\/pre>\n
exits from a script if any command returned a non-zero exit code (see above).<\/p>\n
This outputs the commands that get run as they run:<\/p>\n
set -x<\/pre>\n<\/blockquote>\n\n
zwischenzugs<\/a> (Hacker News<\/a>):<\/p>\n